ED 193 256

TITLE

INSTITOTICN

PUE CATSNOTE

AVAILABLE FROM

DOCUMENT RESUME

TM 800 570

Set Number One, 1980. Research Information forTeachers.Australian Council for Educational Research,Hawthorn.: New Zealand Council for EducationalResearch, Wellington.8074p.; Items 6 and 7 are excluded because of copyrightrestrictions. For previous Set see ED 178 531.NCZEB, Box 3237, Wellington, New Zealand (310.00: 2sets yearly)

EDFS PRICE MF01/PC03 Plus Postage.EESCPIPTORS Classroom Environment: Cost Effectiveness:

*Educational Assessment: Educational Environment:*Educational Improvement: Elementary SecondaryEducation: *Foreign Countries: IndividualizedInstruction: Instructional Innovation: ProgramedInstruction: Student Behavior: Teacher Attitudes:Teacher Made Tests: Textbook Evaluation

IDENTIFIERS *New Zealand: Teacher fxpectations

ABSTRACTTopics and issues relating to educational research

are presented in the form of separate news sheets or essays. A casestud) of the merits and deficiencies of Penrose High School's "WhanauHouse," is included as an example cf curriculum innovation. Anexploratory study describing expectations of New Zealand teachersdemonstrates how teacher expectations cause certain students to bestereotyped and treated differently. An economist looks at what itcosts society to invest in children's education, and what it pays.Some of the effects of school size on the behavior of pupils arereviewed. Two very different techniques to fit the worn to the child(student team learning and self-paced instruction) are described. Onearticle describes research findings on how successful InternalAssessment has teen in Oueensland's secondary school program. Asummary cf the techniques, theoretical basis, and changing nature ofprogrammed learning is presented, as well as its impact on teaching.Other miscellaneous iteis of information include a guide to making u?classroom tests, an analysis of textbook English, and a guide toclass nomenclature. (EL)

************************************************************************ Reproductions supplied by EDRS are the best that can be made *

* from the original document. ************************************************************************

4

setresearchInrormatoonfor teachers

number one 1980

set: research information forteachers, is publishedtwice-yearly by the NewZealand Council forEducational Research and theAustralian Council forEducational Research.

General Editor: Lyn RichardsAustralian Editor: Peter JefferyDesigner: Peter Ridder

Copyright ownership of thispublication, with the exception ofitem 7, is vested in the publishers.With the exception of item 7, theright to copy, reprint, photo-copy,etc.. is granted by the publishersto all schools and teachers, in theinterests of better teaching;simply acknowledge the source.All other people or organisationswishing to copy any part shouldapply to set, NZCER, Box 3237,Wellington, New Zealand.

New Zealand:Subscriptions: $10 per yearStudent subscription $8 per yearDiscount subscriptions for bulkordersTo: set

NZCER.Box 3237,Wellington.

Australia:Subscriptions: $10 per yearTo: set

ACER.Box 210,Hawthorn,Victoria 3122.

3

,-

Wit*iset-4-4-

Item 1(NZ)

Content SheetResearch information foreveryone interested ineducation: teachers,principals, students,lecturers, actively involvedparents...

Leaflets and brief reportsdesigned for private study,staff meetings, in-servicecourses, and discussion insmall groups.

Contents

\ t'

...,ni... we Aiel

set -:l

morose am me

nomair amk WO

Isetz--4

11111i I

goober am MI

norrOgg ang IOW riot

1

I0i'l1

1 Contents sheet

2 The Penrose Whanau Unit: A Case StudyRae MunroeLike open-plan classrooms the Wanau units nowbeing built demand more than just a change inwhere the children hang up their coals. Anexamination of hopes fulfilled and unfulfilled,

3 Teacher Expectations and ClassroomBehaviour Alison St GeorgeWhat do New Zealand teachers expect of theirpupils? Are their expectations culturallystereotyped? Do they treat some pupils differentlybecause of their stereotypes?

'4 The Economics of Children in the WelfareState Brian EastonAn economist looks at what it costs society to investin children and their education and what it pays.

5 Projecting Educational ExpenditureIan McGill and othersWhat does it cost lo educate one child? What will itcost in 1990? How will education cuts hurt? Are theynecessary?

6 Chips With Everything exciu 41 pdT nicies about computers and education, Ageneral a I of where computers it intoschools; an article describing one successful use ofa computer sinfulation in science; a light heartedaccounrelthe trials and tribulations of a ComputerAs listed Instruction expert in Cincinnati. -

7 w Many People Can a Young Child FeelSecu h? Peter Smith e xe foci edis it safe, emo try, to pass the baby from hand tohar91.4traleave it in child care, to take it to parties, tolea in hospital? An article from Britain,

8 Small is Beautiful? Jack CampbellSome of the affects on school size on the behaviourof pupils.

9 Class Size Rides AgainDoes the size of a class have an effect on theachievemeM of the children in the class? Theeconomic situation and new research techniques byGene Glass and Mary Lee Smith. make are-examination of the problem timely.

4

set MNMit (Ma1st tIMIC..

number one 190

Produced twice-yearly bythe New Zealand andAustralian Councils foEducational Research

Subscriptions: $10 petyear; Student subscript.or.$8 per year: from:set,NZCEIR.Box 3237,Wellington.

10 Catering for individual Differences Pew Power& John LongbottomTwovery different techniques to fit the wc. ,o thechild are described in this item* Student TeamLearning, and an example of sett-paced instruction

11 Assessing internal Assessment Ann Lamontand Bruce McBrideQueensland has had Internal Assessment for severalyears now. This article describes research on howsuccessful it has been.

12 What Ever Happened to ProgrammedLearning? Graham WagnerTeaching Machines, Programmed Texts. en-, andtheir contribution to our understanding of howlearning lakes place.

13 Assessing What They've Learned WarwickElleyA guide to making up classroom tests whit . arereliable and valid as welt as quick and useful,

14 Educational English As She Is WrittenCharles TaylorTextbooks as wen as leachers put unnece., i yobstacles in second-language learners' waysTeachers can learn from this analysis of texqu okEnglish.

15 A Guide to Class NomenclatureA 'ready reckoner' chart to help with such mysteritsas what an American book means by Grade 3.. tAustralian book by Year 3, and a New Zeala d bookby Standard 3.

s Item Ne xel v decd,,

Free copies of set Hems

Copies of many past set items are available on requestwhile they last.11 the item is available its numerical listnumber is shown here. (Sea the ser index, item 14 of setNo. 1, 1979,)

63 80 86 97 103 113 116 120 123 127 132 138 141

73 81 87 98 105 114 118 121 124 130 133 139 14277 83 93 99 109 115 122 125 131 134 140 143 148

Also available are copies of the following items from sotNo.1.1979:367810111214and set No. 2. 1979: 4 5 6 8 9 10 11 12 13 14 15

it .#61 1111 s .t 'gait

kit

s. 10.m.kopri It-

'40°.

I

5

444.. 14 1111 pi g.. et .

Item 2(NZ)

The PenroseWhanau UnitA Case Study

Whgnau: 1. v.i. Be born I muri tonu mai teneio tona whanautanga.

2. Be in childbed Ka rite nga ra ewhanau ai te wahine.

3. n. Offspring, family group Ka rangonae Rata e karakia haere ana mai tewhanau a Tana. Used occasionallyin tribal designations, as, TeWhanau a Apanui.

4. Family (modern) It is questionablewhether the Maori had any realconception of the family as a unit.

5, A familiar term of address to anumber of people. Engari, e te whankei aroha tatou ki tona auetanga.

Pronunciation: as in, The lad will gofar now as he has a good job.

Whanau: 1. v.i. Go.2. Lean, incline, bend down Kia

whanau tou taiepa- ka whanau ihote rakau,

Pronunciation: as in. The lad will havefun now.

The Penrose Whanau Unit

A Case Study

By Rae Munroe,Auckland Secondary Teachers' College

IntroductionThe original planning committee of the Whanau Housedesign stated its expectations in Secondary Schools forTomorrow. The expectations were that the House wouldbe 'more intimate', provide 'room for spontaneity', be'guidance centred', and 'give each pupil insights,knowledge and experience so that he can understandhimself and the people he lives and works with, in boththe smaller and the larger community; and to make themost of what he can do personally that is unique andvital to him', and further, that 'Dignity and self-esteemare essential to every human being, as is the feeling ofbelonging. Within the extended family of each WhanauHouse pupils will feel accepted and of value'.

The staff of the Whanau House built at Penrose HighSchool has expectations too. These they wrote about ina booklet early in 1977. This statement was more clearlyrelated to actual conditions in a New Zealand schooland can be summarised as:

General Educational Expectations

learning would take place in a 'variety of ways andenvironments':

learning would be 'a shared experience with ... fellowclassmates, teachers. parents and members of thecommunity';

pupils' social and emotional needs would be met 'in aguidance-centred situation';

the Whanau would be a place 'to which the pupilbelongs and with which he identifies'.

7 1,,

'V

Expectations for Curriculum Organisation andAdministrationStudents should have a contribution to make to

decisions concerning the curriculum andadministration of the House.

Students should be allowed as much flexibility aspossible in the choice of subjects including somereduction of the compulsory core and substitution ofcomplementary courses so a wide range of topicsmay be offered.

Students should have the opportunity within subjects toinvolve themselves in remedial or extensionprogrammes in addition to core requirements.

The facilities of the school and services of subjectsshould be available to the student outside thepresently accepted school hours.

How many expectations were realised?

There were constraints (both anticipated andunanticipated) to the autonomy of a Whanau Housestaff, and some of these made it impossible to realisecertain expectations. It would be unacceptable to judgethe efforts of the staff by the number of expectationsfulfilled. Penrose High School is a fairly typical urbansecondary school, so it was inevitable that there wouldbe a number of constraints on independence. Theschool had well established administrative routines andagreed expectations for the behaviour of pupils andteachers. The major constraints could be seen asassociated with: the school timetable; the school'sexpectations of teachers: the school's expectations ofpupils. Expectations only marginally affected were:learning, environment, sharing, guidance, and identity.Those affected directly were: curriculum andadministrative decisions, curriculum choice. additionalcourses within the core. remedial and extensionprogrammes, and extended hours.

Thus, the General Educational Expectations seemedthe more realistic and most likely to be attained inpractice. Those for Curriculum Organisation andAdministration, although theoretically capable ofrealisation, were the most constrained by the existingconditions in Penrose High School.

Three main themes

The findings of the two-year case-study of the unit canbe seen as having three main themes which stem fromthe expectations: The Whanau Unit as Community;Teaching and Learning in the Unit; and, The Unit asDemocracy. The evidence upon which thesecommentaries are based comes from tape transcripts ofinterviews with pupils and teachers (inside and outsidethe unit), staff and staff-pupil meetings and seminars,classroom activities, and staff meetings with parents.

The Whanau Unit as Community

As soon as the leachers and the pupils entered the unitin April 1977 their comments reflected their belief thatthe Whanau House provided new opportunities for closeassociation between pupils and between teachers andpupils. The comfortable conditions, the intermingling ofteachers and pupils of all ages, and the relaxed andcasual relationships were noted by all teachers andpupils during the intitial round of interviews.

A community developed and was seen in the extent towhich pupils were 'looking after' their new environment.Care of equipment and furnishings was one constantperceived by all teachers and pupils during the two-yearperiod of study. Although the standard of care variedsomewhat from season to season, depending on thenumber of people using the building, teachers agreedthat the pupils' attitude towards their environment wassuperior to that of pupils they had knowr in other schoolconditions. This opinion of Whanau staff members wasfrequently confirmed by visitors to the unit. On thewhole leachers believed that the pupils' attitude was afunction of their sense of identity with the building andwith the people in it and this was confirmed by mostpupils.

The range and the relaxed nature of friendshippatterns among pupils were other aspects of communitylife commented on. However, sixth formers werereluctant to identity with the new unit and thisreluctance persisted throughout the first year. In thesecond year the problem was less noticeable becausethe sixth formers had already had time to adjust to theunit.

Perhaps the most notable development in friendshippatterns was the mixing of different age groups in the

_'a

commons. Teachers often contrasted the behaviour inthe Whanau commons with their experience in the mainschool where 'juniors and seniors simply don't mix'.Nonetheless. teachers differed in their explanations ofthe change. Most attributed it, in part. to the building,but some saw the vertical form arrangement as asignificant contributor to the mixing of age-levels, (yetthis organisation was shared with the main school). Thepupils themselves believed the mixing was wholly due toconditions in the commons 'where you naturally get toknow everyone in the House'. In all, both teachers andpupils saw the socialising of pupils in the Whanau assignificantly different from that in the main school. Theyperceived that barriers had been broken down for pupilsboth at their own age levels and between age levels.

The relationship between teachers and pupils wasanother sign of the Whanau House as a community.Inevitably teachers found that pupils had more access tothem and although they all considered this to be adesirable trend there was sometimes an ambivalence intheir comments. A few teachers, however, did not feelthat the nature of their relationship with pupils hadchanged but simply that their own access to pupils waseasier. Those teachers who did perceive markedchanges in their involvement with pupils, nevertheless,had misgivings which in the initial stages weresomewhat masked by the public commitment to persistwith the goals established during planning. In the event,teachers, initially, made every effort to increase timespent in the unit while seeking to balance involvementagainst the desire to rest and be 'secluded'.

Despite high expectations for the development of apositive social climate in the early days and despiteobvious gains in this area, by late in term three of thefirst year there were obvious signs in the teachers'comments that the experience had left them somewhatdisillusioned. The changed emphasis can be traced inthe transcripts of staff planning sessions. By Novemberwhen the whole staff met for a day to assess progress,discussion tended to focus on some of the negativefeatures of the year's experiences. In particular, theteachers' constant exposure to pupils had become anirritation to some of them. Given that the main thrust .,fthe Whanau was to reduce the psychological distancebetween the teacher and taught, the reality proved to betoo taxing for those teachers who were used to greater

9

privacy in their day-to-day work at school. l'he lack ofprivacy in the staff planning centre led some staff tosuggest that teachers in a Whanau needed less classcontact time with pupils so they could cope with theadditional deMands on their time outside class hours.

In summary, the evidence suggested that for mostpupils the unit promoted a very satisfying socialenvironment which allowed for the establishment ofwide-ranging friendships with peers and for increasedaccess to teachers at both the personal and academiclevel. For some teachers the environment providedequally satisfying relationships, but for most the closesocial interaction resulted in varying degrees of fatigueand strain.

Teaching and Learning in the Unit

How did teachers perceive the effects of the move to theWhanau on classroom events? Their comments lay on acontinuum from 'nothing's changed' to 'a completelynew experience'. Some teachers had been able to usethe facilities in the new building to advantage inextending a particular style of teaching. The extent towhich this occurred varied between subjects andteachers but it was only a minority of the teachers whofelt their teaching style had been markedly affected bythe move.At the same time there was a general consensus that thebuilding was pleasant to work in, that this tended tomake classroom activities more relaxed and that as aresult pupils were better behaved. Again, the flexibility ofthe building was claimed as the main factor affectingchange. The new environment, providing as it didcarpeted areas, attractive decor, and easily movedfurniture, was very much appreciated by those teacherswho were promoting creative activities, group andproject work, and informal discussion between pupilsand themselves. However, a few teachers wereconcerned that flexibility in the unit had led to a certainslackness. Clearly the more relaxed routines did not suitall teachers equally. It was inevitable that those teacherswho, for example, released pupils before the end of aperiod to go into the commons. were seen as erodingobedience to bell times, Other teachers who were morepunctilious in this matter were upset by the apparentoffhand approach of the pupils. 'Whanau pupilswandering about in corridors during class time' were

frequently noted unfavourably by main-school teachers.Thus, in the early months. the main leaching issues

commented on by teachers related to the effects of thebuilding and consequent changes in routines; exampleswere the fact that time was saved between periodsbecause pupils and teachers did not have to 'walk downmiles of corridors', the convenience of dividing classesbetween rooms, the ease with which teachers could findpupils, and the more mature behaviour of pupils. Somecomment was made about curriculum development butlargely in regard to the functioning of innovations whichhad been introduced before the unit was occupied.

For the individualised mathematics programme in thethird and fourth forms the accommodation was thoughtto be ideal and the double activities room wassuccessfully adapted for this work by the twomathematics teachers. The Australian ScienceEducation Project's modules were introducedprogressively in the junior science classrooms as thematerials became available and interdisciplinary studieswere begun for the third formers. Predictably, by thethird term of the first year more detailed comment on thefunctioning of the various curriculum innovations wasbeginning to appear in interviews and in staff planningsessions.

Perhaps the single curriculum innovation whichinvolved most planning. and could be said to typify themain thrust of innovation in the unit, wasinterdisciplinary studies. From the outsetinterdisciplinary studies was intended to embody threemain intentions: lo provide an opportunity forintegration of core studies in junior classes, to allow forthe full teaming of the Whanau teachers, and to givepupils opportunities for independent or self-directedlearning. As well, the interdisciplinary studies weremeant to alternate in blocks of three or four weeks withordinary classwork in English. Social Studies, Scienceand Mathematics. Themes had been prepared at theplanning stage and detailed work sheets with questionsequences were designed before the studies werelaunched. The first session was held late in term two ofthe first year with third formers. Two further sessions,one third form and one fourth form, were held in thefinal term.

By the end of the first year there were differencesbetween staff about the value of the studies, there was

1U

an ambivalence in the response of the pupils, and anegative attitude to the studies by some parents andsome staff members in the main school. These allcontributed In some measure to a reduction of impetusfor the innovation. As welt, staff members found itincreasingly difficult to find time to do the detailedplanning necessary. In the second year only one thirdform session was held. By the end of that second yearsome teachers had reflected on the apparent flilure ofthe innovation and commented on what they saw asshortcomings in practice.

The effect of the timetable structure was frequentlymentioned as a major problem, and there is little doubtthat it was a factor in the run down of interdisciplinarystudies. So closely was the Whanau unit timetabled withthe main school that any development such as thisneeded protracted negotiation, not only with mainschool teachers likely to be affected by Whanau pupilsleaving the school during classtime for study in thecommunity, but also for satisfying the main school'sregulations for pupil's movements inside and outsidethe school.

Teachers perceptions of teaching and learning in theWhanau did centre on some common ground but therewere also marked differences between their viewpoints.The teachers shared the opinion that they were in thepresence of a more informal teaching environment thanpreviously, that relationships with pupils had ')ecomeLess distant and that pupils had more ready access tothem for special attention and tuition. However, opinionvaried about whether or not pupils were learning morethan they would have done in the main school and therewere sharp differences in their perceptions of the extentto which curriculum change had actually taken place.

Comparison between 1978 School Certificate resultsof Whanau pupils and those of a matched group in themain school showed no significant difference inperformance. However, Whanau pupils did pass, onaverage, a greater number of papers and this differencewas statistically significant. On balance, it could beconfidently claimed that the pupils who entered theWhanau unit were not academically disadvantaged atthe fifth form level.

The Unit as DemocracyAt the planning stage. participation by pupils was

11

regarded as a major goal for the Whanau unit and earlyin the occupation of the building efforts wore made toset up a House Committee which would encourage thepupils to participate in the affairs of the unit. However,the Whanau staff members had two clearly differentpositions about pupil-participation. One group acceptedthe original notion from the planning stage that formatstructures should bet provided; the other grouppreferred the participation to emerge on a day-to-daybasis built around teachers' contacts with individualsand small groups. Most discussions about participationwere about the pupils involvement in the generalrunning of the unit. However, a second aspect ofparticipation, which had featured in the planning stage,was the pupils' contribution to curriculum development.This was little mentioned by teachers either in interviewsor in meetings, and systematic feedback from pupils oncurriculum matters was never a prominentconsideration in the unit.

In all, the formal structures of vertical forms, HouseCommittees, House meetings and forums (heldinfrequently, on Friday afternoons), were not seen bystaff as fulfilling the high hopes held for them. TheHouse meetings and vertical forms were seen tofunction successfully for administrative purposes in theschool, the checking of absences, the communicationof routine notices and for teachers to tell pupils aboutcurriculum matters, field trips, school rules and the like.But they were not seen as providing many opportunitiesfor pupils to participate either in the running of the unitor in the selecting and organising of the curriculum.Some teachers believed there were informalopportunities for pupils to participate in these ways butthere was no consensus about the extent to which theseopportunities existed for all pupils.

ConclusionIt is not possible to determine how much more could orshould have been done to attain the goals of theWhanau. At the planning stage it was clear 'fiat thehighest priority was placed on curriculum innovation.Innovation in consultation and decision-sharing, eitherbetween teachers and pupils or between teachers andparents, never engaged the full attention of the staff.Given the constraints noted, the situation that

developed was not surprising.There is no doubt that the Whanau staff worked

extremely hard during the two years of this study andthe range of their activities, both inside and outside theunit, was impressive. However, whether or not the goatswould have been more successfully attained had therebeen additional staffing and therefore less contact-timefor oath teacher, is problematic. There is abundantevidence in the transcripts of the teachers' views thatthere was no clear consensus about which goals shouldhave the greatest priority. Agreement was greatestabout the social goals of the unit, but even here therewere thot,e who would have preferred less contact withpupils and greater protection of their privacy. In goalsassociated with teaching and learning, some teacherstook and maintained an innovative stance while othersdid not. However, in the twin concerns ofpupil-involvement and parent-involvement in therunning of the unit there was considerable ambivalencein the views expressed and no systematic plan of actionwas adopted during the two-year period,

The course of events in the unit illustrated quitegraphically the effects of unanticipated constraints onan educational innovation. A degree of autonomy hadbeen assumed which could not be realised in practice.Without such necessary independence the Whanau staffhad been presented with the impossible task of literallyrestructuring secondary schooling. The extent to whichthey succeeded, particularly in attaining the socialgoals, was clearly attributable to their commitment andtheir determination to persist in the face of considerablefrustration,

NotesDepartment of Education. Secondary Schools for Tomorrow.1975.

Penrose High School. Whanau House. 1977,

Rae Munroe is Dean of the Secondary Teachers' College.Auckland and was Chairman of the PPTA Curriculum ReviewGroup which produced the important book Education inChange In 1969.

The dictionary extract at the begining is from A Dictionary ofthe Maori Language, by H.W. Williams, Seventh Edition, TheGovernment Printer. 1971. slightly shortened, Thepronunciation guide is by the editor.

i, 1 2

A Guide to Class NomenclatureAges, Stages, Classes, Years, Grades, and Systems compared

Approximate Chronological Age

1 2 3

Australia

Item 15 set--wolooewat.ontow twoenrws

number one 1980

Note: This chart gives a roughguide only. All systems arechanging and in all cases thechronological ages areapproximate.

N.S.W.VIC.

TAS.A.C.T.

S A0

N.T.

W.A.

i

KindergartenPreparatory

iKrody PrepKindergarten

Reception

134cb *13 2

Kindy

Play Groups1- 3.

L Kindergarten

Year2

s

Primaryi

Primaryi

Year3

1

Year4

Year5

Year6

i1

-I-I

Year7

Year8

Year9

I

1 1

Year10

Year11

Year12

High School

Li i

High School

Primary

]

IJunior Primaryi

Primary

1

1

{ High School

i

Matriculation Col.Secondary Col.

i

i

Senior High School1

I High SchoolI

Tech.. Further. Advanced Cols. of Ed

I fMost States have some State Preschools. Most States have Rural or Area Schools Crossing Usual Primary/Secondary Breaks.

UniversityI

New Zealand

13[

Playcentres j

Junior Classes

Std2

Std3

Std4

Contributing Primary

Form1

1

Form2

Form3

Form4

Form5

AIM

Form6

Form7

Form One to Seven High Schools

Intermediate Technical Institutes

Full Primary

IArea Schools. Mainly Rural

econ ary c oo

I 1 I

[University

f I

J.

United States and CanadaKindergarten

1..

Kindergarten----- --___-C--__-.1Nursery School

.1

]

Grade 1 Grade 2 Grade 3

i

ISix Grade Elementary

Grade 4 Grade 5 1Grade 6

I Eight Grade Elementary1

Grade 1 Grade 8 Grade 9

Junior High School

rade 11 radell Trade 12

Senior High School

JuniorSenior High School

1IFour Year High School

I

- -----1-------[university

-.F-.-.-

I

[Junior college

United Kingdom ni1H-1 1--1ni r--1 1-1 1--1 i i 1----1 1--1 r---7Each Local Education Authority has its own rules. Scotland is different again. Generalisations are misleading.

England and Wales i II II III 11_11 I LI I___I I__IILIILILIILAI

1 6th Form College

INursery

Primary

Plowden Recommended[Nursery ] 1 Infant

1

1

Scotland 1 11 Primary

Nursery I

1.0 1 II

I

Junior

I Middle

I Secondary

II

I Secondary

1

Junior High School

1

1

1

fI

Secondary

ISenior High Schooli

[ Secondary

16

Item 14 set...tee INChOM

number one 1980

I

e. 0 I1k

I

By Charles Taylor, Sydney University .

Sue Price

What can a teacher, faced by the problem of helping immigrant second-languagepupils in his or her class, learn from recent research? One important researchproject has made us aware of the sort of obstacles which books, as well asclassroom talk, are throwing in the path of the immigrant pupil. In this item isdescribed the EMEC (English for Migrant Education Curriculum) research projectand conclusions are drawn for everyday teaching practice.

17

Educational English as She is WrittenBy Charles Taylor, Sydney University

What is the English of school textbooks like?Impressionistic statements such as 'Far too hard'.'There is a shortage of carefully graded texts', 'Notenough solid meat in them these days', are not goodenough. This research project, which looks at what isactually read in NSW schools, began in 1975. Itpaused in 1978 and these are some of the results.

The Methods

All 669 schools in NSW which have courses in years 7to 10 were sent a one page questionnaire. Fromsubject teachers came information about the mostfrequently used textbooks. A cut off point of 18well-used books gave us 700 000 words of runningtext in six subjects: science, mathematics. history,geography, social studies. and commerce. This wasthe material analysed in 1975-8.

The analysis was computer assisted and looked atsyntax (sentence structure), semantics (meanings)and vocabulary.

The Questions

We set ourselves the task of answering suchquestions as:

What is written English like, especially what is itlike for non-native readers?

Most obviously it is not like the exercisesihat arewritten for foreigners learning English. Thoseexercises usually consist of single unrelatedsentences. Understanding them requires effort for thelearner but it is quite a different art from fluentreading. Where a text has been simplified forforeigners it has usually been done by chopping uplong sentences into short ones. Yet a string of shortsentences is harder to understand than a group ofsomewhat longer ones written so that therelationships between the parts are made clear bysuch phrases as 'Nevertheless', 'In the meantime',One thing led to another and .....

The problem for the reader is making the rightconnections. Suppose for example that your eye hadcaught

'the reader is making'

out of the last sentence. That is a fairly good readingspan bigger than most second-language users inyears 7 to 10 could manage. But of what value is it toread those words unless you have the correctdivisions between them and the correct links with therest of the sentence?

The Readers

We found that non-English-speaking learners differ inmore than one way from long-resident Australianlearners. They tend to respect the textbook more:they tend to give up trying to understand classroomtalk; they are shy of spoken interaction in class: and

they find that using books gives them more time tounderstand for themselves without involving them inawkward social problems. For these reasons, andalso because more research can be done cheaplywith written text. we confined our attention to thepossible problems students who have English as theirsecond language would be likely to encounter in theirreading.

The Analyses

Our first batch of analyses concerned syntax theway sentences are held together. We noted thatsingle sentences avoid the problems that can arisewhen we make a narrative or an argument flow along

their meanings tend to be self contained. Alsosingle sentences are seldom dependent on somesituational or psychological Context.

The pen is on the table.The car is in the garage.The pen is in the box on the table.

These from a typical English-as-a-second-languagecourse. They are not being taught for the content theycontain, but for the rules they exemplify the use ofthe prepositions 'on' and 'in'. They are specific too.Subject textbooks contain very few such sentences.

When you have observed a white precipitateforming remove the beaker from the heat using aheat resistant cloth. stir, and filter the mixtureusing the apparatus you have already set up.

That is a much more common type of sentence, andas you can see it is long. full of connections such as'when' and it refers to a particular situation in alaboratory.

What is a complex sentence? There is a wholeliterature of attempts to describe complexity, but thenotion that emerged from this study was that itconcerns 'interruption'. We used a simpleinterruption index using long interruptions ofcontinuity as the mark of complexity.

'Using a heat resistant Cloth'in the science sentence above is an example of whatis. for a second-language learner. a long interruption.Being able to move along a page from one eye-fix tothe next and retain the meaning depends on yourknowledge of grammar and vocabulary. Vocabularytells you what the paragraph is about and grammarwhat it says about it.

We begin by asking what is interrupted. Usually aparticular line of argument. It is usually interrupted totake in some supporting evidence or an implication.

William and Mary, much harrassed by theMonarchists on one side and the Parliamentarianson the other, pursued a middle course.

The next question is how long must the reader holdpart of the argument in suspense. Here we need helpfrom the psychologists. They tell us that a wait of

Is

more than 1.1 seconds is a bit too long for properunderstanding.

This short-term memory span covers reading up toten words. That is, if you're an efficient long-residentnative reader. But if you're a recent arrival, whodoesn't know a lot of English, or if your family doesn'tencourage you to read English, you'll miss the boat.That means, the 1.1 seconds has gone before you cancatch the drift of the whole sentence. The first bit youread got lost before you finished getting the secondbit.

What happens is this. There are invisiblerelationships between parts of sentences. These arepartly structural, i.e.. they hold sentences together.But sentences are also tied together even when theyare separate.

Look back at the paragraph above, and you'll find aword these. The word these refers to invisiblerelationships, even though that phrase is in anothersentence. We call all these links 'bonding', and it'sbecause good readers get the bonding links right thatthey know what the sentences mean. Some peoplestudy grammar in school still, but the links grammardescribes aren't enough. When we read a continuousbit of writing, we need to process all the links. Thatincludes what linguistics call 'cohesion'.

VocabularyWe looked at an idea called 'collocation'. This vogueword means the closeness of one word to anotherword in a passage. Take these sentences:

The cat was basking in the sun.Cardinal Wolsey basked in the warmth of Henry'sfavour.

You notice that the rather rare word bask has to gowith some word suggesting heat. So we collect allthese words that go together and come up with acluster of words. These words form a sort of topicpool for vocabulary. You can try this for yourself onsome geography book, for example. Notice how eachword has other words that seem to be demanded forcompany. Then collect them into a list, and you've gota vocabulary base for practice material in English.

In our study we used a measure which we called aspan. Count the words from the keyword the oneyou're investigating. Going from bask in thesentences about the cat and Henry, you find sun andwarmth are both 3 places away. They are 'at a span of3'. We used a span of 4 for our research, and you canread some of the results in the Report.

The study also covered word frequency in differentsubjects and it tried to identify words which do a tot ofwork. For example, a question like, 'Who won?' saidaftera soccer match can throw people whose Englishis weak. This innocent looking question really means:

'Which team scored the most goals, and thereforethe winner?'

Then we found that innocent words like cut andarea have different meanings in different subjects. Ingeometry, one line cuts another. But the line isn't nowin two pieces. Cut really means 'cross' in geometry.The area of a triangle is the amount of space in it. Anarea in geography is just a region. So, 'find thearea ...' What does it really mean? All you nativeusers of English reading this think this is child's play.But not so the second language users.

GrammarWe all hate grammar, but we all use it, or else wecouldn't get our ideas across. We looked at grammar,and especially at verbs. Verbs are the key to asentence, Each verb has a different meaningaccording to the nouns that go with it.

In order to show the funding authority that we werescholars and not just playing around, we called thissort of thing 'transitivity'. As far as teachers go, it'sreally just putting verbs and nouns together. Take theword find:

(a) Stuart found a lake at the foot of the hill.(b) Find the value ofx in the following.In (a) becausefind goes with the noun 'lake', we

know it means 'discover'. In (b), it goes with 'value'. soit now means 'look for'. Keeping those meanings intheir right place is a hard game for new readers inhigh schools.

Language for What?Is there a special language for science, and one formaths and so on? We didn't find so much of a speciallanguage for subjects, but we did find that there wasa sort of educational language that tries to beobjective.

Objectivity can be achieved by using:(i) the passive voice: the island was discovered

in 1856(ii) -ing and -ed words: boiling the water;

attached to the top(iii) the 'always' tense: water boils at 100°C(iv) introductory& it is interesting that...(v) introductory there: there are many farms In

the area(vi) forms with what: what is needed is. ..

(vii) verbs like may, could, must: this may be thecause of...; there must be some reason

ProblemsThe passive voice in English has two meanings.Look at:

(a) The solution is placed in the beaker.(b) Balloons are filled with light gases.

In (a). the passive really means:(i) Put the solution in the beaker

or(ii) It you put the solution in the beaker. (the

experiment will go well).

19

But in (b) it means:People usually fill balloons with light gases.

Verbs likeseem and think often come in funnyplaces in English. Thus:

(a) She seemed not to have heard(b) I don't think that's right.

In (a) it's not she who seems. What we mean is: itseemed as if she didn't hear, Foreigners find thisreally queer. In (b) its not that I don't think. I dothink, and I think that isn't right.

Reading spans we've already mentioned. Theyare a problem because readers get the wrongwords together in a passage and so invent newexpressions. In this sentence:

What they found of importance was negligible.Some readers invented an expression 'find of'. Butof doesn't go with find, it goes with importance.Some conjunctions like so that can introducedifferent ideas:

(a) Place the compass so it faces north (in sucha way)

(b) Pull the cord so that the parachute unfolds(with this purpose)

(c) He stopped suddenly, so that the baggagefell off (with this result)

One innocent letter we should all watch closely isthe letter 5. Both verbs and nouns use it in English:

(a) The mist clouds our vision.(b) The clouds moved slowly.(c) The cloud's in the way.(d) The cloud's base was fiat.The comma above the line, the apostrophe, helps

us a bit, but even it means two different things here.and we still need to think so as to sort out themeanings.

SimplificationWe tried to measure what was simple. We used

the idea of lexicality and lexical density (see thenotes) and linguists' ideas like clause depth andclause interruption mentioned earlier.

We noted that simple words were often morevague. So there's a need In all subjects forspecialized vocabulary. This helps sort out theexact meaning.

Some QuestionsAre all high school teachers English teachers? Dosubject teachers overlook language problems. orhope the English teachers will sort them out?Should we use students' own language as the basisof textbooks, or does this make it impossible toadvance further? Is it really possible to teach theEnglish of geography 'situationally'? (i.e.. using areal situation, 'the chalk is in the box on the table'often used by English teachers). Is teachingvocabulary enough?

It seems as if many of these questions are

2U

answered in part by bearing in mind how we read.As we read, we have expectations that thesentences will be completed. Reading isn't just apassive occupation. If our expectations aredisappointed, we get contused. Therefore we mustfind out exactly what we ought to expect in a schooltextbook. And that's what the EMEC project tried toshow.

An ExampleTake the sentence:

The critical solution state of the ore is relateddirectly to the composition of its crystallineelements.

What is it about? The keyword is ore, but that wordis embedded in the sentence at a point where somereaders wouldn't expect it. In particular, foreignspeaking readers might not notice it. It's a smallword. And yet it's the right place. Our students needa lot of practice in this sort of English.

Students need practice in learning:(i) to find the links between words.(ii) to discover which vocabulary words go

together.(iii) to understand objective language.(iv) to identify keywords in sentences and in

passages.

NotesSome of these references are rather technical, butthey repay study:

About Cohesion:M.A.K. Halliday and R. Hasan: Cohesion in English,

Longman.1976.

About Context, or how much work a word has to do:B. Bernstein: 'Social class, language and

socialisation' in Open University: Language inEducation, pages 102-110. esp. page 106.

About Lexicality and Lexical Density:J.N. Ure: 'Lexical density and register differentiation'

in G.E. Perren and J.L.M. Trim (eds.): Applicationof Linguistics. Cambridge University Press, 1971.

About the EMEC Project:C.V. Taylor: The English of High School Textbooks -

Education Research and Development CommitteeReport No. 18, Australian Government PrintingService, P.O. Box 84. Canberra. A.C.T., 2600. andobtainable at Government Bookshops .

The AuthorDr C.V. Taylor is Director of the Department of

Education's Diploma in the Teaching of English asa Foreign Language at Sydney University.

Assessing WhatThey've LearnedBy Warwick B. El ley, University of the South Pacific

AMinrINIMENIMMINNIMIEMENIPrimMEM"MMIMEMMIlrillAMPIMENVAM=Mi

NAWINSIrgialMMEOIN111110M111=11MINE'21111111MMINIIIMNIIIIII4MUMMEMEMMENIMMEWAMIMMMUNIM ME INIMMEMMEIIMM

Sue Price

1. IntroductionWhy do we test pupils?Most teachers spend a lot of time and effort preparingtests. giving tests. marking tests and using tests for onepurpose or another. Why? The main functions of tests andexaminations in school are:

(a) Mastery. Classroom tests are often prepared byteachers to see whether pupils have mastered a particularunit or skill that has just been taught.e.g.. Teacher gives a quick quiz to see whether pupils havelearned how to multiply fractions: or know the main eventsand characters in the book "Animal Farm": or havelearned the main vitamins and the foods that contain them.

(b) Diagnosis. Tests are often used by teachers todetermine the major weaknesses that a child shows inbasic skills.e.g.. In mathematics. to see whether pupils know basicnumber facts. or can handle zeros. or understand how todivide fractions; in reading. to see whether pupils'difficulty is due to poor vision or hearing. or word attackskills. or vocabulary weakness, or some misconceptionsabout print.

(c) Reporting Progress, Formal examinations are oftenused to report on the progress made by pupils over a term.or school year. The results are of interest to parents.pupils. potential employers, and teachers in next year'sclasses.

In addition to these three main uses, formal tests andexaminations are often used

(i) to place children in ability groups:(ii) to select pupils for further education.

scholarships. etc.;(iii tp match pupil materials with pupil abilities:

(iv) to evaluate one's own instruction:(v) to assist with vocational guidance:(vi) to determine a child's readiness for learning:(vii) to undertake research on pupils' abilities. or on

teaching methods:(viii) to determine whether educational standards in a

school. or total system are changing.

As teachers, we should be clear why we are testing. Weshould not test just because it is always done. Certainly.tests often do help motivate students to work harder. Butthe results can be discouraging too. if the results are poor.

In fact. the purpose of the test should affect the kind oftest given. Thus, a formal selection examination willnormally cover a large number of skills and topics lightly.and most questions should be of middle difficulty level.Mastery tests on a panicular unit will he more intensive,with several questions on each ore few topics to seewhether they are mastered or not. Some tests should berelatively hard for all pupils. (e.g.. Diagnostic tests). Somewill be lengthy and formal with elaborate marking schemes(e.g.. End-of-year examinations): at other times theteacher will give short, informal quizzes with emphasis onquick feedhack. Sometimes the need is for many shunquestions which can be objectively marked. At other timesthe teacher's purpose will be better served with a few longanswer questions. It is imponant therefore that we thinkthrough our reasons for testing.

2. Qualifies of a good examinationNot all examinations are well set. Often they :ire too hard.or too easy, or not enough time is allowed. Sometimes thequestions are vague, or trivial, or provide clues for the'test-wise' pupil. Many examinations are unbalanced,providing too many questions on some aspects. and too

Item 13SetNOMINI

number one 1980

few on others. Such weaknesses means that decisionsbased on the examination results will he unfair, orpedagogically unsound. How can we tell if ourexamination is a good one? Two important features areReliability and Validity.

tat ReliabilityTests are reliable if they product:cm:fix:ear results. if theyproduce similar marks on different occasions. If a pupilgains 75%. in a reading comprehension test today. andonly 50% tomorrow, then the results are not consistent;the tests are not sufficiently reliable to base judgementson. Ira pupil is placed first in his class in a test ofmultiplication and division of decimals today and is 20th init similar test tomorrow, we can conclude that the tests arenot reliable indicators of his ability.

To be reliable a test must normally be /mg enough tominimize the effects °reliance factors in the content andskills included in the test. With a short test, a pupil may helucky, because he happened to know or guess correctly thefew questions that were asked. whereas he knew very littleabout the areas untouched by the test, A standardized lestof reading. mathematics or language normally needs at least40 good objectively marked questions. to reach asatisfactory level of reliability. To make decisions aboutindividual pupils. for placement. or grouping. or diagnosis.a teacher-made test will probably require more questionsthan this; for judgements about the performance of wholegroups. a teacher can get by with fewer.

Just how long a particular test should be. depends on thetype of material tested, the amount of supplementaryinformation available, and the imponance of the decisionsbeing made. Thus a test of a highly specific skill. such asarithmetical addition, or spelling, or typing. may producereliable results within ten minutes. if however, we wish toexamine a pupil's grasp of a variety of mathematicalrelationships, or his understanding or a period of history,and then to make decisions about future schooling on thebasis of the results, we may wish to extend the test overtwo or three hours to gain masuntim reliability. For suchgeneral skills as essay-writing ability. or oral expression. itis commonly fouad that pupils vary so much in theirperformance from day to day and from topic to topic. thatthe only sure way to gain adequate reliability is to test the

22

pupils on several Occasions and several topics. andcombine the marks given by two or three independentmarkers. This may not always be practicable, but weshould realize when oar results are likely to he fallible.

Other requirements of a rani& test are clear, precisedirections. and reasonable time limits. If students arerushed, their performance may not he typical. Thequestions should be clear and unambiguous. neither tooeasy nor too difficult: they should normally discriminatewell between good and poor pupils, and they should hecapable of reasonably objective marking. Otherwise theresults will vary according to the values and whims ofmarker. If a choice of questions is allowed reliabilityusually drops, because markers cannot compare answersso consistently.

These are some of the more important faetors indetermining how reliable a test will he. It is possible toassess the reliability of a test statistically, but that is a topicfor another lime.

(h) ValidityA good test must he valid. This means that, in addition tomeasuring a pupil's achievement consistently, it should herelevant to the Main Objeriire ofthc course. It shouldcover the unit or course adequately, sampling each contentarea and skill in appropriate proportions. I f a teacherknows precisely what his objectives are, he can usuallytell, by analysing the questions of a test, whether theyconform closely to the objectives he has adopted i.e..whether the lest has 'content validity' for his purposes.

To illustrate, a 60-item test of addition in arithmetic maybe highly reliable, and yet be quite invalid for measuringaehievment in a course of modern mathematics whichemphasizes concepts, relationships and reasoning. Theobjectives of the, test do not match the teaching objvctives.A 3-hour written examination in manual arts may givereliable results. But if it does not require students to showthe actual skills they have learned, it will have poorvalidity. The students who do well on a writtenexamination may not he those who do well in the practicalskills.

Again a test of geograph y which focusses on factual]details about populations, areas, climate, exports, capitalcities, and the like would produce irrelevant results for ateacher who stressed broad concepts, generalized skillsand underlying relationships. A valid test of suchobjectives may require novel or fictitious situations onwhich to base questions so that a pupil can demonstratethat he has attainernse objectives, regardless of the

particular factual details he has actarcd.Sometimes tests lack validity because of 'eultural bias',

Questions may be unfair hecause they assume that thepupil has had particular experiences w hich he has not had.or read hooks which were not accessihie to him, or seenfilms or TV. programmes which he has not seen.Sometimes the test may he invalid because the print isillegible, or the diagrams unclear. or the paperinadequately proof-read. Such problems may distractpupils and cause changes in their rank order. Likewise,cheating will result in invalid results. If pupils van copyone anothers answers. or gain prior knowledge of thequestions the results will not reflect their true grasp of thesubject assessed,

To ensure maximum validity for his tests. then, it isimportant for a teacher to spell ow. as clearly as possible,precisely what his objectives are. and to Wild hisquestions around these, in the appropriate proportions.Tests which develop without such planning oftendegenerate into factual quizzes of the low-level, isolated.easily testable fragments of the course.

Guidelines ,fin (Voting Reliability

I. Is the test long enough?'2. Are the questions clear?3. Are the time limits realistic?4. Are the questions olappropu late difficult)?5. Is the marking effective?6. Are the instructions elem.?7. Has the choice of questions been kepi to a minimum?

Guidelines Inr Chilling Validity

Arc the questions relevant. important?2. Have all topics been assessed in appropriate

proportions?3. Have all skills been assessed in appropriate

proportions?4. Are there clues to the right answers?5. Is the typing and presentation adequate?6. Have all students had an adequate opportunity to learn

the material tested?7. Is security adequate to as aid cheating?

3. Planning the testif a test is not well halanced it will not he valid. Therefore,to ensure proper balance, it is a good idea to draw up itplan or blueprint. List the main topics to he covered on

one axis anti the major skills to he developed on the other.Thos, a blueprint for a unit on Mathematics might look

like this:Skills

Tories Mental) Campliratitel 'split:Wan Tidal

Sels 3 _ 5 IIFractions t .5 5 'J

Xiensurtment 2 3 5 10neettnnts 1 4 3 HO

Soothtic. 3 5 5 13_ -____.23 50'Fatal tot i5--- --------------------.-----------

The teacher who prepared this plan has clearly decidedthat the most important objectives in his course are thoseconcerned with applying the skills learnt in new situations,rather than memory work or rooting: computational skills.Therefore, 2 of the 50 questions are devoted toaPPlieation. Likewise, statistics is given more weight thanthe other topics, although all receive sonic weight.

Planning ofthis kind should he undertaken in everysubject and the weights given should reflect the amount ofemphasis given to the topics and skills during the leachingof the unit (*course. For example, in Social Studies, thetopics to he weighted might he:

Location, Climate, Discovery. Early Settlement.Industry, Transport, Culture, The skills assessed might heRecall. Comprehension. Application, and Evaluation.

An English test might have as its main topics: WrittenLanguage. Oral Language, Grammar, Fiction, Poetry. Theskills to he tested might he Knowledge. Comprehension,Application. Synthesis (production of original work), andEvaluation.

4. Test questions

Question writing is an art t hat depends on clearunderstanding of the stibject and of the pupils heingassessed, us well as a grasp of the general principles ofitem writing. It helps, also, to have plenty of time. someimagination, access to other people's questions as models,and an opportunity to have your questions edited bycolleagues.

Several kinds of questions can he used, and none is idealfor all circumstances. Written test questions can he simplydivided into two types:al Objective Questiems: These have right or wrong

24

answers and markers should agree on which are right andwhich are wrong.(b) Suk lectire Questions: Vssaytype tests in which thepupils must respond to open-ended questions bycomposing their own answers. There are varying degreesof completeness and correctness.

°MECUM MST QURS1 MINS

to) Multiple-choiceThese consist of a nem, stating the question, and 4 or 5possible options to choose from.

e.g. (i) What is the area au rectangle which is 5 cm longand 3 em wide?

A. li sq cm.*B. 15 sq cm.C. 16 sq cm.U. 30 sq. cm.

Note that the "distractors- (A, C and D) should heplausible answers for the pupils who might be unsure.

(ii) If bread is placed in a refrigerator, it will notbecome mouldy so quickly, because:*A. cooling slows down the growth of fungi

B. darkness retards the growth of mouldC. cooling prevents the bread from drying outD. mould requires both heat and light forgrowth

This question requires the pupil to apply his knowledge ofthe relationship bet wren temperature and the growth ofmoulds,

(19 Matching Questions:These consist of two columns of items selected so thatpupils can match the words or symbols in one column withthe appropriate word or phrase in the other. Matchingquestions are useful for testing homogeneous sets of factse.g.. matching hooks with their authors, chemicals withtheir formulas, words with the pans of speech theyrepresent. etc.

Country Capital City

1. Fiji ( )2. Tonga ( )3. Western Samoa ( )4. Cook Islands( )

25

A. Rarotonga!B. SuvaC. VilaD. Nuku'alofaE. HoniaraF. Apia

Note that both lists should he liottontntIS, and oneshould be /anger than the other. The main fault inpreparing these questions is that each list often containsterms which are heterogeneous. They should all heauthors, or cities, or chemicals, or ports of speech. etc. Ifcities are mixed with minerals, and people. andorganizations, they provide obvious clues to help theuninformed pupil.

(Cl True-hike Questions:These consist of a single statement which the pupils are tomark true or Jidse, or right or wrong. They are usefulquestions for a quick quiz, but guessing can he a seriousproblem with this type of question. This can he reducedsomewhat by asking pupils to correct the false statements.e.g.,

king 111: writeTut F the correct

answer

i it .1'Y/ 0144 is 4 'Car 1:no The volume 01a Mass

of gas tends toincrease as itstemperature increases I' or 1:

iiiii Fiji's chief viportis wpm "for 1:

kb COMA* full, QUOSHIMS:These consist of a question or sentence containing a blank,for which the pupils must supply the appropriate word.symbol or phrase. These questions arc actually"semi-objective", because there is often more than oneacceptable answer.

e.g.. (i) What is the name (tithe instrument used tomeasure temperature?

tin The device used to tell whether an electric chargeis positive or negative is:

liii) What is Z5% of 44?

Whirl' Type olObjective Test Question Should You Use?

There is PIO OW hO.S1 type of item. All are appropriate atone lime or another, but multiple-choice questions aremore widely used than others in standardized tests andimportant examinations. The following advantages areoften claimed kw multiple - choice questions.

(i) They are more objective and reliable than essaytests or completion questions.(ii) They make possible the testing of a larger sampleof the pupils' knowledge and ability in a shoe timethan does the essay test.(iii) They enable the teacher to measure processskills as well as recall of simple knowledge. Bycontrast, true-false, matching and completionquestions are largely restricted to simple recall.(iv) They are easy to mark in large numbers.(v) They make it impossible for a pupil to gain a highscore by guessing.(vi) Common weaknesses in pupil knowledge andskills can he readily diagnosed by the teacher.(vii) The questions themselves can be readilyevaluated and improved by means of item analysis.

On the other hand, multiple-choice questions do havethese disadvantages:

(i) They cannot measure pupils' creative skills, orability to organize material in a coherent manner.This is particularly important in language, literature,and other expressive subjects.(ii) They take much time and skill to con4ruct.Poorly prepared questions may produce more invalidresults than completion or essay questions.

5. Suggestions for preparing test questions

1. Essay Questions:

(i) Specify clearly what is to be ineludedin the answer.Compare: (Poor) : Write an essay on the French

Revolution.(Better): In not more than 500 words,

(a) Outline the main causes of theFrench Revolution.and(h) Explain why reform could not beobtained without violence.

4. 26

(ii) Use several short questions rather than one long one.(iii) Avoid optional questions where possible, as theymake marking more difficult.(iv) Before the test. prepare a model answer, outlining themain criteria and weights to be attached to each.(v) Mark one question for all pupils before beginning thenext.(vi) Mark without knowing the pupils' names. wherepossible.(vii) Obtain independent assessments. wherever you can.The average of two markers is more reliable than theresults from one.

2. Objective Questions:

(A) GENERAL(i) Keep your questions brief, simple in expression, and

free from complex verbal instructions, doublenegatives. etc.

(ii) Test only the important facts and skills. Avoid trivialquestions. "catch" questions. and irrelevant material.

(8) MULTIP LE-CHOICE QUESTIONS(i) The problem should be clearly stated in the stem of

the question.e.g.. (poor)BatsA. drive off harmful birdsB. are enemies of man

'C. eat insectsD. eat ratsThe pupils must read all options before theyunderstand what the problem is.

(ii) Use only plausible distractorse.g.. (poor)The Prime minister of Fiji isA. Mr MuldoonB. Mr Fraser

*C. Sir Kamisese MaraD. The Shah of IranMany pupils could guess the answer with verylimited knowledge. Names of other prominentFijians would provide better distractors.

(iii) Ensure that there is only one correct answere.g.. (poor)The population of Hamilton isA. less than 50 000

27

B. between 50 000 and 70 000C. over 70000D. over 80 000Both C and D are correct.

(iv) Avoid the stereotyped tanguage of textbooks in thecorrect answere.g.. (poor)The Renaissance in Europe was characterised byA. a decline in tradeB. many religious wars

*C. an unusual efflorescence of creative talentD. the loss of colonies

(v) Beware of grammatical clues and verbal associationse.g.. (poor)The French scientist who discovered the basis forpasteurising milk was

'A. Louis PasteurB. Isaac NewtonC. Francis BaconD. Alexander Graham BellThere are two clues to the right answer here. Thequestion should be rephrased.

(vi) Make the correct option the same length as thedistractorse.g.. (poor)Sweets are not recommended for eating betweenmeals as theyA. cause diabetesB. supply excess energyC. stimulate the bile

'D. dull the appetite for foods rich in other necessaryelements

D sounds right because it is longer. and so makes fora fuller statement.

(C) COMPLETION QUESTIONS(i) Use a single blank in each question

e.g.. (poor)The "_ of _" was written byA pupil may know the facts required, but beconfused by the question.

(ii) Place the blanks near the end of the sentencee.g.. (poor)

is the name usually given to thebreakdown of the soil by various processes.

e.g.. (better)The breakdown of the soil by various processes isusually called

(iii) Make all blanks the same lengthe.g., (poor)Villa is the capital city of theA pupil who was not sure whether to chooseSolomon Islands or New Hebrides would have anobvious clue here.

(iv) Make sure that there are a finite number of acceptableanswers

e.g., (poor)Columbus discovered America ine.g.. (better)In which year did Columbus discover America?

6. ConclusionsMuch more could be said about writing soundquestions. However, a careful reading of theprinciples outlined above and some meticulousediting by your colleagues should make for betterreliability and validity than that in a test which grows'Topsy-like' without planning and forethought.

Teachers who wish to improve their assessmentskills further can learn much from studying examplesof well prepared examinations and standardisedtests, and from analysing the results of their owntests, using item analysis. This and other topics canbe followed up in such books as:

Ebel. Robert L. Essen:1'14 of Educational Meamirenteni. NewJersey. Prentice-Hall. 1977 tthe test developer's bible).

Peddie. Bill, and Graham White Testing in Practice. Auckland,Heinemann Education. 1978. t short. pithy. a practicing teacher'sguide).

Queensland Department of Education. School Assent:wit:Prix. edures Titles: t An Introduction. 2 The Multiple Choiceitem. 3 Assessment in English,4 Moderation Within Schools. 5Assessment in English. 6 Assessment in Foreign Languages. 7Planning a Su mmative Assessment Programme. 1971-5.Available from ACER.

izard. J.F. Construction and Analysis ofCleisamont TestsMelbourne. ACER, 1977.

28r

29

Item 12[SetOIMrelt

$0101111~IV 100004

number one 1980

Whatever Happened toProgrammedLearning?By Graham Wagner, NZCER

In the late 1950s programmed teaming was set torevolutionise teaching. B.F. Skinner started themovement in 1954 with a provocative paper entitled TheScience of teaming and the Art of Teaching'.

We are on the threshold of an exciting andrevolutionary period, in which the scientific study ofman will be put to work in man's best interests.Education must play its part. It must accept the factthat a sweeping revision of educational practices ispossible and inevitable.

The scientific approach Skinner had used in his ownexperimental work with animals could, he believed, beapplied to human teaming by the use of teachingmachines:

If the teacher is to take advantage of recent advancesin the study of teaming. she must have the help ofmechanical devices.

Six years later Stolu row wrote:

The teaching machine seems to be a majorbreakthrough in education comparable to the book,radio and television.

30

Almost immediately. however, there was a change ofemphasis away from machines (hardware) to theprograms written for machines (software). Theprograms were soon (for convenience) presented intextbook form and the machines became less dominant.This was a move in the right direction but even so somereputable scholars still seemed to be mesmerised by therevolutionary possibilities of Programmed Learning andas late as 1962 we were still reading such statements as:

Programmed Instruction is essentially a revolutionarydevice in that it has the potential to free man frombondage.

It was easy to see the truth of the arguments Skinner putforward in his attack on the shortcomings in theeducation system of the USA: other Western countrieslike Australia and New Zealand were also vulnerable,The education systems were static, tradition bound,content dominated, placing the teacher and thecurriculum first and the learner a very poor second. Ingeneral, the teaching methods revolved around theteacher's knowledge of the syllabus and previous exampapers. The most common teaching aids were thetextbook and a chalk board, and the most recentinnovation, in New Zealand at least. was to change theboards from black to green. The syllabuses gave littlehelp to teachers: they provided lists of what was to betaught, but little, if any, help with how one idea shouldfollow on another. Further, in the examinationdominated climate, any reinforcement (such asinformation given by the teacher to the pupils aboutwhether they were getting their work right or wrong)usually came too late. too infrequently, and toounsystematically.

The teacher was in a poor position to remedy any ofthese shortcomings Skinner suggested without someform othelp. However, the new technology available tothe teacher, he believed, made it possible to researchthe complexities of human behaviour in a controlledlaboratory-like' environment with the help of machines:

Now, the human organism is, if anything, moresensitive to precise contingencies than the organismswe have studied. We have every reason to expect.therefore, that the most effective control of humanlearning wilt require instrumental aid.

Skinner saw teaching machines as a part of animprovement to teaching techniques. In this respect hewas saying nothing really new. A machine had beenused in history as far back as the middle ages for thetraining and amusement of knights. Called a 'quintaire itwas essentially a figure of a swordsman mounted on apivot. In one hand the figure had a sword and in theother a shield which served as a target. If the horsemanwith his lance did not strike the the shield squarely in thecentre the quintain would pivot quickly and slap theknight with his sword as he sped by. This form of'negative reinforcement' was a strong inducement toimprove the knight's performance.

Sidney Pressey is regarded as the inventor of the firstmodern teaching machine although Thornlike iscredited with having thought out a way to programtextbooks using a machine-like system in 1912.

Researchers investigating the effectiveness ofmachines and programs have turned away from talk ofprogrammed learning, which directs attention at whatlearners do with their machine or programmed book,and are now looking more closely at what programwriters ask learners to do. This emphasis leads Them toprefer the term programmed instruction. As Hartley. aprominent writer on the movement points out, they areboth sides of the same coin.

The Techniques

In theearty days of programmed learning the list oftechniques and principles advocated was somethinglike this:

(a) The learner must work on his own, with a machine orprogrammed text.

(b) He must work at his own pace.(c) He must work through small steps.(d) These steps must be carefully sequenced.(e) The learner must respond in some way to each step

usually by writing his answer down, or by pressingan appropriate button.

(f) The learner must receive immediate knowledge ofresults concerning the correctness orappropriateness of his answer,

An additional principle was also borne in mind, but thisdid not affect the learner's way of working. This was:

(g) The program must be tried out and revisedaccording to the results of the trials so that the finalversion of the program teaches specified objectivesto a specified group of learners.

These techniques were simply and practically translatedinto the following programming steps which writersemployed when constructing a program for either abook or a machine:

(1) Performance objectives were clearly defined.(2) A task (or topic) analysis was carried out.(3) A teaching sequence was selected.(4) The program was written (in, for example, linear or

branching format).(5) Then the program was evaluated and revised (if

necessary), before it was used in a classroom.

Theoretical MattersThere is now enough evidence to support the use ofprogrammed instruction in selected areas of the schoolcurriculum without spending a lot of time justifying itsusefulness. But 'mechanised learning' is a sort of bogyin our educational folklore and it will be as well to lookat where it comes from and consider it in the light ofwhat we now know.

Skinner backed up his description of how learningtakes place in laboratory rats and pigeons with a thongthat of reinforcement. A reward is a positivereinforcement of the action that brought the reward; theaction (or response) being what the experimenter, orteacher. is trying to get learnt. A punishment is anegative reinforcement for some response which the'teacher is trying to avoid. Liberal minded people werepleased to learn that rewards for the 'right' actions get'better' results than punishments for the 'wrong'actions. (A little awkward was the discovery that ahungry rat finds food quicker if you give him a littleshock 1punishment?l as well as the food.)

Rewards can be, for humans, just the pleasure ofgetting the right answer. So, the argument goes, give allLearners the right answer right away, everytime, and

32

learning will proceed apace. Keep the steps small andthe positive reinforcement will come more often andlearning will be even faster. This sort of theory. whichmakes no mention of what might be going on in thelearner's mind, is often called the 'behaviouristic theoryof learning.'

Skinner's theory came under close scrutiny. Onecriticism centred around whether studies of animallearning tell us anything about how humans learn.Ausubel in 1963 said:

The methods of learning employed by animals inmazes do not necessarily correspond to methods oflearning children use in grappling with verbal mattersin classrooms.

Another criticism was that 'laboratory results had littlerelevance to 'natural' learning situations since thesimplification of behaviour necessary for laboratorystudy hides the complexity of human learning to theextent that the results become misleading. Today.studies of human learning favour realism overexperimental control in research methodology.

Social and intellectual factors in learning areimportant. Critics have attacked programmed learningbecause it is a method that isolates the learner anddeprives him of the chance of teaming from others. Thisis a fact and must be taken into consideration when ateacher introduces programmed learning. Isolation canbe detrimental to some learners if not used inconjunction with other teaching approaches. Pastexperience. and the effect it has upon what is to belearned, must also be considered. Similarly the need forcreativity to be encouraged and acknowledged.

Although Skinner's simplistic description of howlearning takes place has been largely discredited,programmed learning survived as one useful techniqueof teaching. Several myths about it seem to havesurvived also. One of them is that programmedinstruction is incompatible with humanistic approachesto teaching. Many readers may be surprised to hear thathumanist theoretician Carl Rogers in his book Freedomto Learn writes that there is a clear place forprogrammed instruction in those instances in which thestudent encounters 'gaps in his knowledge. toots which

33

he lacks. information which he needs to meet theproblem he is confronting. Here the flexibility ofprogrammed instruction is invaluable'. Whileadvocating caution against using programmedinstruction as a substitute for thinking he believes 'that itis one of the most powerful tools which psychology hasas yet contributed to the field' (p.141), that is. when it isused flexibly.

Impact on Teaching

Successful though programmed learning has been, thedebate surrounding the laws of learning and theimpersonal machines made school teachers wary of thetechnique. Twelve years after it all began there was littleevidence that programmed materials were being usedextensively in American schools. However, theoristswere still experimenting with different methods ofsequencing materials.

In New Zealand programmed teaming made someheadway in industry. a few government departmentsand the armed services. There were attempts. largelyunsuccessful. to introduce the techniques into schoolsthrough the universities and teachers' colleges. TheDepartment of Education was not interested. There wasLoud debate over the impersonal nature of machines, theboring sameness of programs. the inroads of abehaviouristic philosophy which made no allowance forcreativity or discovery teaming, and these factors helpedto turn teachers off the innovations. Although not theonly reason for teachers' resistance. the bogy ofmachines replacing teachers in classrooms was thoughtof as a very real threat. especially at a time of growingunemployment. In New Zealand it seemed prudent for asmall country with limited resources to put its moneyinto 'less controversial' resources such as New Maths,PSSC Physics. Open-plan classrooms. and the like.Teachers were too busy to spend time grappling with anew method of teaching about which there were gravedoubts.

Even if there had been no controversy. the timerequired for training teachers in the new techniques.and the resources required to implement such

programmes were in most cases insurmountablehurdles. For the hardy few who managed it, there wasalways a suspicious inspectorate not kindly disposedtowards innovations that handed over the learning tothe student and left the teacher doing something otherthan teaching. To be a teacher you had to be seenteaching.

Today

The concept of what programmed instruction is haswidened over the years: what started out in 1954 as arelatively simple technique of step by stepselfinstruction has become a complex systematicapproach to teaching in general. The four inter-relatedsteps (1. stating objectives, 2. analysing tasks. 3.selecting a teaching method. 4. evaluating theeffectiveness of the teaching and revising to improve)became a flexible curriculum plan enabling teachers touse large scale teaching methods in a way that is nowcalled a systems approach to teaching. Thoseacquainted with computer jargon will recognise'system' as the word used to describe the way computerexperts go about their planning. Indeed the excitingdevelopments in Computer Assisted Learning are anoutcome of the programmed learning movement.Examples of the large scale learning packages whichcan be designed with the help of a 'systems approach'are various: they would include resource andcurriculum packages designed to last a term, or even ayear. and containing a variety of audio-visual aids andmaterials. all to be used with a variety of teachingmethods. Some 'systems approaches' resulted in largescale curriculum plans. for example. the Keller Plan. theDalton Plan. performance contracting. the integratedday. *Sesame Street'. 'Man: A Course of Study'. to namejust a few.

Table 1. on the back page. is a good summary of thechanging nature of programmed learning. Researchersstarted off by looking at the techniques of programmedinstruction but now concern themselves with the wholefield of educational technology. course planning andcurriculum development.

34

O

cr.

r

Table 1 The changing nature of programmed instruction(each cell indicates areas of research)

1954 1961

behavioural objectivestask analysis

1968task analysisbehavioural objectivessubject analysis

1974systematic analyses and constraintslevels of objectivestechniques of structuring material

small steps

overt responses

reinforcement

self-pacing

small steps

logical sequencing

active responding

immediate feedback

self-pacing

task structures andrelated leachingstrategies

presentation as acommunication problem

task structures and related teachingmethods. strategies and tactics

presentation as a communicationproblem; problems of learning fromtext, diagrams. etc.: appropriateinstructional decisions

individual differences and methodsof learning

revision evaluation(i) revision(ii) studies

evaluation : revision evaluation of(i) objectives and methods(ii) techniques: retrospective and

concurrent decision makingevaluation as illumination

installation and implementationinservice teacher training

The Contribution of Programmed Learningto Education

With the blurring of the boundaries between media, itbecame necessary to look beyond the teachingprogramme and think instead of programmed teaching.This is where 'systems' technology took over from the

35

machines and this change of emphasis may prove tohave been programmed learning's greatest gift toeducation.

However, we got more than a re-emphasis onobjectives and evaluation from the experimental work ofprogrammed learning enthusiasts we got anintroductory understanding of educational technologywhich will help us cope with further technological

advances and the complexity of learning; mostcomputer assisted instructional (CAI) programs are.after all, just very sophisticated programmed learningsequences. We are much wiser now about thelimitations of programmed instruction. We are wiser tooabout theories of learning. We accept that programmedlearning is only one teaching method out of many andthat it works best when the teacher is part of theprogramme. But we also know that the skills we gainedin producing programs are transferable to the wholefield of education under the heading of a systemsapproach to education.

The education system of the future will look verydifferent from the present one: it will be different in largepart because of Skinner and the programmed learningmovement.

Notes

Skinner's provocative paper, 'The Science of Learning and theArt of Teaching' is found in:

Lumsdaine. A.A. and Glaser. R. (Eds.) Teaching Machinesand Programmed Learning, Washington, N.E.A.. 1960.

The quotation by Stolurow is found in the same book.

The quotation 'Programmed Instruction is essentially arevolutionary device..,' is from:

Schramm. Wilkin. (Ed.) Four Case Studies of ProgrammedInstruction. New York, Fund for the Advancement ofEducation. 1964.

The second quotation from Skinner is from the same paper asis mentioned above.

For criticisms of Skinners ideas see:

Armed. J. 'Psychological Bases of Educational Technology'in Budgett. R. and Leedham. J. (Eds.) Aspects ofEducational Technology VIII, London, Pitman. 1973.

Hartley, who is quoted for his good summaries of techniquesand the changing nature 01 programmed instruction see thesection Techniques. and Table 1 can be read in:

Hartley. J. 'Programmed Instruction 1954-1974: A Review'.in Programmed Learning and Educational Technology, Vol.Xt.. pp.278-91.

36

M.*

itpni 11 [setnunfiner 1 jPo

hl \t Lit,LI -I in tillH1.1 :1,11 (' \s.'kl II-IL:\ ,11e

L in ]0-1 L=.CF.1}.-)r,,L:t1,11111:1!

.11111 1:Th 1 111", INC e\p,_'I1f111'llt

Assessing InternalAssessment

By Ann Lamont and Bruce McBride.,[)C1 }(71-tHICHf of L Iiicctiott, (.21tCCH'=;IWRI

1(17? H1F. In-nil...1,P', 1:1 l)l1t'L'f-1.1.1:1:i `-,e[i:(..tIL)I1 ET-1H, ITI.M\

C%L"IT,11L1rl, L_]['[ f1111fICL; t'',' tho. rt.[11t. in Lin t_'',terrikll CN.11-11iI1,1tLt)11

1111, e\k1f11111.11:L111, 1-1.11L11 1_1:11,,'Cr-,J (1t

_11[11Lrl.lit'Ll '-eCLIF,d.111 tAllIC411'4)11 :11 C ,I)L1t.:('Il'.1.111tii

tl'C (LIF:CH)r1 Ht ft'L-01-11r1leilL1.1t1011', in

IC)(-1", L",1:11[1',WCL' 1L1 \.1111111.11011,,

H-LI, 1 C11(_ir.,1r1 1.11H 1., L'J. Pkii.!ILIT1ILTIt" er-Fi,-)l

,Lir-1114; 1'171 ant9 rt'LILHIng tnncx l 11-,71,

In 1°- ; 11'4111,11 L111-1h,i1 11Lir,'T n1,it1c

n hr75 '11( N1 \1'H'

This paper describes the major aims andfindings of the MAPS Study conducted by theauthors for the Department of Education,Queensland.

Aims

The major aims of the study were:

to map the decisions that have been made aboutassessment procedures;

to find out how satisfied those who actually do theassessing are

to compare the assessment situation in 1978 with thesituation as it was in 1974.

Semester 4 at the Year 12 level was chosen and threesubject areas (English, Mathematics and ForeignLanguages) in a representative sample of 12 Statesecondary schools were investigated.

Data were collected by the researchers duringinterviews with school inspectors, principals, subjectarea co-ordinators and Year 12 teachers. Also, in eachschool, group discussions between the researchers anda class of Year 12 students provided data aboutstudents.

Background

The third aim of the study was to compare thesituations in schools in 1974 and 1978. Data had beencollected in the 'Schools Under Radford' (SUR) studyconducted jointly by the Department of Education andthe Board of Secondary School Studies. This study hadnot investigated assessment procedures in detail butwas concerned with a global evaluation of thesecondary school system which operated after theimplementation of internal assessment.

The subjects investigated were chosen to represent awide range of characteristics. English and Maths I weretaken by large numbers of students. The skillsdeveloped in these subjects differ from each other antiare used by students in a wide range of other subjects.Foreign Languages were chosen because each foreignlanguage is studied by a small group which, it ispurported, is drawn 1...m a relatively elite group ofhigh achievers.

The final year of secondary schooling, Year 12, waschosen because:

the results of three different measures ofachievement are given to students in Year 12(moderate school assessments, the TertiaryEntrance score and school reports); and

achievement measures provided to students in Year12, whether internal or external. are likely to havea more significant effect on students' future careerchoices than measures provided in earlier years.

Moderated School Assessment

For each of the four semesters in years 11 and 12, thelast two years of secondary school, students are given arating on a seven point scale for each subject studied.An elaborate system of moderation has been set up bythe Board of Secondary School Studies to ensure thatratings given by different schools are of equal worth.

The moderation system involves meetings of theteachers involved in each subject. This system isdescribed in detail in information kits distributed bythe Board.

The Tertiary Entrance Score (Ti; Score,

The TE Score is based on the Special SubjectAssessment (SSA) and the Australian ScholasticAptitude Test (ASAT). An SSA is an achievementmeasure given by the school to each student in Year 12for each subject studied in eitherYear II or Year 12.The calculation of then Score is described in detail byMcCaw, Wary and Marycle (1976). The main stepsare:

1. The SSAs for each subject are scaled to the mean forthe students taking that subject in their school. Thiscompensates for the differences in ability levels ofstudents taking different subjects.

2, The aggregate of the scaled scores for the highest 20semester units is calculated,

3. The aggregate scores for students in each school areresealed to the ASAT mean for the school. Thisadjusts for different ability levels between schools.

4. The resealed aggregate scores of all schools are usedto rank all students in Queensland in orderof merit.TE Scores are then assigned to students inaccordance with their position in the order of meritlist.

Seltoot Reports

The type of information appearing on schtx)I reportsand the frequency with which they are issued isdetermined by each individual school,

ResultsStimulative Assessment

(Assessments made during or at the end of a course ofstudy which contribute to the final grade or rating.)

Information was obtained about the structure of thesummative assessment programme. Changes desiredwere noted. The decision - making; processes were alsoinvestigated. The major topics studied were:

weighting on various assessment elements;weighting on modes of assessment used; weightingon types of item used; use of common test, (withinand between schools).

Weighting

The weighting given to various lo,..ess men t elements,modes of assessment and types of item variedconsiderably between subjects and between schools.1 lowever, teachers were largely responsible for thechoice of what was used, and consequently weregenerally satisfied.

38 '

In Maths I, ratings were based almost entirely onformal tests underexamination conditions, whereas inEnglish and foreign languages the ratings were basedon Marge number of assessments Ufa widely varyingnature.

t fse of Consmon Tests

Within all schools. common Maths 1 tests were usedwhere two or more classes were studying the sameunit. Many mathematics teachers indicated thatcommon tests between schools would he desirable sothat schools' results could be compared.

Common assessment for all Year 12 English classeswithin a school was not universal and the amount ofcommon testing varied markedly between schools.Must schools, however, used more than 70% commonassessment between classes. Almost all Englishteachers indicated that they did not want commontesting between schools.

No school in the sample had more than one classstudying the same foreign language at Year 12 level.Some foreign language teachers suggested commontests between schools fur parts of the course. such asgrammar and comprehension.

There was a sharp contrast between the Maths 1teachers, many of whom wanted a return to an externalsenior examination and the English and foreignlanguage teachers, none 0( whom wanted a return tofull external examinations.

Overview of Sumomtive Assessment

In general. teachers were satisfied with the presentend -of- the - course (summative} assessment pmgratn fortheir Year 12 classes. Those who did desire changeindicated that they wanted a small change in emphasisrather than a change in its structure. In English therewas some support for an increased emphasis on oralwork and other communication skills with a decreasedemphasis on literature.

Most decisions concerning the structure of thesummative assessment program were made by theteacher in charge of the subject. usually in cow ultat ionwith the other teachers. In all cases, the principal couldover-rule these decisions. but it appeared that thispower was seldom exercised.

Formative AsSeSSttieta

(Assessments made while a course is in progress toshow gaps in knowledge.)

Two of the aspects of formative assessment probed inthe study were:

the use of summative assessment for formativepurposes. e.g., middle of term exams used todiagnose what a pupil needs to learn;

the balance between summative and formativeassessment.

39

'l'Iu Me of Sourialive Assessoseet ft), Purposes

The major purpose of the summative assessmentprogram is to provide measures achievement at theend of the pupil's school life, but it can also providesome feedback on student performance. Three aspectswere investigated; delays, students' attitudes hi marks,and the making of corrections on marked scripts.

Teachers and students generally agreed that the delaybefore getting Maths 1 and foreign language papersback was satisfactory. Delays in English were longerand considered unsatisfactory by students in about halfthe schools. In contrast. almost all English teachersthought the delay was not too long.

Most students indicated that they were moreinterested in where and why they lost marks than intheir overall scores, However, teachers felt thaistudents were more interested in the scores!

In some cases, students were allowed to makecorrections on their papers while in other cases thepapers were kept 'clean' for moderation meetings.Neither teachers nor students appeared to beconcerned with this issue.

The thrlimee Between Soon:Wive emit Forosolive Msestrient

The majority of inspectors, principals and teachersindicated that they were satisfied with the presentfrequency of summative assessments. While studentswere reasonably satisfied with the summativeassessment load in these three subjects. some informalcomments by students indicated that those studentsdoing mainly humanities subjects had a heavy workload. (Data relating to these subjects Was not sought inthis study.)

Sperial School Assessments (SSAF.,

It appears that very few teachers have much knowledgeof. or interest in. the procedures used for thesecalculations in their school. Usually majorresponsibility for the calculation is given to one or twopersons (generally, the principal, deputy principals.head of the mathematics department or a mathematicsteacher) with a small degree of participation required(rum the other subject heads. While principals reportedthat input from 'other teachers' occurred frequently.these teachers felt that an input from them was seldomsought.

..3chool Reports

The contents of the end of semester report variedmarkedly among schools. All schools provided at leastone achievement measure. and all except one schoolpr wided at least one affective measure.

Comparisons between 1974 and 1978tkes of formative k;sssoseiriData collected in 1974 in the 'Schools Under Radford'study indicated that most teachers held very strong

views about the emphasis on summative assessment.They felt it prevented them from devoting any time toformative assessment. This contrasts sharply with thepresent study which indicates that, in 1978, formativeassessment was frequently used, and that over-testingis no longer a serious problem in the subjectsinvestigated.

Flexibility in Assessment Programs

The results of the 'Schools Under Radford' studysuggested that teachers in 1974 were not takingadvantage of the flexibility in assessment proceduresafforded by the implementation of the Radford Reportin 1971. By contrast, in 1978 some teachers particularlythose of English and foreign languages appear to havecapitalised on the flexibility available through internalassessment for modes of assessment and types of item.However, moderation meetings still base theirjudgements mainly on written work. This emphasis onwritten work still appears to be constraining teachers tolimit their use of oral assessment techniques.

Mechanics of the Moderation System

in 1978, respondents were generally satisfied with theprocedural aspects of the moderation system whereasthis was not generally true in 1974.

Principals interviewed appear accutely aware of theproblems of over-testing and many are taking steps toensure that over-testing does not occur.

When asked if they would like to change theemphasis on summative and formative assessment,many teachers indicated a preference for a small shiftfrom summative to formative, but generally they didnot see the summative assessment program asdominating their teaching. Most teachers in thesubjects studied indicated that they were able tointegrate formative assessment into their teachingprograms. Half of the principals and most of theinspectors wanted to increase the amount of formativeassessment.

Validation ProceduresMost principals and teachers indicated that they weresatisfied with the quality of the assessment instrumentsin their school or subject. However, on beingquestioned further, most of these indicated that theirview was based purely on their own perceptions ratherthan on any specific validation procedure. In contrast tothe satisfaction in schools, only one of the sixinspectors interviewed was satisfied with the quality oftests used in schools.

Achievement and other Exit MeasuresData was collected on the following aspects ofmeasuring and reporting student achievement:

moderated school assessments; special schoolassessments; and school reports.

Moderated School Assessments

About two-thirds of the respondents were dissatisfiedwith the moderation system. Many English teacherswanted an alternative which decreased the emphasisplaced on moderation, giving schools moreresponsibility for maintaining standards. However,many mathematics and foreign language teachersdesired some form of external assessment.

A considerable number of respondents have doubtsabout whether schools can be compared accurately andmarks adjusted accurately. Many teachers felt that onceratings had been given in Semester 1, it was verydifficult to adjust the distribution upward insubsequent semesters if some students showed amarked improvement in their achievement.

ConclusionsThe major findings are:

Teachers are generally satisfied that mechanical aspectsof assessment and moderation are functioning fairlysmoothly.

There is substantial concern that comparabilitybetween schools is not being achieved that thesystem of moderation is not ensuring that results arefairly scaled to reflect the standards in different schools.

Schools are utilising the flexibility available to them intheir choice of modes of assessment.

Respondents generally felt that summative assessmentdid not dominate their assessment program and thatthey were generally able to integrate formativeassessment into their teaching program.

Only half of the schools used standardisationeffectively when combining semester scores during theassigning of Special School Assessments which are sentto the Board for input into the Tertiary Entrance (TE1Score calculations.

Between 1974 and 1978 there have been considerableimprovements in: the use of formative assessment; theuse of appropriate modes of assessment; and themechanics of the moderation system.

Overall, there is greater satisfaction with the system atpresent than there was in the years immediatelyfollowing the implementation of internal assessment.

Further information concerning the studies may beobtained from the authors d- Research Branch,Department of Education, Queensland, P.O. Box 33,North Quay, 4001.

.... I

Standard 3, Newtown, 1935

After Play: 11.05 Everyone doing the 7 times table1

i 11.10 Everyone doing the 8 times table

11.15 Everyone doing the same 20 mental

11.20 Everyone doing the sums on page 45,

..........._..... _......_ ____

only one or two of them fitted.It was the class trap. We put the children in classes: but

It was just right for nobody, well, maybe, for one or two.

you have finished.

It was too slow for half the class.

It was too fast for half the class.

pens down and fold your arms when

Mb Ai Ai

. .a Mb

' .AO -"I

- - s*

I'.

- II'.' .

. . -

ft .

.1

- . .

-

. _

Student Team Learning

By Paul Power, Melbourne State College

1 can't go on vacation now. My team needs me.' That'swhat 9-year-old Lisa at Old Forge School, Maryland, saidwhen her parents suggested a trip to Mexico. And theteam she was referring to was her learning team of fiveclassmates in her Year 4 (Std 3) class.

Lisa's class is trying out Student Team Learning inthree variations, each creating the same kind of peersupport, excitement, effort, and camaraderie normallyassociated with team games. The variations are calledTGT, STAD, and JIGSAW. The experiment is under theCentre for Social Organisation of Scheele. JohnHopkins University, Baltimore, They have spent eightyears developing their own techniques and adapting theJigsaw technique of Elliot Aronson of the University ofCalifornia. Santa Cruz.

Lisa's class is one of many in the evaluation, whichhas included thousands of students. Student TeamLearning techniques have equalled and in many casessurpassed traditionally structured classes In promotinglearning, particularly in basic skills. In addition they havebeen shown to have positive effects on race-relations Inschools that wasa major reason for the developmentof Jigsaw and on students' concern for each other,and on self-esteem.

Further, the techniques are easy and inexpensive.

TGT (Teams-Games-Tournaments)

Like the other two techniques, TGT is designed aroundthe learning team which is composed 014 or 5 boys andgirls of different levels of ability and social backgrounds.The teacher assigns the children to teams, each teamchoosing a name, Teams stay together for al least 6weeksto build cohesiveness and learn spirit,

After the teacher has presented a lesson, worksheetsare handed to each team. The team members then studytogether trying to make sure that each member knowsthe material. The teacher is free to co-ordinate, assistindividual pupils who are having difficulties, supervisethe teams and so on. She acts like a head-coach but allthe players help their teammates polish their 'game'.

At the end of the week, or the end of a learning unit,the teams clash in a tournament covering the work theyhave been boning up all week. The games use simpleobjective-answer questions taken usually from readingand language programmes. 1. Grass goes with groundasbark goes with (a) dog. (b) tree, (c) wolf; 2. Sleepygoes with tired as confused goes with (a) mad, (b) afraid,(c) bewildered; 3. Closed tightly = (a) clipped, (b)clenched, (c) cleaned; 4, The farm belongs tofamily, (a) their, (b) there, (c) they're. The tournaments

42

are well organised with each child competing againstothers of similar ability from the other teams winningis equally valuable so low ability students have just asmuch chance of contributing to their team's success.There are no individual losers since only the learnscores count. The winning team is announced in theweekly class newsletter. Questions Come in randomorder, challenging is allowed with penalties if thechallenger is wrong.

Consistently high scorers in the tournament are'bumped' to meet opponents of higher ability next week.Low scorers are moved to meet their peers. Aggressivechildren lake to the system better than others but thesystem encourages co-operation and interaction acrossability levels and across racial and sex lines.

STAD

Student Teams Achievement Divisions uses the sameteam structure as TGT but instead of the tournament thestudents take quizzes and short tests individually. If youscore as well as usual you get a maximum score for yourteam, and there are bonus points for those who dobetter than usual. STAD takes less time and is easier toorganise than TGT.

JIGSAW

Each team is given the same four or five pieces of workusually biographies, short stories, book chapters, and

so on and each team member becomes an 'expert' onone piece. Members of different teams who have thesame topic meet in 'counterpoint' groups, for aboutone-third of the time available, to work over the material.Then they return to their original groups and pass ontheir information to the other members. There is nosystem of inter-team competition for marks in JIGSAW.Groups should be heterogenous as mixed as possible

with best friends and worst enemies in separategroups.

There is mixing of abilities, sexes, and ethnic origins,and a great deal of practice in reading for meaning,verbalisation, listening, and explaining skills. Enjoymentof school, co-operation, self-esteem, concern for othersand the number of black-white friendships (compared tocontrol classes using ordinary teaching methods) grewsubstantially.

TGT and STAD are most appropriate for leachingbasic material: JIGSAW is more appropriate for subjectssuch as literature, history, social studies, and theconcept-learning aspects of science. Low-cost materialsare available which provide sufficient detail for teachersto use the techniques with their own curriculum material(US$3,00). Specific curriculum material has beenprepared for language arts, mathematics, reading,social studies, and nutrition for years 3 to 8. These tooare inexpensive (US$8.00 per set) but are geared toAmerican schools. Although not directly adaptable inAustralia and New Zealand they may provide usefulguides to teachers who want to develop their ownmaterial for use with the basic techniques, which dooffer much promise.

Letting students work on theirown: an experiment in 4th FormScience

By John Longbottom,Aranui HighSchool

In 19781was faced with a particularly 'deactivated' 5thForm science class. To catch even half the pupilsattention was difficult, to hold it was impossible. Severalpupils would do no work at all unless someone stooddirectly behind them. In a desperate attempt to gainmore time to circulate and prod along the morereluctant class members I transcribed five lessons,including instructions for experiments, explanationsand notes to be copied, onto worksheets. Then, makingmaximum use of the Hawthorne effect, I told the class itwas the subject of an important experiment andpresented the sheets to the pupils as a new way ofworking. The responses ranged from neutral through tovery encouraging, with one pair of pupils who had donevirtually no work all year suddenly seeking me out to askquestions on the text and experiments.

Later I had the opportunity to try out such anapproach under more controlled conditions, andattempted to isolate and measure the effect of allowingstudents to pace themselves through lessons.

The Programme

In a study, undertaken as Research Affiliate at theUniversity of Canterbury, I prepared worksheets for afive-week block of individualised work in science andtested them under controlted conditions. The resuttsindicate that pupils tend to work harder when left tothemselves and their achievement at least matches thatof pupils taught by a teacher.

About fifty objectives were developed for Section 4(Force and Mass) and Section 22 (Forces in Action) ofthe New Zealand Form 1 -V Science syllabus. Based onthese objectives, tests and worksheets were writtencovering a five-week block of work (20 x 1 hr periods).The worksheetswere not difficult to write as theycontained nothing different in content or approach tothat of a normal lesson.

Four middle-ability 4th Form classes worked throughthe first two-week block and then sat a post-test. Two ofthe classes (controls) then reverted to 'normal teaching'while two experimental classes continued with the samematerial in worksheet form. A second post-test was heldat the end of five weeks and the results of experimentaland control classes were compared (with the results ofthe first post-test being used to adjust for any initialdifference between the classes).

A questionnaire was used to gaugevarious aspects ofpupil attitude towards the method.

The Results

Pupils achieved equally well whether they worked underself-paced or teacher-paced conditions. Thequestionnaire clearly indicated that pupils put in morework under the self-paced conditions, even when theyhad a negative altitude towards this method of working,

Commentary

Using the worksheets under more normal conditionsafter the experiment was over, and where there hadbeen free interaction between pupils, teachers andworksheets, I have encountered a generally positiveattitude, with most pupils expressing the desire to usethe same method for at least a proportion of theirscience course. No one method will ever suit everyoneand some pupils will dislike working on their ownbecause they miss the security of a teacher telling themwhat to do or because they have difficulty gettingthemselves organised. However, it is hoped that, if madeto work in this mode for at least some of the time, thesepupils will become a bit more self reliant and a bit moreorganised.

Teachers commented on the improved work habitsand attitudes of classes working on the material andparticularly on the smooth and rapid start to the lessonas pupils came in, collect the material and beginworking. There is an obvious advantage in the methodfor dealing with pupils who are absent fora time.

I had expected that the pupils working throughmaterial by themselves would be much slower thanthose in classes paced by the teacher. The surprisingfind was that the teachers reported that it was in their'ordinary' classes that they had difficulty in covering thework in the scheduled time. This could be due toimproved work habits by those using the individualisedmaterial there is increased efficiency when a pupilspends time only on those parts he has difficulty with.

Although the method obviously enables teachers tospend more time with small groups and getting to knowindividuals, an additional comment made by oneteacher was that it was the first time he had been able tostand back and watch the class as a whole working andinteracting.

Teachers enjoyed the feeling of being organisedahead of time and also a non-specialist commented thatusing the programme enabled her to teach fully andenjoy a topic that she normally skipped through quickly.

The material used in this study contained no unusualmaterial or equipment; it is 'ordinary' and as such couldbe produced for any science topic by any school.

However, teachers may hesitate to embark on thepreparation of such materials because of the timerequired. There are long-term savings once the materialhas been produced. and the load can be reduced bysharing it amongst the members of a sciencedepartment and teachers can produce material fortopics in which they have paricular strengths.

There is also the problem that the school has to findthe cost of providing printed material. We provided eachchild with about 10 sides of A4 per week (excludingtests) and it seems unlikely that this could be reduced

43

significantly. One possibility isto extract all thosesections on which pupils write and produce this onseparate sheets. However. this would provide addedcomplications to the instructions and those sheetswhich were collected back may well have a very limitedlife anyway. With the cost of text books rising rapidlymany schools have had to move from pupil-texts toclass-sets which can be circulated around severalclasses; the use of printed material fits nicely into such ascheme and some saving on text book expenditure maybe made.

A complaint about using prepared material is that itprescribes the teaching programme and leaves no roomfor teacher innovation, but this can be avoided by usingthe prepared materials for certain topics only or for thecore work in each topic. Providing there is no attempt tobuild a completely integrated science programme ofprepared materials, teachers should have plenty ofscope for using their own initiative and ideas and pupilswill not become bored by a single method of working.

The Worksheets

The worksheets were designed to follow the course of alesson and contain:

1. blocks of information;2. questions that either:

(a) asked pupils to speculate before reading thenext block of information,Or

(b) led pupils to a conclusion via a set of cuequestions;

3. Instructions for performing experiments:4. summaries of theory or experiments that were

to be copied by pupils into their books.

There are a number of problems in producingworksheets:

1. Language difficulty: a conscious effort has tobe made to keep simple the general language(as distinct from any necessary technicallanguage).

2. Feedback on progress: a major criticism wasthat pupils had little information on how theywere progressing until the actual tests. Anumber of built-in, self-administered tests isprobably required.

3. Form of the worksheet: there seemed to be ageneral tendency for the less able to enjoycopying out material from the worksheets intotheir books whereas the more able found thisrather frustrating and would have preferred afill-in-the-blanks style of sheet which would letthem work faster.

Under the experimental conditions, pupils wereasked to restrict their co-operation with neighbours toa) the gathering of data in experiments where morethan nne person was required. and to (b) asking smallpoints of clarification of the textual material. Theserestrictions over five weeks Red to a boredom effect:

four times as many pupils to the experimental classesexpressed a negative attitude towards the self-pacedstyle of teaching. Teachers were similarly asked torestrict their activity to making equipment available.answering small points of clarification and, only whena pupil was obviously wasting a large proportion ofhis time or distracting others, act to control a pupil'sactivity. Teachers also expressed some frustration atbeing restricted.

The control classes worked through the samematerial under teacher direction with textual materialand instructions being read and displayed on an OHPor at times written on the blackboard. Teachers wereasked to restrict themselves to the textual materialprovided and not to encourage questions that mightlead to them giving large-scale interpretations orexplanations.

Teachers with control classes reported frustrationwith these conditions so regularly that the extent towhich they managed to maintain the conditions forthe whole time must be questioned. It is very probablethat at least on some occasions they providedexplanations and materials in addition to those on thesheet one is only human, a child's genuine requestfor information and help is hard to refuse.

Notes:

Two good articles about Student Team Learning are byJudge. Jean. in American Education of December 1978andPearson. Craig. in Learning of March 1979.

Information on STL (including an overview film strip andcassette tape at US$15.00) is available from Dr RobertSlavin, Student Team Learning Project, Centre forSocial Organisation of Schools. The John HopkinsUniversity, 3505 N. Charles Street, Baltimore, Maryland,USA 21218.

Jigsaw is described in several articles, for example, byAronson. E., in Psychology Today. Vol. 8, No.9, 1975.Further information on the technique is available fromDr E. Aronson, University of California. Kresge College,Santa Cruz, California, USA 95064, particularly inBender. R.D., Lewis. K.E., Schwartz, J.C. Talks to JigsawTeachers: A teacher's guide to the implementation ofthe jigsaw co-operative learning method. 1976.

Paul Power, PhD, is a lecturer in the Department ofEducational Psychology. Melbourne State College, 757Swanston Street. Carlton. Victoria. Australia 3053.

John Longbottom is Head of Science, Aranui HighSchool, Shortland Street. Christ church 6, New Zealand.He is able to supply examples of the science worksheetshe has described.

44

Item 9-- .,

J. eit........., , -=*-."...- i.-1...`"t..4*. . .0j,p, . : .o.:.,...;:.....4..,... Adek.... .

.J. :::* 111itallti,a4.'*". "1 ...

esd.. ,-

SOOAixe'

set Imeateto

ler INC WS

number one 1980

.. - ,.e.le"on:4km

tr,

The Class Size Issue RidesAgain

Starring Gene Glass and Mary LeeSmith, with a cast of thousands theresearchers and the researched.

Our Story Opens to Reveal that:

Teachers are almost universally agreed that large classes"are exhausting, a cause of frustration, and a reason forfailure in basic subjects". Their views about whatconstitutes a "large" class have changed considerably inrecent years, however. The same author quotes figuresfor England to show that the optimum size of class thatteachers thought most desirable was regularly about 3to 5 pupils less than the current average! Parents tend tosupport the teachers' views about the desirability ofsmall classes and frequently send their children toprivate schools for just this reason. Both teachers and

parents argue that with smaller classes there is lessstrain on the teacher, and he can give more attention toindividual pupils, and so improve achievement levels.Unfortunately, tie research shows that the situation isnot nearly as simple as this.

The Plot Thickens: A Brief Review of SomeResearch StudiesThe first experimental investigation was carried out bythe American, J.M. Rice, in 1902. He studied thearithmetic achievement of 6,000 children in grades 4-8in relation to size of class, amount of instruction time,and several other factors, and found that dass size hadno bearing on the results. He repeated the experimentwith a test of language in the following year, with thesame outcome. Rice concluded that there was norelation between size of class and the results, that someof the best work was done in the largest classes, andsome of the poorest in the smallest classes.

Countless similar investigations have beenconducted since Rice's day, but many have been poorlydesigned and executed. Writing in 1954, Blakereviewed 267 published studies on the subject, andfound only 22 that met the criteria of "scientificadequacy". Eleven of these had been evaluated forachievement test scores, and of these, 5 favoured small

45

classes, 3 favoured large classes, and 3 showed nodifferences. The other I 1 studies were evaluated by ananalysis of desirable class activities and teacherpractices, and all found small classes superior in theserespects.

Several large-scale surveys in the U.K. have notresolved the issue. The Scottish Mental Survey of 70,000eleven-year-olds (Maxwell, 1958) found no differencesin test scores in relation to pupiUteacher ratio. Alongitudinal study of a national sample of 5,000 U.K.children found that children in smaller classesperformed slightly better, but nearly all these childrenwere in private schools, so the results were notconclusive. Morris studied the reading progress ofEnglish children aged 7-11 years in 51 schools, andfound that "schools with an unfavourable pupil/teacherratio returned higher scores on the whole than thosewith smallerclasses!" She added that the large classesusually had other favourable circumstances, however.Peaker in 1967 found the same pattern in his nationalsurvey for the Plowden Report, and made the samequalifying comments.

More recently the problem has been tackledinternationally by the surveys of the InternationalAssociation for the Evaluation of EducationalAchievement (IEA). In the mathematics survey,students in countries with larger classes scored higherat the 13-year-old level, but the reverse was true at the17-18 year level. It was suggested in explanation that, atthe lower level, backward classes were reduced insize to assist their progress, and good teachers weregiven larger classes because they were able to cope withthem. At the higher level, there would seem to be atendency for the more advanced students to havesmaller classes.

The lEA surveys in reading, science and literature,published in 1973, showed some relationships betweenclass size and achievement but these were reduced tozero once the effects of home background and type ofschool/programme were taken into account. It is clear,then, that the problem of class size and achievement isfar more complex than appears at first sight. Many otherfactors must be taken into account, and only soundexperimental research will allow these to be fully tested.

The vast majority of research studies have usedachievement of pupils on tests as the main criterion. AnAmerican study by Olson in 1971, however, looked atother "indicators of quality", Teachers in 18,500primary and secondary classrooms were rated on fourcriteria of quality: individualization, interpersonalregard, group activity, and creativity. On these factors,smaller classes did produce significantly better results,and the relationship was consistent with classesranging from 5 to 50 pupils.

However, as no measure of pupil achievement wasused, no conclusion can be drawn about the importanceof these4 factors in promoting better learning amongstpupils.

. . . And Thickens: Factors Which Intervenein the Relationship Between Class Size andPupil PerformanceSeveral factors are believed to complicate the researchstudies and explain why children in small classes do notshow up to better advantage. Some argue that whenclasses are homogeneous in ability they can be taughtmore efficiently in large groups, although there is noactual research which supports this view,

Again, small classes are more commonly found inrural areas and in older urban districts with bad socialconditions. The children in such classes are less likely toperform as well as those from urban middle-classschools which frequently have large classes. Of coursethe opposite would be true of the small classes found inelite private schools. Slow learning students placed insmall classes would also complicate the researchpicture.

Size of class also varies according to level ofeducation. Smaller teacher/pupil ratios are found in theupper secondary school and universities, although theresearch has rarely supported this trend. Justification isfound rather in the need for more space and morespecialised facilities for the older, more advancedstudents.

Anothercomplicating factor is the use of ancillaryassistance in the classroom. English studies haveshown that only about 43 per cent of the teacher's timeis spent on lesson instruction. Therefore, if teacher aidswere used, the teacher's efficiency would be increased,and the class size maintained at a higher level withoutloss of pupil achievement.

More important still is the behaviour and method ofthe teacher. Most would agree that different methods ofteaching are best suited to different sized groups, andthe more modern methods which focus on the waypupils learn rather than on the way teachers teachnecessarily require smaller classes. Thus the lecturemethod can be used with equal efficiency for 200students or for 10, but when used with large groupsthere can be little individual interaction betweenteacher and student. Even at primary and secondarylevels, large groups of 100 or more are quite satisfactorywhere pupils are watching films or listening toillustrated talks. As soon as pupil participation isrequired, however, smaller classes become necessary,and with methods demanding close and continualinteraction between individual pupils and theirteacher, the smaller must the classes become to remainefficient. As one teacher put it, it is impossible to takean oral approach to French with classes of 40 pupils.

However, there is research in both England andAmerica which shows that teachers do notautomatically change from traditional lectureprocedures when their class sizes are reduced. This factwould help explain why smaller classes do not showbetter in research studies. Improvement in pupilachievement will only follow from a reduction of classsize if the teachers change their methods to takeadvantage of smaller groups, and if these changes areaccompanied by genuine changes in the attitudes andbeliefs of the teachers concerned.

Thus, the main reasons why smaller classes have notshown better achievement in research to date are:

46'

(i) The continued use of traditional teaching methodsin small classes.

(ii) The placing of low ability students in small classes,and vice versa.

(iii) The failure to take socio-economic and urban/ruralbackground into account.

(iv) Theconcentration of research workers on a narrowrange of curricular objectives.

(v) The use of class size rather than pupil/adult ratio,thus ignoring the flexible useof staff for varying sizegroups.

Our Heroes Arrive and Start Unravelling thePlot . . .Gene Glass and Mary Lee Smith of the Laboratory ofEducational Research, University of Colorado, wroteMeta-Analysis of Research on the Relationship of Class Siteand Achieoenient in 1978. It reports on a new technique ofintegrating the results of previous research. Theirconclusion: that average pupil achievement increases asclass size decreases.

Their research began with a four month literaturesearch which turned up 300 reports, articles, theses, etc.They found the Australian Education Department reportof 1974 and the Ontario Ministry of Education report of1975 particularly helpful. Only 77 of the documents couldbe used. These yielded 725 comparisons of achievementin different class sizes. The data came from studies of900,000 pupils back to 1900. Included were studies of theeffects of tutoring in very small groups of one, two, or afew pupils. Class size was defined as the number ofpupils being taught by a single teacher, giving aPupil:Teacher ratio (P11) for a class of thirty taught by oneteacher as 30, fora supplementary maths class of fourwith a teacher as 4.

The results of the 77 studies were all expressed indifferent scales. These had to be reduced to show whatGlass and Smith call achievement advantages, andexpressed in a common way. This score (called a "delta")was the mean achievement score for the smallerclass in astudy minus the mean for the larger class, the differencebeing divided by the within-group standard deviation.To illustrate, Class A has 10 pupils, Class B in thesamestudy 20. The students have been given a test with 50items. The mean for Class A is 35, the mean for B is 30.The standard deviation for Class A and Class 13 is 10. Thedelta for this hypothetical case is (35-30) + 10 = 0.5.

Such calculations are easy when means and standarddeviations, etc., are given. Hut many were not and Glassworked out ways of estimating them from other datagiven.

Of the 725 deltas calculated 60 per cent were positive,indicating that achievement was higher in smallerclasses. The average was 0.9. Further analysis revealedtwo important interactions, The size of the differencedepended on the size of the classes being compared. Italso depended on the quality of the research design.Effects were stronger in studies having good designcharacteristics, Further mathematical analysis producedthe now famous graph.

Imagine a typical pupil in a typical class of 40, He canbe described as being better than 50 per cent of all thepupils of his class-level in a country-wide achievementtest. The same pupil in a class of 20 would be better than

55 per cent of this group. In a class of 15 he would bebetter than 58 percent. In a classof 10 he would be betterthan 65 percent. And being taught in a group of five hewould improve his achievement to exceed that of 74 percent of his class peers.

Another example: the average pupil in a class of 40might be expected to gain one year's school knowledgein a year. In a class of 20 he could be expected to gainapproximately I 1/2 year's knowledge. In a class of five hecould be expected to gain 1.7 year's knowledge in a year.

In the figure, as in the examples, the curve starts to risemost dramatically when class size is reduced to below 20pupils. Such improvement is not of course automaticand reties on the better use the teacher makes of theopportunities smaller classes give. Instances ofsubstantially larger classes out-performing smallerclasses have been recorded. Glass and Smith point outthat "researchers must take account of what actuallytakes place in smaller classes: the instructionalprocedures used the beliefs and capabilities of theteachers, the demonstrated abilities and backgrounds ofthe pupils, the subject matter, and the like. Theseultimately determine whether the potential for increasedlearning that smaller classes create will be realised".

Glass and Smith found further that the relationshipbetween class size and achievement was slightlystronger in secondary school than in primary school, butnot strong enough to lead to the conclusion that primarypupils would profit less than secondary pupils if classsizes were reduced. There also appeared to be nodifference in the results for reading, mathematics, andlanguage.

The overall difference in results between the wellcontrolled and the poorly controlled studies wasdramatic. The curve for the poorly designed studies wasalmost flat, indicating, at best, a very smalladvantage tosmaller classes. There was no correlation between classsize and achievement in studies done before 1940.

1 low does one judge the importance of the differencesshown in the figure? With such a small decrease inachievement between class sizes of 30 and 40 could it beargued that in these straightened times a return to 40pupils per teacher would do little harm?Glass and Smithare continuing with their work and are now relatingclasssize to classroom practice, and field studies are beingconducted to see what changes take place in classroomswhen they are reduced in size. In two schools four infantclasses, each with a single teacher will be observed forhalf a year and then extra teachers will join the staff, theclasses will be made smaller and the classes observed forthe second half of the year. Such research as this willperhaps show why the increases are so dramatic as classsize decreases below 20, However, in the meantime, it is

47

clear that increasing class size to save money will causepupils to learn less. The effects such moves would haveon teacher morale can be best spelled out by teachers'unions.

Would it be economic to increase achievement byreducing class size dramatically to say, below 15?Already we reduce class sizes for some pupils, forexample, backward readers, and reading recoveryprogrammes report that the gains children make areworth the expense. The employment of less expensiveparaprofessional or ancillary staff can free teachers forprofessional duties (that is, for teaching, which is oftenpushed aside by administrative chores). Grouping andindividualised programmes are also methods of cuttingdown the instructional group size, if not the class.Staggered hours so that children "glide" in, somecoming early for reading instruction and going early,others arriving later for school and reading, and goinglater, would also cut down the size of the instructionalgroup.

In the end, however, weighing up outcomes againstcosts is a question of values. Glass shows a clearrelationship between class size and achievement. Whatvalue does society, and therefore the school, put upon aparticular magnitude of improved achievement, andupon those values the school can give that are notmeasurable?

After the Cheering Has Died Down: OurHeroes Methods Are QuestionedIn essence Glass and Smith's techniques boil down toan attempt to quantify to put into numbersreview of all the available literature on the subject. Thedanger of this is that it lends the results an appearanceof more accuracy and veracity than they may deserve.There are two important and related points which needto be made and which should influence anyinterpretation of the findings.

(a) There are very few studies which contrast smallclasses (i.e., 1-10) with classes a little bigger thanthemselves (i.e., 2-16) thus the graph as drawn mayreflect the influence of a very small number ofresults.

(b) A substantial proportion of the effect produced bythe analysis may result from the comparatively smallnumber of studies which contrast individualinstruction with larger classes. The inclusion ofthese studies may give the Glass Smith graph anunwarrantedly sharp bend and exaggerate the realgains to be made from reducing class size where, forus, it is most realistic, e.g., from 30 to 25.Nevertheless, the graph would be expected to followthe same shape. even if the one-to-one classes wereomitted. However, the slope would be less steep atmoderate class sizes and apparent gains fromreducing class size much reduced. The slope at verysmall class sizes would be, however, much steeper.

Redrawing the graph to take these considerationsinto account could possibly reveal that Glass andSmith's results have tended to overestimate the gains inachievement resulting from a reduction in class sizefrom 30 to 20, but have underestimated the gainsresulting from individualised instruction.

To fill out (and perhaps to justify), the Glass andSmith predictions of large gains in very small classesmany more pieces of research on the differences

between classes with few members are needed, forexample, classes with three, four, or five memberscontrasted with classes of six, seven, and up to 16members.

Overall the Glass and Smith finding of increasedperformance from smaller classes holds. But the effect isprobably less than they show. Alas, the fate of ourheroes is, in the meantime, to have their importance,like their graph, lowered.

Credits:The first part of this article was written by WarwickElley,now Reader in Education, University of the South Pacific, Fiji,for New Zealand set in 1475, summarising the work ofDouglas Pigeon in 1973 for the OECD, published by them as"Class Size as a Factor of Pupil Performance: A PolicyAnalysis".

The second part, Our Heroes ArriN ., is a summary oftwo articles: the first by Gene Glass, Leonard Cahen, MaryLee Smith, and Ni kola Filby, "Class Size and Learning newinterpretation of the research literature" in Today's Education,April-May 1979; and "The Class Size'Achievement Issue: NewEvidence and a Research Plan" in Phi Data Kirwan, March1979, by Leonard Cahen and Nikota Filby.

The third part, Mier flee Cheering . . ., is by Peter Jackson ofthe Test Development Unit of the New Zealand Council forEducational Research and is based on examination of Glass,G.V and Smith, M.L., AlefaAnainsis of Research on theRelationship 01 Class Size and Achievement, Boulder, Laboratoryof Educational Research, University of Colorado. 1978.

A report drafted by Mudina Campbell of the New ZealandDepartment of Education Research and Statistics Division wasof help to the edi tor.

References:The quotation about large lasses being exhausting, etc., isfrom Fleming. C.M.. "Class Size as a Variable in the TeachingSituation" in tihicational Research, 1, 2, 1959.Olakt's 1934 review of class size research is found in Blake,11.E., "Class Sire: A Summary of Selected Studies inElementary and Secondary Public Schools", Ph,D,dissertalion. New York, Teachers' College, Columbia, 1954.The Scottish Mental Survey can be found written up inMaxwell, J., Educananal and Other Aspects 416e J9.17 secul,A,Alettht1 Surrey, London, ULP, 1958.The longitudinal study of 3 000 U.K. children can be found inDouglas, J.W.11., The dame not the School, London, McGibbonand Kee, 1964.Morris, who found higher scores when: the teacher/pupil ratiowas unfavourable, can be read in Morris, J., Reaelinx in thePrimary School. London, Newnes, 1959,reaker. C.F.. wrote "The Regression Analysis of the NationalSurvey" which is Appendix 4 in the Plowden report, Chddrendod the Primary Schools, London, 11.M.S.O., 1967,The American study which looked at indicators of qualityother than academic success was described in Olsen, N.M.,"Ways to Achieve Quality in School Class ((((( rns: SomeDefinitive Answers" in Phi Dena Kappa,' , 53, n., pp. 63-65.1971.

The Australian report on class size is Lafleur. CD.. Sumner,R.J., and Whitton, E., Class Sin. Survey, October. 1974,Canberra, Australian Government Publishing Service for theDepartment of Education, 1975.

The Canadian report is Ryan, D, W., and Greenfield, T.0., TheChess Size Qneshati: flevelopmern Of RiNeare.h StnItes Related tothe EffraN Si:.'. Pipit/Won mid PopilrfaherToronto. Ontario Ministry of Education, 1975.

Item 8Set110801100%la *Wan

number one 1980

,',- Small is BeautifulSome Effects of School Size

Upon the Behaviour of Pupils

49 50

Small is Beautiful: SomeEffects of School Size on theBehaviour of Pupils

By Jack Campbell,Professor of Education,University of Queensland.

Introduction

Is the issue worth investigating? After all, numerousstudies have shown that the home environment is amajor influence on behaviour, and, within the school.more promising variables for study are teacher-pupilrelationships, the curriculum, and specific teachingstrategies. I concede at the outset that the size of aschool which a child attends is unlikely to be a MajOrinfluence on his behaviour, but it could be a verypervasive one millions of children in Australia andNew Zealand attend schools, and even a minorinfluence would, in total, be very substantial. Moreover,however minor its effects might be. school size. unlikecertain aspects of home background, for example,could be controlled if need be.

The question of whether or not the issuecan beinvestigated convincingly is more difficult to answer.Although many of our everyday actions are based on theassumption that our physical situation can affect uspsychologically, it is notoriously difficult to present aconvincing demonstration of this relationship. However,I want to try to proceed as far as possible in identifyingeffects which flow almost inevitably from school sizeand are not influenced to any great extent by otherschool factors or even the personalities of teachers orpupils. Sooner or later, of course, I will have to leave the

51

shelter of allegedly inevitable relationships, and enterthe stormy water of probabilities.

Characteristics of Large and Small Schools

The one absolutely sure thing is that, in comparison withlarge schools, small schools have fewer pupils! Bothlarge and small schools can be further characterized bywhat they offer their pupils, for example, regularclassroom lessons, special-occasion events (such as avisit to Parliament House), and extracurricular activities.In the jargon, these offerings are called 'behavioursettings'. Study after study has shown that although thebigger schools tend to have a bigger list of behavioursettings, the difference in number is often small, and thedifference in variety is frequently nil. The smallerschools tend to offer the same kinds of things as do thebigger ones, but there might not be as many instanceswithin each variety. Thus in a recent study which someof my colleagues and I have just completed, largeschools with enrolments above 700 offered their Year 7pupils an average of 60 behaviour settings (there were 6varieties) whereas small schools with enrolments below400 offered their Year 7 pupils an average of 50behaviour settings within the same number of varieties(6). On the one hand, leachers in the small schoolsappear to make special efforts to ensure that their pupilsare not disadvantaged in comparison with theirlarge-school counterparts, and. on the other, largeschools appear to be unable to increase offerings in linewith increasing enrolments.

Some 'First-Order' Consequences

The first 'almost inevitable' consequence is that pupilsin small schools participate in more activities, a greaternumber of key positions. and a wider range of activitiesthan pupils in large schools. There is nothingmysterious, or even psychological, about this; it is littlemore than a matter of arithmetic. With fewer pupilsavailable to 'man' the behaviour settings, it follows that

greater involvement will be required from those who arethere. Ina very small school, every pupil might be calledupon to play a key role in every activity, whereas in avery large school many pupils are likely to be redundant.

A second first-order consequence is that pupils insmall schools will experience greater challenge andgreater support than pupils in large schools. EachbehavioursettIng-cdntains a system of controlsdesigned to maintain it intact and functioning at a stableor near-stable level. Thus when pupils are in abundance,the less competent are likely to be relegated to onlookerroles, but, when pupils are in short supply, some arelikely to be involved above their levels of competence. Inorder to ensure that the setting functions adequately,these latter will be given supportive guidance.

The picture that has been built up, then, is of pupils insmall schools participating at higher levels in a greaternumber and variety of settings whose psychologicalclimates are more challenging and supportive. Everystudy of school size, irrespective of teachers, pupils, orother variables, shows this to be so and the reason isthat these firskirder consequences are based on howbehaviour settings function under conditions of'over-manning', 'optimal-manning', and'under-manning'. In a sense, they areprocesses ratherthan outcomes.

Some people may be impressed by these first-orderdifferences in behaviour, but others are inclined to say'O.K., but so what? Do these differences in behaviourwithin schools of different sizes persist and/or influenceother more significant behaviour?'

Persistence of First-Order Behaviour

Two of our recent studies suggest that there is at leastsome persistence of behaviour generated by large andsmall schools. In the first study we looked at one smallschool which had been established, a year before thedata were gathered, by drawing pupils from a very largeschool (1600 +). When we presented the computer withall the data in the study and asked it to cluster theschools according to the behaviour profiles of their

52

pupils, it made an almost perfect separation ofsmall-school and large-school groups but placed thisnewly-established small school with the large schools.Obviously the pupils from this particular small schoolwere still, alter one year, displaying the first-orderbehaviours which are typical of large-school pupils.

In the second study attention was focused on thebehaviour of pupils who had attended small, ruralschools for the first ten years of their schooling, but hadThen had to move to large secondary schools for Years11 and 12. Typically, a small group of around lenstudents from a 'decapitated' school had joined aboutone hundred and fifty others who had been together forat least three years. In these circumstances, one mightnot expect the ten commuters to make a major splash inthe social and extracurricular pools of their new school,but we are finding that they do. By very much more thanchance would suggest, the Year 12 prefects andcaptains are being drawn from the groups of pupils whohad spent their first ten years in small schoolselsewhere. Before coming to their new large schools,they had, of course, been the most senior pupils in theirsmall schools, and, almost inevitably, had occupiedpositions of leadership.

Influence of First-Order Behaviour on OtherBehaviour

II one assumes that processes of development arebasically similar for most people, and that experiencesimpinge similarly upon most, it is possible to make somepredictions from the first-order behaviours tosecond-order ones which are 'probabilistic'. Animportant prediction is that:

pupils who participate at higher levels in a greaternumber and variety of settings whose psychologicalclimates are more challenging and supportive (inother words, small-school pupils) will develop:1. a stronger sense of cohesion, 2 a more favourableattitude towards school. 3. a greater concern forpersons.

53

Working with Year 7 pupils who had spent al least fiveyears in their small or large primary schools, we haveconfirmed these predictions. After account has beentaken of all other factors, it appears that between 10 and20 per cent of the variance in scores on those threeoutcomes is explain-Jo by the first-order consequencesof school size. Perhaps it was these second-ordereffects (stronger sense of cohesion, a more favourablealtitude towards school, and a greater concern forpersons) which led to the ex-small-school pupils quicklymoving into positions of peer and school leadershipwithin the large schools to which they moved after Year10.

How Significant are the Findings?

Our findings are that:

1. in comparison with large-school pupils, small-schoolones participate at higher levels in a greater numberand variety of settings whose psychological climatesare more challenging and supportive;

2 this orientation towards involvement persists, at leastfor a short time, when pupils enter contrastingsettings;

a the first-order effects appear to generatesecond-order ones pertaining to sense of cohesion,attitude towards school, and concern for persons.

I suspect that the significance ascribed to thesefindings will depend largely upon the educational valueswhich one holds. Industrially-oriented readers will notethat I have presented no evidence to hint that schoolsize effect work-force skills. Nor have I presented anyabout different knowledge in the basic subjects ofreading, spelling, writing, and arithmetic. These readersare likely to conclude from this that the economies ofscale. which bigger schools are said to have, might aswell be enjoyed. Similarly, academically-orientedreaders are likely to dismiss the findings as insignificantbecause they say nothing about either the processes orproducts of thinking. I respect the value stances of thesetwo groups of readers. and agree that the educational

outcomes which they hold dear are important. Ouryoung school leavers should be able to earn a living andcontribute to the economic growth of our countries.These are, of course, important assets both for theindividual and for society. Young people should, too(particularly as we begin to move into a post-industrialera), possess the intellectual skills needed to gather,store, and retrieve information. These are all necessaryattributes. But they are not suflicient for living in thismodern world. In addition, our young people need themotivation which will assist them to build a morehumane world and overcome the challenges of war,poverty, pollution, ignorance, disease, and diversity ofvalues within and between nations. I believe that thismotivation is likely to be associated with a concern forpeople.

A favourable altitude towards school is no lessimportant. As Bill Renwick has said, 'The extent towhich each person is equipped to respond positivelyand confidently to the educational challenges that willface him as an adult will greatly depend on the altitudesongendered during his initial encounter with theeducation system.' In the interests of continuingmotivation, which in our changing society is emergingas a more important educational outcome than specificitems of information, we need to foster a favourableattitude towards school among all our pupils.

Few would wish to claim that by changing the schoolswe can guarantee the creation of a significantly morehumane world, and the fostering of continuingmotivation in the young people. Nevertheless, if, bychanging the size of schools, some improvements canbe effected in these areas without loss to otherlegitimate goals, change is warranted.

What should be done?

Three common-sense solutions suggest themselves:

(a) expand resources and facilities in large schools;(b) create a number of semi-autonomous schools within

each of the large ones; and(c) ensure that the schools are kept small.

54

The first of these solutions is frequently difficult toimplement there is a limit to how many football fieldsor netball courts one can place on a school site and isunlikely to work, anyway. Better resources and facilitieslead to better educational experiences only when theyactually impinge upon the pupils, and all of the researchevidence on school size suggests that higher densitygenerates lower involvement on the part of the pupilsdespite the resources made available.

The second solution sounds plausible. but ineducational matters we have all been fooled before byplausible-sounding. 'common-sense' solutions. A recentstudy by Elizabeth Campbell suggests that 'vertical-unit'schools are much the same as others of the same totalsize in matters of 'flexible learning structures'. It appearsthat the negative relationship between school size andflexible learning structures is deeply based and difficult,if not impossible, to avoid. Nevertheless, there shouldcontinue to be active experimentation with differentstructures within the large schools which currently exist.The whanau-house concept in New Zealand seemsparticularly promising.

The third solution. which involves the creation, ormaintenance, of small schools, appears, at the moment.to be most worthy of support. Research does notprovide data on what the optimum size of schools is. butit suggests that when schools have enrolments above700 it is difficult to ensure that the pupils participate athigh levels in warm and challenging learningenvironments. At the other end of the scale, there is ahint that very small schools within communities with a

low rating on a socio-economic scale frequently requirecompensatory assistance.

It might be claimed that I am flogging a dead horse,for most of what is contained in this article has beenacknowledged by teachers for a long time. Despite thisacknowledgement, however, one of the most strikingtrends in education during the last half- century, whenpopulations have been rapidly increasing, has been areduction in the number of schools. This reduction. ofcourse, has been accompanied by an increase in theaverageenrolments. ft is a depressing experience evento visit some of our large inner-suburb schools and seeup to 2,000 pupils being packed into areas that containonly a ramshackle set of buildings and no playing areas.What must it be like to learn there? At no other stage intheir lives will persons occupy such densely-populated,constrained settings for such extended periods of time.The industrial arguments of scale still dominateeducational thinking in this matter: big schools providebetter resources; better resources mean betterexperiences for pupils. f4eitner of these arguments issustainable; it is time to announce that the Emperor hasno clothes.

Notes

Characteristics of Large and Small SchoolsThe recent study mentioned is by Campbell, WA,Cotterell, J.L.. Robinson, N.M.. and Sadler, D.A.. Effects

of School Size on Some Aspects of Personality. Reportto the Education and Research and DevelopmentCommittee. Canberra, 1979.

Persistance of First-order Behaviour

The first study mentioned is that just above. The secondstudy is by Campbell, W.J.. and Robinson, N.M., SomeEffects of Consolidation Upon the Duality, Efficiencyand Economy of Education in Rural Communities,Report being prepared for the Education Research andDevelopment Committee, Canberra, 1980.

What Should be Done?The study on vertical-unit schools is by Campbell,Elizabeth M., The Relationship of School LearningStructures to Some Reactions of Students. Report to theEducation Research and Development Committee,Canberra, 1979.

How Significant are the Findings?Bill Renwick is Director-General of the New ZealandDepartment of Education. His remarks on attitudes toschool are from Renwick. WI., Social Change and theObjectives of Secondary Education in Cods]. IA.. andHermansson, G.L. (Eds.), Directions in New ZealandSecondary Education, Auckland, HodderandStoughton, 1976.

J6 6

Item 5,fit (NZ)

number one1980

Projecting EducationExpenditure

By Ian McGill,Department of Education,and others.

' In 1979180 the government set aside S1,00,1-10,000 for

..,

sedpuecnadtiinogn.ofintEadt uisc.5u.aloificosubreeGnminsscreD.oisminegstbicuPt ronodwuct.!.."

4, ._Am. immigration

ceased, emigration is at an all time high, times are tougher

Experts on population changes and finance can look,--- into the future and show us our options reasonably

.....- clearly, but they cannot predict the future. Calculated.., ..- guesses about school enrolments will, for example, help

us invent possible ways of spend ing our taxes in thefuture, but they will not tell us how the money will in fact

..-- be spent. Everyone must decide which possible solutionfulfils their hopes best and then tell the politicians; theyhave the job of finding the most acceptable solution.

_.....------ '...., Demographers (population experts) like to keep their....

...... options open, so they present us with different models..,"For example, we can assume that the birthrate stayssteady for some time to come and the effect, coupled with

... zero migration (immigrants and emigrants cancelling each...., Fiona Kelly

... other out) will give us the model of population growth- --..:-,.:.,... ,.. -..z...,:. 40 seen in Chart 1 on the next page.1..

,

N

58

Chart 1 Total Population Model A - steady birthrate,zero migration

4-

3-

2 -

0 I1976 1981 19ti6 1091 1096 21101 2006

However, we could assume that the birthrate willcontinue to fall, and that we will lose 10 000 people peryear. This will give us model 0: Chart 2.

Chart 2 Population Model D - birthrate dropping,emigration continuing

.4

3

0 -197h 1081 1986 3991 1996 2001 2006

Sy

There are models in between.Since these projections were prepared actual emigration

has been greater than assumed in D and in 1979 thepopulation has fallen already. Is this decline a trend or ahiccup?

Even if it is assumed to be a hiccup from the point ofview of educational planning, it is not so much the size ofthe total population that is of special interest but of certainage groups within that total:

1. pre-schoolers 0-4 years2. primary schoolers 5-12

3. secondary schoolers 13-14 (compulsory)

4. secondary schoolers 15+ (voluntary)

If we compare the two models with our base year of19Th we get chart 3.

Chart 3 Projected population in 2006

500 -

450 -

4110 -

350 -

31111 -

250 -

200 -

150 -

1011-

50

11

1076 {bast) Model A

n 0-4 years

5.12 years

13.14 years

Model D

Model A shows a decrease in the size of the age groupsof about 10% despite the steady rise in the population ofthe whole country.

Model D shows a decrease in the size of the age groupsof about 40% although the total population is bigger thanit was in 1976.

Most demographers suggest that Model D is the mostplausible for planning purposes- they know that in thepast they hai.e consistently over-estimated and even Dmay be rather optimistic considering our 1978 decline. Allother things being equal it implies a one-third reductionin our demand for maternity beds, paediatricians,teachers, schoolrooms, training college students, and soon.

All other things are not, of course, equal. Some sectionsof the community will be hit more than others; the effecton rural areas may he devastating. First, however,finance.

In 1978-9 $936,671,000 was spent on education underthe title 'Vote Education' by the government and therewere 1,006,864 students of one sort or another fromKindergarten to full-time University students andpart-time adults at night school. The simple division sumgives us $930.29 per student but we should look deeperthan that. After all each playcentre child costs thegovernment about $82 per yearand the Minister of Healthrecently gave the cost of training a doctor as $12,500 peryear.

Education Expenditure

'Vote education' is divided into seven descriptiveprogrammes further divided into activities. A condensedversion for three years is shown in chart 4.

Chart 4 Vote Fducation in millions of dollars

Programme

1 Ad min i-.I rat ion

1071-2 1976-7 1079-80sestimated)

and General 7.6 13.3 20.7

11 Pre-Chon I 2.4 5.9 15.6

III Primary andsecondary 189.4 384.7 614.5

1V Continuing 77.2 175.2 279.5

V NationalLibrary 2.1 4.11 7.4

VI Research - - 1.8

VU Buildings 52.2 118.5 107.6

TOTA I. 331.0 704,7 1047.2

If we look more closely at primary and secondary schoolcosts we get chart 5.

Chart S Primary and Secondary Education

Year-Numberof State

1971-2 197h -7 1979-80

Primary Pupils 408 207 473 113 407 51111

Number of StateSecondary Pupils tot 442 197 912 107 410

Total Pupils 029 709 073 023 004 910

- - - -

State PrimaryExpenditure $ 011.8m S196.4m 5301.8m

State SecondaryExpenditure $ 06.1m 5t31.4m 5213.3m

Joint Primary andSecondary Evend i ton::un administration.special education,buildings, etc. S 31.6m $112.8m $145.0m

Total Expenditure $219.3m $440.6m 5662.7m

Cost per StatePrimary Pupil 217 S 413 S 040

Cost per StateSecondary Pupil S 409 5 00-I S 1091

Average Joint COMper State Pupil S 82 S 108 5 219

Average Cost perState Pupil 5 349 S 633 S 997

State Primary, Secondaryand Joint Costs:as a perm, logeof total cost ofVote Education

61

050%

Since 1971 direct expenditure per primary student hastrebled and for secondary students increased abotit twoand a half times. The increases have not been even 'acrossthe board' administration and special education haveincreased more than school buildings.

These figures must now be adjusted for inflation.

Chart 6 Cost per Student (in 1979-80 dollars)

Year 1971.2 1970-7 1979.80

Cost per StatePrimary l'upil 373 S 018 S

Cost per StateSecondary Pupil 5 1079 ti 993 S 1091

Average Joint Costper State Pupil $ 216 5 251 5 219

Average Cost perStale Pupil 5 921 5 980 S 997

Not simple averagesfrom the figures above.

So in real terms, at most, there was an 8.3% increasebetween 1971 and 1980 (the highest average cost perstudent was in 1978-9: 51021. The small decrease of 2.4%from 1978-9 to 1979-80 was mainly due to a reduction inspending on buildings.).

The proportion of Vote Education going to Primary andSecondary Schools has not varied to any great degree.

The Future

Research in the USA and Canada has shown that whenrolls fall, spending on education does not fall at the samepace. This has been taken into account in the following'scenarios'. Primary and secondary school costs have beenkept together for statistical reasons, despite the different

02.5"0 pace at which their rolls will fall. The average cost perpupil from 1976 to 80 has been used for the calculations.

Chart 7 Four Scenarios for Educational Spending

AverageChange in Cost perSpending Enrolments Pupil

W Severe cutsFalling enrolments

X Moderate cutsSteady enrolments

Y No cutsSteady enrolments

Z Spending upEnrolments up

- 15":. low S 835

medium 5 953

0 medium 51003

+10 ". high $1103

These scenarios are further complicated if we take intoaccount possible changes in school enrolments; forexample, in times of economic hardship children may stayon at school trying to get higher qualifications, orenjoying the social life, rather than joining theunemployed.This gives us four scenarios worth considering.

Chart 8 The Four Scenarios' Effects on Future Spendingon Primary and Secondary Education (in 1980dollars)

19811-1 1964.5 14911-1

w $302.7m 5520.8m S443.9m

X $028.0m S593.7m 5539.4m

Y $60 I.2m S624.9m $507.7m

$22ti.om $090.4m 5049.$m

The most surprising result is in scenario Z where anincrease in spending per pupil of 10%, and a highprojection of roll numbers still results in a real reduction inspending on education of 2% by 1991.

But before we point out this automatic decrease ineducation spending to those who wish to cut theeducation vote there are a lot of other factors to beremembered.1. Total school rolls decrease in all these scenarios but the

mix does not remain constant. For example in scenario

62

W in 1980 secondary schools are 30% of the totalenrolment, in 1988 they are 36% of the total school role.The cost of a secondary pupil is 60% more than aprimary pupil so scenario W may be a further 3% moreby 198S. If many of the extra secondary pupils are 6thand 7th formers more teachers will have to beemployed, raising the costs further.

2. Any increase in the proportion of the educationallydisadvantaged through social, economic, or otherreasons may require further expenditure per student.

3. At out 70% of educational expenditure is on people sosalaries negotiated above the 'ordinary' inflationarymoves for example to attract scientists andmathematicians into secondary school teaching willchange the costs per pupil.

4. These figures make no allowance for the impact of rolldecline on individual schools and teachers. Teachers'associations and parents will press for smaller classes,strong pressure groups will oppose the closing of smallrural schools, early retirements will have to be paid for,busing pupils costs money and closed schools are notusually sellable, senior staff will probably be kept onrather than new inexperienced teachers employed andsenior staff cost more, teachers' associations willnegotiate salary protection measures as school rolls andgradings drop, and a significant part of schoolexpenditure is not directly indexed to roll numbers.

5. Private schools integration costs money.

6. Unless the number of administrative and support staffdrop proportionally with school rolls, costs per pupilwill rise.

Implications

Staff

As some 60% of state school expenditure is on teachers, iffalling roll numbers mean that salaries are renegotiatedand put the salary bill up 2%, then in 1990 we may havethe situation shown in Chart 9.

Chart 9 The Four Scenarios' Effects on Teaching in 1990

7el:. 4. 'f, '7, 7.. ." = .t. a,

J

19795307,nm 512.258

1990

Z

32 438 1,1.4 glo

tz is

1:20.5

,. e e a .:f

11. S2nn3rn 512,503 21 300 5114 300 1:23.7 .16% 34 "d

X $323.om 512.503 25 880 558 400 1:21.6 - 20";,

Y 5340,om 512.503 27 240 558 400 1:20.5 0 lt,".Z 5.380.0m 512,503 31 180 583 700 1:18.7 + 9" 40;.

Competition for funds

Vote Education competes with other major centralgovernment activities such as 'Health', 'Transport andCommunications', 'Social Services' and 'Development ofIndustry' fora share of government expenditure.Although Gross Domestic Product increased 22% in realterms over the last decade expenditure on educationincreased 32%. These rates of increase appear unlikelyover the next ten years.

State school education must in turn compete with othereducational activities; pre-school, university, technicaland community education and, of considerable futureimportance, integrating private schools. In 1979-80 theprimary and secondary private school appropriation isestimated at S27,650,000 for 80,480 private students or anaverage of 53-1-1 per student. By 1990 some 80% of privateschool students are expected to have integrated into thestate system. This will create, on the basis of averagecosts, an additional demand in school education for some528m under scenario W assumptions or 546m underscenario Z.

Resource Switching

An increase in Vote Education in real terms seemsunlikely it would involve higher taxes or a change inpolitical priorities. An alternative, pleasing to those who

get the increase, is to switch resources and expenditurefrom one area of education to another: the same sized cakebut changes in the size of the slice you get. This hasalready happened, for example, as expenditure onteachers increased with improving staffistudent ratiosfrom 1978 to 80, and spendinb on school buildingsdeclined.

Conclusion

There is no conclusion in this p .per. Our governors needguidance in the face of falling mils and economichardship. A clear set of goals for eaucational expenditureis needed. What do we want education to do? WilliamTaylor, Director of the University of London's Institute ofEducation, whose declaration that education must beaccountable has delighted many, said at the same time,that the case for school education rests not just oncompetence but also on culture and citizenship.

Notes

The material in this article has been abstracted from threepapers. The first two are by Peter Ramsay and James O'Neill ofthe University of Waikato and the third by Ian McGill of theResearch and Statistics Division of the Department ofEducation.

All three papers include extensive bibliographies of sourcesand further reatling.

O'Neill, C.J. and Ramsay, P.D. K.. 'Educational Futures: Theimplications of Demographic Change for Educational Policy' inNew Zealand lintrinil of &hint:num' Studies, Volume 14, No. 2,1970,

C.J. and Ramsay, P.D.K., Demographic Tracts in Newzealand and their Relation hi to filactilional Policy Malattg.Unpublished 1980,McGill, Ian, Projections of Education Even/Wiwi'. Unpublished,1980.

The two unpublished papers are available in photocopied formfrom NZCER.Ian McGill's paper (and the part of it used in thisset item)expresses the views of the author and does not purport torepresent the views oldie Department of Education or anyother official body.

Peter Ramsay and James O'Neill's papers contain carefullydetailed descriptions of the effects of whotesa)e cuts ineducational spending on rural schools and the quality ofteaching and draw conclusions about the opportunities forbetter quality education which falling rolls, and a good supplyof teachers provide.

63 6

ilenpd 43K199

096L auo Jaciumu

elm alemem a Ul

LialP1143 JoSOIWOU003 ayl (ZN)

t, well

The Economics of Children in aWelfare StateBy Brian Easton, University of Canterbury

The Concept of Human CapitalIn recent years economists have extended the notion ofcapital to cover similar phenomena in humans. For aneconomist a capital item generates a stream of futureincome. The traditional idea of capital included suchitems as machinery, buildings, installations, and the like.However, juit as a capital good put to productive usegenerates income so does a human. With each personwe can associate a future stream of income (but in somecases, such as parenting, it is generated in the informaleconomy and so does not appear as a financialtransaction). As in the case of physical objects we candiscount this income stream into a capital value.However, unlike objects, humans may not be boughtand sold in the market (with some exceptions as in acontract to supply services), so the market value ofhuman capital is notional, It can also be very high. Thecapital value of a person in his or her early twenties islikely to be twenty times his or her annual earnings.

One of the earliest uses of the human capitalapproach was to integrate education into economicanalysis. Education was seen as a process by which theindividual acquired more human capital, that isincreased his future earnings. Since education involvedresources and the student foregoing acquisit;*.n ofhuman capital involves an investment process that isthe foregoing of consumption (or income) now, in orderto generate higher consumption (or income) in thefuture. Economists were able to measure the return onthis investment, which they found to be very large bynormal standards. For instance in New Zealand latersecondary school and university education gives a realreturn of 15 to 25 percent per annum (above the rate ofinflation) to the individual and S to 10 percent perannum to the society.

Another use was in the aid programmes fordeveloping countries where it was shown that goodnutrition and health services for children and adultsimproved their potential to earn income. This suggestedthat it may be just as sensible to invest in human capitalfor foreign aid as it is to invest in physical capital. Thusthe human capital model gives a justification for 'soft'aid programmes, that is, aid programmes directlyassisting humans, as well as 'hard' aid programmes ofmore 'objective' items.

Investing in ChildrenWe can extend the notion further by arguing thatchildrearing is also an investment programme. That iseach society maintains and promotes its income in thefuture by investing in its children today. The investmentbecomes realised as the children grow up to adulthoodand produce. Part of that production is retained by theadults that were the children (as is was by their parents)but part is commanded by society for the use of others.including social welfare provision for the elderly. sick,

66

and disabled. and for further investment in the nextgeneration of children.

This process of social investment in children whichis realised in adulthood. is vital for the working of amodern welfare state. The ability to provide adequatelyfor the retirement of its elderly is dependent upon theproductivity capacity of its working adults. If someyears before, during their working life, the elderly hadinvested wisely in the children who are today's adults,then, they provided for their retirement. Moreover it isnot merely a matter of economic inputs and outputs.The investment activity places a moral obligation ontoday's adults to provide for the elderly who supportedthem in childhood, and a moral example for today'schildren who are tomorrow's adults, to provide fortoday's adults in their retirement.

Another important social consequence of theinvestment in children approach is its transformationof the role of parents. Because parenting is mainly inthe informal part of the economy. (exceptions includechild minding for money and teaching) the parentactivity as an economic activity is easily neglected.However, from the point of view of investment inchildren, parental care is an integral economic activityto be measured by the income parents forego in orderto care for children. By far the larger part of thisforegone income is by mothers. The implication is thatparenting in general, and mothering in particular, canbe treated as an occupational activity. Perhaps it lies, inprofessional standing. between medicine and teachingat one end, and economics and politics at the other.The implications for women's life cycle andemployment patterns are manifold, but would take usoutside this discussion.

It has been possible to make some estimate of thesize of the child investment programme in NewZealand. The total can be divided into broadmagnitudes as follows. The maintenance costs ofchildren in food, clothing, housing, recreation, etc.,comes to around 12 p,ircent of Gross National Product(G.N.P.) i.e. total economic output. Public serviceexpenditure, mainly on education and health, comes toaround a further 5 percent of G.N.P. And the costs ofchild care, calculated as parent's market earningpotential foregone, is conservatively measured as 8percent of G.N.P.

In total, investment in children comes to around aquarter of Gross National Product, or about the samemagnitude as all the other investments combined.However, ordinary investment activity is moreheterogeneous than child investment, so thechildrearing industry is our largest single industry.Moreover it is a very productive industry. It appears thatthe rate of return on investment in children is aboutthree to five times the rate of return in physicalinvestment, although we get such a low return on ourphysical investment the ratio is not as impressive as itmight appear.

What we do not know, is which facets of child rearingare most productive. It would be great if we couldreport that the return on breast feeding was X percent.on child beating was minus Y percent, and onfluoridation of the water supply was Z percent.Unfortunately, we do not know returns on specificactivities. Until we do we must be cautious atadvocating increased expenditure on children in thevague hope that this will increase the productivity ofthe industry.

Traditional capital theory was often illustrated byexamples from the wine industry, and students werefaced with considering the economic return on payingfor the annual priestly blessing on the stored maturingwine. Similarly much of our expenditure on childrenmay have as much effect on the maturing children.

One interesting result from my research is that NewZealand has been investing relatively heavily in humancapital throughout the post war era. Like the winemerchant building up cellar stocks, much of thisinvestment has not yet begun to generate cash income.The implication was that as the children enteredadulthood in the 1980s New Zealand would haveexperienced a marked increase in its growth potential.However, it now seems likely that much of that humaninvestment wilt be dissipated because of unemploymentand outmigration.

Human Investment and Household IncomeHaving identified the size of the human investment thenext step is to examine how it is financed, that is, whoforegoes consumption today in order to create thehuman capital which produces the consumption oftomorrow. While the state. that is, the communitycollectively, contributes about a third of the resourcesrequired. the remainder. amounting to 16 percent ofGNP. is provided by the parents.

Thus the cost of the human investment programmefalls unevenly across the community, with a childlesscouple both of whom work. contributing relatively littleeven though they are on high incomes while the familywith young children and only one adult in employmentcontributing a tot despite being on a lower income.

Three points can be made in terms of the earlieranalysis. First the childless couple will expect its socialsecurity benefits during retirement and disability to bepaid by others' children and for those children tohonour the moral commitment not to expropriate(without compensation) the financial investment thecouple have accumulated for their retirement. Thisimplies that the childless, (and those with few children)have a reciprocal moral duty to support those withchildren (and those with many children).

Second, while the family with young children may be aone-income family, it is not a one-worker family since,although not in formal employment, the mother caring

for young children is a worker contributing to socialoutput.

Third, if the family was producing physical goods thenit could finance the project by borrowing against thesecurity of the income the goods would produce in thefuture. However, since we do not permit slavery, afinancial institution cannot loan against the security ofthe future income of children. Thus a family cannotfinance its human investment programme analogous tothe financing of an ordinary firm.

This is a rather academic way of deriving the obviouspoint that the material standard of living of families withchildren is likely to be below the national average. Databased on the 197314 Household Expenditure Surveysuggested that 63 percent of children and 57 percent ofparents were in households with below median materialstandards of living.

That families are in this situation is not in dispute. Norcould there be dispute, given the preceding analysis.There has been a tittle dispute about the numbers offamilies below the poverty line. Leaving aside the erraticintervention of politicians, at least one of whom hasargued that the problem is a 'poverty of intellect', thusperhaps demonstrating what he was talking about. thedispute has been over the definition and measurementof poverty.

Family Poverty in New ZealandIt is now broadly accepted that there are two sorts ofpoverty. Absolute poverty is when the individual isunable to sustain life and health. It is a phenomenon weassociate with poor countries which may havewidespread disease and inadequate nutrition.

However, within more wealthy countries, it has beenrecognised that while there may be little absolutepoverty. the share some groups get of communityconsumption may be so low that they do not. in thewords of the 1972 Royal Commission on Social Security,'feel a sense of participation in. and belonging to, thecommunity'. This is clearly a crucial problem, given themoral underpinning of the economics of children thatwe have referred to earlier. Such people are in relativepoverty. and the Royal Commission on Social Securityargued 'that there is a need to ensure within limitationswhich may be imposed by physical or other disabilitiesthat everyone is able to enjoy a standard of living muchlike that of the rest of the community'.

Thus there are two poverty principles and byimplication two poverty lines. Following traditionalanalysis we might call the absolute poverty line 'thebreadline', in which case the relative poverty tine mightbe called 'the meat line' since it is frequently associatedwith an inability to afford meat regularly at meals.

In practice the meatline is set as the material standardof living Oa couple on the basic invalid social securitybenefit. The argument is that it is unlikely that the

benefit level is meant to be below the poverty line, and itis also unlikely that we mean it to be substantially aboveeither. Moreover, there has been a bit of survey researchto suggest that households below this income level dohave difficulty maintaining an acceptable life style muchlike the rest of the community, such as regular meatmeals.

If we adopt the social security level as the poverty line,then the evidence points to around a quarter of childrenand a fifth of parents are in households below thisstandard of living. This is by tar the largest group of poorin the community. Moreover the bulk of poor familiesare not solo parent families and others living on socialsecurity benefits. The vast majority are primarilydependent upon the wage their father can earn in themarket.

The research has gone a stage further in an attempt toelaborate a description of what determines householdinc- -ne, and therefore, household poverty. Quite a lot isknown about various aspects of the mechanicsincluding father's incomes. mother's workforceparticipation. the family life cycle, the impact Of taxationand benefit, and some of the factors about family size.However. there have not yet been funds available tobuild up a model which integrates these facets togetherinto a comprehensive description of householdeconomic behaviour.

The Consequences of Family PovertyWe know lessabout the consequences of this poverty.In principle there are three major issues. First itsexistence undermines the implicit morality of thewelfare state as we discussed earlier. If one groupserving a worthwhile function in society is poorly off.then other groups which benefit from social support arelikely to be treated as illiberally.

Second is the question of family morale. No matterhow worthwhile a social function families may becarrying out, if they are not socially rewarded theirpsychological state may be weakened. The rewardsneed to be two-fold: income which keeps their standardof living much like the rest of the community. andgenuine public respect for the role of childrearing. Itseems likely that low morale can have real effects onchildrearing performance. The tired, slightly hungry,slightly sick or tense parent is unlikely to manage thechildren well during the difficult developmental periods.The same phenomena can lead to marital tensions andeven marriage breakdown when other difficultiesalready exist.

On the whole, family morale seems to be much betterthan we might expect from the treatment we givefamilies. This is partly because its public image issuppressed, particularly in order to protect children. Asone headmaster remarked:

If there is a struggle in the home situation you mightnot see it reflected in the children... there are a lot of

68

parents who have their priorities in the right place.The children are the last to suffer.

On the other hand the housewives' boycott outburst of1977 almost certainly reflected a pathetic response ofdemoralised mothers to the economic pressures onthem.

The third consequence is the direct effect of povertyon childrearing. The previous two paragraphs indicatethat deterioration in parental morale can result in lesseffective child rearing. It is also evident that poverty canresult in the child's nutritional standards. comfort, orhealth being sufficiently below the acceptable level toresult in poor performance, at school for instance.Moreover, an overcrowded, dangerous. ornonstimulating environment for a child may result inpoor development and even injury. A final direct effectwould be where poverty leads to lower opportunity forthe children. There is evidence this applies to bothhealth care and education.

On the whole there is very little systematic evidence tosupport suspicions that family poverty is causingmarkedly inferiour child development in New Zealand.But this is a conclusion of agnostism rather thanatheism. It is an argument for minimising family povertyuntil we have evidence that poverty does not have theseeffects. rather than leaving family incomes alone untilthe research evidence is collected (knowing full well thatthere would be no urgency to do the research).Moreover whatever the conclusion of the research. theeffect of the existence of family poverty undermining thewelfare state cannot be ignored.

Recent Trends in Family IncomeMaintenanceThe evidence for family poverty was becoming apparentby the mid 1970s. It involved the salutory lesson that thepast solution for maintaining family incomes had failedto eliminate poverty. This approach. which appears fromthe beginning of the welfare state in the 1890s. was toexpect the wage system to provide adequate support forintact families without disability. (There was specialprovision for other sorts of families, commencing withWorkers Compensation in 1900.) In 1914 a small incometax concession for children was introduced. In 1927there was the beginnings of the family benefit. Theperiod of the first Labour Government saw thebeginnings of systematic assistance with health andhousing. ('Free' education began by 1877.)

The approach failed partly because it lackedcoherence. but perhaps more fundamentally becauseany family policy which is based on the labour-earningsstructure is bound to fail since the worker's earningcapacity has little relation to his family's needs.

Since family poverty became evident the approachhas remained ad hoc. but more attention has been paidto modifying the income tax structure in favour ofchildren.

In 1976 the young family tax rebate was introduced forfamilies with a child below the age of 5. Currently it is$9,00 per week but abateable after the first $150 perweek of income of the principle earner (father).

In 1977 the single income family tax rebate for familieswith a child below the age of 12 was introduced.Currently it is $5.00 per week, but abateable after thefirst $25 per week of the income of the secondary earner(mother).

In 1978 the tax scales were changed to levy more taxon the secondary earner, and hence lower tax rates onthe principle earner. This change benefited one incomefamilies at the expense of one and a bit income families,

These changes were reinforced from October 1979,and the family benefit was doubled, It seems likely thatfrom October 1979 the state will be contributing more tothe financial support. measured in real terms, of intactfamilies without disability, than at any time in the last 20years and perhaps longer.

Nevertheless. such ad hocery is not likely to succeedunless it is very expensive, Some comprehensiveapproach to income maintenance for families isnecessary, including the institution of tax credits (ornegative income tax) and the integration of subsidies forchild care into the system, It should also be possible toinclude non standard family situations such as soloparents and disability into the system.

The task of constructing a comprehensive incomemaintenance scheme is not an easy one. Besidesempirical evidence it requires a sound analysis. of whichthe economics of children is likely to be part. and asocial philosophy,

The Family as a Production UnitAnother policy area which has developed out of theeconomics of children arises from the closer attentiongiven to the family as a production unit. If child rearing isan industry, perhaps our largest and most importantindustry, then the families are its firms. It is a natural stepto considering how we may increase each firm'sproductivity.

The effect then is to shift attention from the childrearing activities which occur in the formal part of theeconomy, such as in hospitals, schools and children'shomes, and are supervised by paid workers such asmedics, teachers, and social workers, to the childrearingwhich occurs in the informal part of the economy ofhouseholds supervised by unpaid parents.

Once the point is made it is evident that by far themost childrearing (and the most important childrearing)occurs in this informal area. Yet the most resources andthe most attention isgiven to the formal areas,For instancein a total public education budget of one billion dollars,the actual amounts spent on the education parents giveto their children, including parent education. is less than0.1 percent of the total. The implication is that we can

substantially improve childrearing by greaterconcentration on the family, not because it is moreefficient per se. but because we are doing so little thatwe must be able to do better at little cost.

Economics has little to say on what measures shouldbe taken, except to reiterate the importance of anadequate economic environment in which the familyshould operate. It is for other disciplines to identify othermeasures. Despite our lack of research of the intricaciesinvolved in family behaviour, there is sufficient evidenceof major differences between optimal behaviour andpractice in a number of areas to suggest effectivemeasures based upon improving the public'sunderstanding of parental responsibilities and activities.

For instance. the children of young mothers are atgreater risk, Yet there is no public educationprogramme among adolescents to discourage teenagebirths. We have become so obsessed with the physicaldetails of contraception, that we neglect thepsychological and social aspects of family planning,Spacing between children is another area where publicunderstanding could be usefully brought more into linewith scientific knowledge.

There are also vast gaps between parentalunderstanding of children's health. nutrition andbehaviour, and the range of practices consideredoptimal by expert opinion. This is not the result ofdeliberate ignorance on the part of most parents whoare probably keener to learn than the experts are toteach. The problem is the creation of adequately fundedinstitutional means to enable parents to obtain theeducation they seek. To do this may well involve arevolution in our approach in the formal economy.Perhaps schools should be offering courses to parents,on fully paid bursaries, and leaving more of the teachingof children to the parents.

One way of summarising this approach is therecognition of the 'pre-primary care' area for childdevelopment. Our tendency has been to think of the firstinstitution, be it school, medical centre or social worker,as the point where care begins. Put this way the view isabsurd. but an astonishing level of advocation andpolicy implicitly assumes this acknowledged absurdview. For instance it is proper to be concerned withthose activities labelled as 'preventative', but, that labelreflects a view from the institution/professional world.rather than the family.

One of the unfortunate aspects of the welfare statehas been to shift power, responsibility, and resourcesaway from the pre-primary care area. that is the family.to the institutions and professions. A reversal of the shiftis required both for human liberty and for economicefficiency.

It is very easy to prepare authoritarian measures tomeet some problem area in child development, but theauthoritarians might do well to reflect on the recentlyreported treatment of children in state social welfare

69

homes. Obviously some family units are far from ideal.Apparently the alternative may be little better or evenworse. And the fact there are a few failures does notjustify unnecessary impositions on all failures (whetherwe are talking of families, doctors, teachers, or socialworkers).

DirectionsOne interesting feature of the economics of children isits tendency to analyse children in typical, althoughdiverse circumstances. rather than to concentrate onatypical circumstances such as solo parent families. exnuptial births, child abuse, or parental and childdelinquincy. Research on pathological situations isnecessary, but it can result in an over preoccupationwith such situation at the expense of the typicalsituation. Too often new social policy in New Zealand is,proposed by psychiatrists. social workers andpoliticians responding to the pathological cases theymeet. The result can be the tail wagging the dog.

It may be that the economic approach is going to theother extreme, subverting the richness and complexityof the psychological, social and economic phenomenawe call family life into a simple economic model.Obviously there is such a danger.

Much policy making. particularly in the social andindustrial areas, effects behaviour within the informaleconomy. To neglect such behaviour can easily result inthe policy prescriptions being inefficient, ineffective.and even having the opposite effect to that intended.

Economists need a theory of the informal economy. Ifsuch a theory is also beneficial for other groups involvedin the informal economy or in developing policy therein.then we have an example of the advantages ofco-operative enterprise.

NotesThe following is a list of publications which generallyenlarge upon the points made in the text. Some of thearticles listed are available in mimeograph from theauthor, but it may be necessary to make a charge. Other

people's research used in the text is identified in theseprimary sources.

The Needs Index (mimeo, 1973)Family Incomes (mimeo. 1973)Poverty in New Zealand (mimeo. 1973)"Staying on At School" NZ Monthly Review (Nov. 1973)-The Welfare State: Principles and Future". New Zealand

Politics:A Reader ed S. Levine (Cheshire. 1975)-The New Zealand Housing Market" New Zealand Economic

Papers, 1976"Poverty in New Zealand: Estimates and Reflections"

Political Science (Aug. 1976)-The Economic Lifecycle of the Modern New Zealand

Family" Australian and N.Z. Journal of Sociology (Feb.1977)

-income Tax on Employee Earnings" Quarterly Predictions(March 1978)

-The Economics of Children" impact: The Burden of Choice(Akld Principles Association. 1978)

"The Married Women's Properly Act and the DomesticPurposes Benefit" Woman Studies Association Newsletter(June 1978)

"Income and Fertility in New Zealand: A Cross-SectionalStudy" The Proceedings of the Annual Conference of theNZ Demographic Society 1978

"The Rise of Family Policy" (mimeo. 1978)-The Economic Needs of Young Children" Parents Centre

Bulletin (Spring 1978)"Population and the Economy" New Zealand Population and

C.J. O'Neill and W. Neville (Longrnans Paul 1979)-The Family Costs of Children" The Proceedings of the

Annual Conference of the NZ Demographic Society 1979-The Economics and Politics of New Zealand Poverty"

Poverty in Godzone? ed G.W. Bryant (1979)"New Directions in Health and Economic Management" N.Z.

Medical Journal (forthcoming)-Household Equivalence Scales" New Zealand Statistician

(forthcoming)Social Security in the Seventies: Policy Formation and

Failure (University of Canterbury. forthcoming)Social Policy Analysis in a Welfare State (Allen and Un win,

forthcoming)"Some Household Expenditure Functions" (with C.M. Low

and S.F. Connor. forthcoming).

Brian Easton is a lecturer in Economics at the University ofCanterbury. Thisarticle is taken from a talk given to the EarlyChildhood Care and Development Convention, Christchurch.1979.

70

Item 3(NZ)

number one 1980

Teacher Expectations andClassroom Behaviour

By Alison St. GeorgeMassey University

Impressions of Other PeoplePeople are always making judgements about people. Itis hard to imagine how we would getalong with otherpeople without our perceptions of them as a guide. Theideas we have about others influence the way webehave toward them and the way they respond to us.There are several important points here:

a. We are active perceivers, selecting, organizing, andstructuring what we notice about people.

b. The situation and who we are as well ascharacteristics of the person affect what we see.

c. We make inferences and attribute characteristics tothe person on the basis of immediate cues. Theattributes may concern temporary states (e.g. he isbored) or more permanent characteristics (e.g. she isintelligent). Many of these characteristics cannot beobserved directly.

d. We also respond in an emotional way, liking, beinginterested in, having sympathy for, rejecting, and soon.

e. Expectations are generated when we perceive otherpeople and these are particularly important inface-to-face situations because we need to preparefor responding.

Teachers and ExpectationsAny school-learning situation is essentially one ofsocial interaction: teachers and pupils form impressionsof each other, react emotionally, and anticipateattitudes and behaviour. judgements and expectationsguide behaviour, though those involved may notalways be aware of it. Further, we do not alwayscarefully select and weigh up evidence. Often we makehaphazard judgements.

For teachers, forming expectations about their pupilsis an integral part of their task. It is necessary in order tobe able to plan work for pupils, to assess their progress,and to relate to them. Teachers are also required tomake deliberate judgements and write them down forschool records, for parents' information, or testimonialsfor employers' guidance. Thus the whole business ofperceiving and judging other people has particularrelevance for teachers and often has long lastingconsequences for their pupils.

Teachers have expectations about their pupils'intellectual potential and academic achievement.However, they also face the task of social control, andtheir expectations about general classroom behaviourare also important. Expectations should not be confusedwith hopes, desires, and aspirations. What a teacherhopes a new class is like, and what he or she expects itwill be like, are often not the same thing!

Expectations as Self-Fulfilling PropheciesIn the late 1960s it was suggested that a teacher'sexpectations fora pupil could be self-fulfilling.Pygmalion in the Classroom, published in 1968, causedconsiderable controversy. In it ROsenthal and Jacobsondescribed a study iii which they attempted tomanipulate teachereexpectations to see if thoseexpectations would be fulfilled. The findings were

interpreted as demonstrating that favourableexpectations could lead tqtains in pupil achievementand general ability. By impfication, the study alsoseemed to support claims made by critics of the schoolsthat the lower academic achievement of poor andminority group pupils was largely the result of thenegative stereotypes and low expectations of theirteachers.

Popular reports of the research made enthusiasticclaims which were often exaggerated and the notion ofteacher expectations acquired a mysterious, almostmagical, quality as if expectations always becameself-fulfilling. A teacher only had to make a predictionand it would come true. Not surprisingly, manyteachers did not accept this. The study was also stronglycriticized in academic circles for problems with theexperimental design and methodology, the data, theway it was reported, and the overgeneralization of theresults.

Since that controversial piece of research a largenumber of investigators have conducted studies ofteacher expectations using a variety of methods. Insome studies attempts have been made to manipulateexpectations, for example, by providing false testinformation or fictional reports. In others, teachers ownnaturally formed expectations have been used. Someinvestigators have looked at teacher expectations andmeasures of pupil performance or self concept, whileothers have been more concerned with the way inwhich expectations may be communicated throughinteractions between teachers and pupils. The evidencenow available supports the idea that teacherexpectations can sometimes function as self-fulfillingprophecies but that this is by no means automatic orinevitable. For an expectation to be self-fulfilling it isnot enough that it just exists. The person concernedmust recognise what the expectation is and theteacher's behaviourmust also be effective in movingthe person in the expected direction.

A New Zealand StudyA recent study in New Zealand has investigated theperceptions and expectations of five Standard Threeteachers and their pupils. No attempt was made tomanipulate the teachers' expectations about theirpupils. Rather, the teachers own expectations werestudied. Early in the year they were required to ranktheir pupils' general ability. Using these rankings ahigh, middle, and low expectation group was selectedfrom each class, 90 children in all (67 Pakeha, 20 Maori,3 Pacific Islands), with equal numbers in each class andexpectation group and equal numbers of boys and girls.

Perceptions and ExpectationsThe teachers were later asked to give further much moredetailed perceptions of each child. Children for whomthe teacher had high general expectations were seen aswell off in:

readingreaction to new workindependence (need for teacher guidance)task concentrationuse of English in conversationhome environmentinterest in the world aroundperseverance 72

participationconfidenceparent attitudes to school and education.

Not surprisingly pupils for whom the teachers heldhigher expectations were closer to the notion of an'ideal' pupil (as identified in the research literature)than those in the middle and low expectation groupsThey were perceived as having academic work skillsand an eagerness to learn which makes it easier for theteacher to feel some success at teaching. They alsotended to be seen as coming from a 'good' homebackground.

Teacher perceptions also varied with pupil ethnicgroup. Generally Pakehas were perceived as beingcloser to the 'ideal' pupil concept than the Maori andPacific Island children.

Specifically, the teachers associated poor parentattitudes to school and education, less interest in theworld around, a poorer use of conversational English,and a less stimulating home environment with Maoriand Pacific Island pupils. Poorer reading skills, lowerparticipation in class activities, a more negativereaction to new work, and a need for greater teacherguidance were also part of teachers' views of thesechildren. To a lesser extent, the Polynesian pupils werealso perceived as having a lower level of perseverance atand concentration on school work. The teachersexpected less of the Maori and Pacific Island children,ranking them lower in general ability. Most were in thelow expectation group.

The findings for teacher perceptions and expectationsthen do not look particularly encouraging as far asminority group children are concerned. it has beensuggested by some that stereotyped expectations forMaori pupils may become self-fulfilling prophecies.Are teachers in New Zealand clearly prejudiced and as aconsequence do they engage in discriminatorybehaviour toward Maori and Pacific Island children inthe classroom?

Expectations and Classroom BehaviourHow did the teachers behave toward their pupils? Inorder to study this we carried out classroomobservations. Teacher and pupil behaviour wasrecorded whenever the teacher was interacting with apupil but not when the interchange was with a group asa whole. Some of the situations were public, where thewhole class was looking on. Others were private, forexample, they were about a piece of work, behaviour,or classroom procedure concerning a particular pupiland not meant for the whole class.

Behaviour and ethnic groupThere was no statistically significant differencebetween the way teachers interacted with Polynesianand Pakeha children. In fact there seemed to be anactive effort on the part of teachers to give all pupilssimilar amounts of attention. While there may be a fewteachers who actively discriminate against minoritygroup children there is little evidence that this isgenerally the case. Certainly it was not seen in the studybeing described here.

Some interesting differences in classroom behaviourdid emerge though when we looked at it in relation towhat teachers expected of their pupils.

Public interactionsWhen children could respond in public there was nosignificant difference in the number of opportun itiesgiven to pupils in the high. middle, and lowexpectation groups. However, there was a difference inthe type and complexity of questions asked. Sometimesteachers posed a question to a group and waited forhands to be raised before selecting a child to answer. Inother instances teachers addressed a question directlyto a particutarchild. When these two types of questionswere compared it was found that high expectationpupils received most of the first 'open' type. while thelow expectation pupils were asked most of the 'direct'questions.

The finding that teachers asked the low expectationgroup to answer fewer questions when they were posedto the class (and to which they expressed somewillingness to respond) may be the result of theteachers' experience of the pupils' differences inknowledge and ability. But this trend is balanced byteachers addressing more questions directly to lowexpectation group pupils and selecting them to respondeven when they have expressed no willingness to do so.The questions posed in this way were more likely to berelatively simple, requiring a short answer, rather thanmore complex process questions involving goingthrough steps to solve a problem or explain aphenomenon. It may be that teachers have differentreasons for asking public academic questions ofchildren from the different expectation groups and thatthese reasons influence the complexity of the questionasked. For example, in some cases the aim might be tosee if the pupil understands the material, in others itmight be to encourage participation.

What are teachers' reactions to pupils' publicacademic responses? Our findings were limited by ourdata. It was clear, however, that all teachers criticisedpupils to a greater degree when they did not attempt aresponse than when they gave an answer which waswrong.

It ntay be that teachers view academic failure as morereprehensible when it is accompanied by an apparentlack of effort. However, frequent criticism for notresponding does not seem likely to encourage pupils torespond in the future unless strategies for reachingsolutions are offered and encouragement given. Theinformational value of criticism may be offset by thelower perceived warmth of the teacher. Further, theteachers as a group criticized the pupils for tvhom theyheld high expectations a lot less than either middle orlow expectation group pupils both in relation toacademic work and to general classroom behaviour.

Private interactionsAs with poblic responses, there was no significantdifference between expectation groups in the overallnumber of private interactions they had with theirteachers. However, a clear pattern did emerge when weasked, 'Who sought the interaction?' Children in thehigh expectation group created more private contactswith the teacher for themselves, whereas the teacherinitiated the greatest numberoi private interactionswith tow expectation group pupils. Once again, therewas an attempt by the teacher to compensate for the lowrate of initiating interactions in the low expectationgroup and thus to reduce any inequalities in theamount of teacher-pupil interaction. This can be further

understood by referring back to teacher perceptions ofpupil attributes. in public response situations it mayrepresent an active attempt to get the children toparticipate and to compensate for their lack ofconfidence. In private contacts it is in line with theperception of children in the low expectation group aslacking perseverance and task concentration, reactingpoorly to new work and needing teacherguidance.

Expectations and AchievementThe teachers did not have the NZCER ProgressiveAchievement Tests (Reading Comprehension, ReadingVocabulary, and Listening Comprehension) resultsavailable when the teacher rankings were collected.However, teachers' expectations for their pupils basedon these rankings were strongly related to results on thePATs and this provides evidence that the teachers'expectations were relatively accurate.

Further, expectations were also strongly related to anend of the year composite measure of schoolachievement based on teacher ratings. Does this meanthat a self-fulfilling prophecy was operating? Notnecessarily, unless there is clear evidence that thoseexpectations were consistently communicated to pupilsin a way that shaped their behaviour and achievement.Here no clear-cut evidence of this nature is available.

What of ethnic group differences? The teacherspercehied Maori and Pacific Island pupils in lessfavourable terms than Pakeha pupils and had lowerexpectations for their general ability to do well atschool. Nevertheless, there were no statisticallysignificant differences in teacher-pupil interactionalong ethnic group lines. Thus the teachers expectedless of the Maoris and Pacific Islanders than the law ofaverages suggests they should have, but they treatedthem similarly to others in their expectation groups.

As an aside we can ask, does this similar treatmentmean that the Polynesian and Pakeha children achievedat similar levels? Unfortunately, no. Overall, Pakehachildren scored higher on the PAT tests and their end ofyear school achievement was rated higher by theirteachers. While the greatest differences in levels ofachievement lay between expectation groups, withinthese groups Polynesian pupils tended to score slightlylower. This difference may be due to somedimension(s) of behaviour tapped by thePolynesian/Pakeha dichotomy used in this study butnot reflected in the teacher expectation measure used.Cultural differences in communication, motivation,and self concept, or preferences for different kinds oflearning environments could be involved.

Implications for TeachersThe study described here was essentially exploratory. itwas naturalistic rather than experimental and does notencourage definite casual inferences. However, it doesprovide some important data on the nature of teacherexpectations in New Zealand ethnically mixedclassrooms and the relationships between teacherperceptions, expectations, and interactions with pupils.

It is encouraging to find no evidence that teacherstreat children for whom they have lots expectations in agenerally negative way. Despite the more negative andstereotyped perceptions of Maori and Pacific Islandpupils as a group, teachers seemed to put into practice aphilosophy of equitable and fair treatment of all pupils.

"/...13

(It would be a pity, however, if in the effort to providefair and equal treatment that teachers are unable todraw upon and react to positive aspects of ethnic groupdifferences. A greater similarity in level of interests,attempts to learn, and a positive self-regard in allchildren might result.)

Being aware of the possible effects of theirexpectations can help teachers reduce the likelihood ofundesirable self-fulfilling prophecies occurring. It is notappropriate, however, for teachers to try to avoidforming expectations altogether. Even if teachers do notlook at achievement records or test information andthey avoid discussing pupils with previous teachersthey still tend to form strong and general impressionsquickly, through interaction in the classroom.

It is also sometimes suggested that teachers shouldonly have positive expectations. This is unrealisticwhen carried to the extreme wishful thinking willnot get rid of all differences in pupil abilities andinterests. More important than having highexpectations all the time is that expectations should beappropriate and followed up with appropriatebehaviour. There is nothing wrong with expecting thata student will have difficulty if help is provided toremedy the difficulty. Such help should be positive,providing the pupil with courses of action, and the wayit is given should avoid communicating negativeexpectations.

Teachers should continually check their expectationsfor pupils and adjust them where necessary. It is notalways easy. When expectations become strong andstable there is a tendency to notice only those aspects ofa person's behaviour that 'fit' the expectation. Inaddition, the way we interpret what we see may beaffected. For instance the behaviour of two childrenmight be exactly the same but the teacher perceives the'bright' one as concentrating on working out theproblem and the 'slow' one as being hopelessly lost.

When teaching children from a variety of ethnic andcultural backgrounds it should be remembered thatcultures are systems of shared understandings,especially about what words and actions mean and howimportant values should be expressed. Unintentionalmiscommunication and misunderstanding may occurin the classroom if teachers do not know the importantdifferences that exist.

Where does the information we use to judge thelikely success or failure of our pupils come from? Oftenfrom observation of academic skills and work habits inthe classroom. Consider what will happen wherestereotyped judgements are adopted on limitedinformation with little attention being paid to how achild is actually behaving. Focussing on a label mayswitch attention away from teaching and fosteringprogress. When poor academic skills can be 'explained'by saying thata child is a slow learner or is a member ofa particular ethnic group then there is little room for theteacher, What effect can he or she have?

ConclusionWhat does this study have to say to New Zealandteachers, particularly those in multi-culturalclassrooms? While it is not representative of .:11 thoseteachers, we think it does provide some indications ofwhat is going on and what to look out for.

Some of the study's findings are heartening, othersless so. It is unfortunate that teachers still hold negativestereotypes about Maori and Pacific Island pupils and

are more likely to expect them to be of lower ability andto do less well in school than Pakeh a pupils. Further,despite the fact that Pakeh a and Polynesian pupils havesimilar kinds of classroom interaction with theirteachers they do not achieve at the same level.

On the positive side it is encouraging to find thatovert discrimination against minority group childrendoes not seem to be the general case. It is also good tosee that teachers are attempting to even up the numberof interaction opportunities available to pupils. This isnot to say that all children experience the sameclassroom environment, for the type and quality ofinteraction is also very important, but it does seem toindicate that the teachers are trying to provide for alltheir pupils.

What are the things to watch out for? Clearly thosewho teach children from a variety of ethnic and culturalbackgrounds should try to avoid making premature,inflexible judgements on little information. Judgingchildren on the basis of a single characteristic such asethnic group will result in many inappropriateexpectations being formed. It is all too easy to teach onthe basis of untested assumptions about children.Check them out and remain open to noticing what yourpupils are really doing. Otherwise you may only seewhat you expect to see.

Notes:

Impressions of Other PeopleDiscussions can be found in:Downey, M. interpersonal judgements in education.

London:Harper and Row, 1977.Hargreaves, D.H. Interpersomd relations in education. London:

Routledge and Kegan Paul, 1972.

Teachers and ExpectationsDiscussions and relevant research are contained in:Brown, C. Teacher expectations: SociopsychologIcal dynamics.

&Tient of Educational Research, 1976, 46, 185-213.Brophy, I.E., and Good, T.L. Teacher-student relationships: Causes

and cortsequettces, New York: Holt, Rinehart and Winston,1974.

Dusek, J.B. Do teachers bias children's learning? Review ofEducational Research. 1975, 45,661-684.

Good, T.L., and Brophy, J.E. Looking in classrooms (2nd ed.). NewYork: Harper and Row, 1978.

Nash, R. Teacher expectations and pupil learning. London:Routledge and Kegan Paul, 1976.

Expectations as Self-Fulfilling PropheciesThe controversial study by Rosenthal and Jacobson is in:Rosenthal, R., and Jacobson, L. Pygmalion in tire classroom, New

York: Holt, Rinehart and Winston, 1968.

A New Zealand StudySt. George, A. Perceptions, expectations and interactions: A study of

leachers and pupils in five ethnically mixed primary classrooms.D. Phil. thesis, University of Waikato, 1978.

Implications for TeachersFurtherhelpful suggestions are in Downey, Good and Brophymentioned above.

Discussions of problems of cross-cultural communication withillustrative examples relevant to teacher expectations and theschool setting are found in:McDonald, C. Working with Pacific Island and Maori parents.

Set, 1979, No. 1, Item 4.

Merge, J., and Kinloch, P. Talking past each other: Prolsktus ofcross-cultural communicatiott. Wellington: Victoria UniversityPressiPrice Milburn, 1978.