Post on 20-Feb-2018
7/24/2019 Relations Between Teacher and Classroom Activity Variables
1/30
JOURNAL OF APPLIED DEVELOPMENTAL PSYCHOLOGY 16,253-282(1995)
Relations Between Teacher and
Classroom Activity Variables and the
Classroom Behaviors of
Prekindergarten Children in
Chapter 1 Funded Programs
MARTHA B BRONSON
Boston College
TERRENCE IVNAN
Harvard University
PATRICIAS SEPPANEN
RMC Research Corporation
This study of 586 children in 54 Chapter 1 funded classrooms in five geographical
areas wos part of o national observational study of early childhood programs. The
study examined relations between eight teacher and classroom activity variables and
14 child variables. Correlational analyses indicated significant relations between
program and teacher variables and childrens experiences in the classroom. More
program ond teacher variables were associated with childrens mastery than social
experiences. Results indicated some benefits ond trade offs when more teacher and
classroom time wos spent on teaching and on cognitively focused activities.
A growing body of research has demonstrated that children from low-income or
high-risk environments benefit from participation in high-quality early childhood
programs and that these benefits translate into reduced risk of school failure and
economic benefits to society in terms of lower school expenditures for special
education, savings in the criminal justice system, and greater economic indepen-
dence through employment and earnings (Balasubramaniam & Tumbull, 1988;
Bamett & Escobar, 1987; Berrueta-Clement, Schweinhart, Barnett, Epstein, &
Weikart, 1984; Lazar, Darlington, Murray, Royce, & Snipper, 1982; McKey et
al., 1985; Weikart, 1989). Research has also suggested that differences in pro-
gram philosophy and curriculum have differential effects on childrens behavior
Support for this study was provided by the U.S. Department of Education, Offke of Policy and
Planning under Contract No. EALC890980.
Correspondence and requests for reprints should be sent to Martha B. Bronson, School of
Education, Boston College, 201B Champion Hall, Chestnut Hill, MA 02167-3813.
253
7/24/2019 Relations Between Teacher and Classroom Activity Variables
2/30
254 BRONSON, TIVNAN, AND SEPPANEN
and development (Clark-Stewart, 1980; Prescott, 1973; Snow, 1983; Weikart,
198 l), but relatively little is known about how specific aspects of programs affect
children. Researchers have lamented their inability to connect specific program
elements with outcomes (Barnett, Frede, Mobasher, & Mohr, 1987; Powell,
1987; Sigel, 1990) and to disentangle the effects of curriculum content, specific
activities, and teaching techniques (Miller, Bugbee, & Hybertson, 1985). This
study examines relations between a set of observed teacher and classroom activ-
ity variables, and child characteristics, including observed social and mastery
behaviors in the classroom. Teacher and classroom activity variables with direct
policy implications were selected for analysis.
The study was part of the Observational Study of Early Childhood Programs,
a national study of programs for disadvantaged children which had two related
components: (a) a descriptive study of 119 early childhood classroom environ-
ments, including Head Start programs, school-sponsored programs, and day-care
centers (Layzer, Goodson, & Moss, 1993) and (b) a substudy of approximately
750 children enrolled in 55 prekindergarten classrooms funded all or in part
through Chapter 1 of the Hawkins-Stafford Elementary and Secondary School
Improvement Amendments of 1988 (Seppanen, Godin, Metzger, Bronson, &
Cichon, 1993). The overall intent of the national study was to provide
observation-based empirical data on programs for young children.
This study was part of the Chapter 1 substudy. Chapter 1 of Title 1 of the
Elementary and Secondary Education Act of 19652 is one of the oldest, most
well-established federal education laws intended to provide supplemental in-
struction to low-achieving students in schools with low-income students. A
small, but growing segment of the student population at the prekindergarten level
is served through Chapter 1 funded programs. The increasing role of Chapter 1 in
early childhood education has raised a number of issues for state and local
education officials. Among these has been how best to create programs with
characteristics that promote basic and advanced skills for young children at
risk for school failure.
To remedy the lack of connectedness between the long-term effectiveness of
programs and specific program variables, there is growing interest in trying to
determine how these effects occur by examining childrens experiences in these
programs (Clarke-Stewart & Fein, 1983; Powell, 1987, Takanishi, 1979). A
major goal of the Chapter 1 substudy was to provide empirical information about
relations between Chapter 1 funded prekindergarten environments and childrens
cognitive and social-emotional development, as revealed in a variety of
observation-based measures. This study focuses on describing the relations be-
U.S. Department of Education, Off& of Policy and Planning, Contract Number EALC890980.
ZReauthorized by Congress as part of the Hawkins-Stafford Elementary and Secondary School
Improvement Amendments of 1988.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
3/30
OBSERVATIONAL STUDY OF CHAPTER 1 PREKINDERGARTEN PROGRAMS
255
tween specific child-level and classroom-level observation data gathered in the
Chapter 1 component of the Observational Study of Early Childhood Programs.
METHOD
Participants
The sample consisted of 586 children in 54 classrooms3 distributed over 16
school districts in five states (Florida, Texas, California, Michigan, and New
Jersey). The sample of classrooms was considered to be representative of the
range of quality in Chapter 1 funded prekindergarten programs in urban and
suburban settings.
Demographic characteristics of the sample of children4 are listed in Table 1.
Slightly more than half of the children were girls. Eighty-five percent were from
racial or ethnic minority groups, but teachers reported that English was the
primary language spoken at home by over three fourths of the mothers (or
primary care givers). Seventy-eight percent of the sample were eligible for free or
reduced lunch.
Thirty-five percent of the mothers of sample children and 27 of the fathers
had at least some college experience. Approximately 22 of the mothers and
fathers had completed less than 12th grade. About 30 of the mothers were
under 19 years of age when they gave birth to their first child. Only about half of
the parents were currently married.
Teachers reported that approximately 14 of the families of sample children
needed assistance due to substance abuse, economic problems, or behavioral,
emotional, or mental problems. Teachers also indicated that 12 of the sample
children needed assistance for a behavioral, emotional, or mental problem and
that 9 had resided with someone other than their natural mother for more than 1
month since school had started. Only small percentages of sample children were
reported by teachers to have physical, emotional, or mental conditions that
required frequent medical attention and treatment or that limited and prevented
regular school attendance (6 ) or participation in all school activities (5 ).
Classrooms
Staff in the Chapter 1 funded prekindergarten classrooms were relatively highly
educated. All teachers had at least a 4-year college degree and 42 had a
masters degree or higher. Forty-five percent of the assistants and aides had some
3The total number of children in the Chapter 1 substudy was 783 in 55 classrooms, but not all of
these children had complete sets of test scores and classroom observations.
4These characteristics are not necessarily representative of all Chapter I preschool children for
two reasons: (a) the Chapter 1 substudy did not randomly sample Chapter 1 prekindergarten programs
and (h) children were eligible for participation in the Chapter 1 substudy if they were simply enrolled
in a class funded all or in part with Chapter 1 funds.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
4/30
256
BRONSON, TIVNAN, AND SEPPANEN
TABLE 1
Demographic Characteristics of Chapter 1 Samplea
Background Characteristics
Gender
Female
Ethnicity
African-American
Hispanic
White
Other
Primary language Spoken at Home
Mostly English
Mostly Spanish
Mixed English and Spanish
Other
Eligibility for Free or Reduced-Price
Lunch
Eligible
Data not available
Education of Father
Grade 11 or less
Grade 12 or GED
At least some college
Unknown
Education of Mother
Grade 11 or less
Grade 12 or GED
At least some college
Unknown
Mothers Age at Birth of First Child
13-15
16-18
19-21
22 or older
Marital Status
Married
Evidence Family Needed Help Due
to Substance Abuse, Economic
Problems, Emotional, or Mental
Problems
of Sample
52
48
31
15
7
77
8
10
5
78
14
22
44
27
7
22
41
35
1
5
24
34
35
50
14
Evidence Child Needed Help Due to
Behavioral, Emotional, or Mental
Problem
12
(continued)
7/24/2019 Relations Between Teacher and Classroom Activity Variables
5/30
OBSERVATIONAL STUDY OF CHAPTER
1
PREKINDERGARTEN PROGRAMS
257
TABLE 1 (Continued)
Background Characteristics
of Sample
Child Resides Apart
From Mother
(> 1 Month)
9
Evidence of Conditions That
Limit/Prevent
Regular school attandance
6
Participation in all classroom activities 5
Usual childhood activities
4
Evidence of Conditions That Require
Frequent medical attention or treatment
6
Regular medicine or drugs 2
Use of special equipment
1
aParent and teacher questionnaires and review of student
records.
college, 18 had a bachelors degree, and only 8 had less than a high-school
diploma. Seventy-one percent of the teachers held a state early childhood certifi-
cate or license and 87 held a state elementary education certificate or license.
Eighteen percent of the assistants and aides held a state early childhood education
certificate or license and 6 held an elementary license.
Teachers had taught in early childhood for an average of 9.6 years, assistants
and aides had 7.8 years of experience, and both had taught about two thirds of
this time at their current programs. Both staff groups had an additional few years
of grade-school teaching experience; teachers averaged 3.4 years and assistants
and aides averaged 2.8 years.
The average enrollment across all classrooms was 19.1 (SD = 4.3) but the
average attendance was 16.0 SD = 3.2). Almost all classrooms had two staff
members (one teacher and one assistant or aide) so the observed child:staff ratio
children per class on any given day was 9.3 SD = 2.7), which is within the
National Association for the Education of Young Children (NAEYC) guidelines
for 4-year-olds (10: 1 for a group of 20).
The majority (71 ) of the Chapter 1 funded classrooms were half-day pro-
grams, with the rest providing an extended day (with lunch and a rest period) of 4
to 6 hr.
Measures
A number of instruments were used to collect both child-level and classroom-
level data. Because the overall study was observation based, the measures were
selected after extensive review of available classroom-level (Goodson, 1990) and
child-level (Seppanen & Love, 1990) observation instruments. In addition to
7/24/2019 Relations Between Teacher and Classroom Activity Variables
6/30
258
BRONSON, TIVNAN, AND SEPPANEN
observational measures, a standardized cognitive test was selected and informa-
tion was collected on child and family background characteristics.
Child and Family Background Characteri stics.
Background information was
collected from three sources: a parent questionnaire, a teacher questionnaire, and
examination of student records. A number of children in the sample had charac-
teristics associated with risk for school failure. Other studies have suggested that
although the effect of any one background variable may be negligible, the cu-
mulative effect of a number of them may affect childrens performance on the
child-level measures (Hauser-Cram, Pierson, Walker, & Tivnan, 1991; Sam-
eroff, 1975; Sameroff, Seifer, & Zax, 1987). Schorr and Schorr (1988) found that
it takes more than a single risk factor to elicit an adverse outcome (p. 25).
To analyze the cumulative effects of child and family background risks on the
child-level variables, a risk composite variable was used that was composed of
nine child and family background variables. It was not possible to collect data on
some of the more potent risk variables examined by Schorr and Schorr (1988;
such as low birth weight, substance abuse, or child abuse), but nine available
variables conceptually associated with risk were included. The variables in-
cluded in the composite were: (a) English not spoken in the home, (b) limitation
of school activities, (c) special education requirements, (d) teacher rated behav-
ioral or emotional problems, (e) family problems, (f) living away from mother,
(g) mothers age, and (h & i) parents education. The criterion for each risk
variable component was based on analysis of the Fall 1990 family background
data and on analyses of the distributions of background data collected during
Spring 1991. Children were given one risk point if an aspect of their background
met the established risk criterion for that variable. Each child received a compos-
ite risk score that could vary from 0 to 9.
Chi ld Level Measures. The design of the Chapter 1 study called for individu-
al assessment of children in the fall and spring of the prekindergarten year. In a
fall pretest and spring posttest, child characteristics were measured with a stan-
dardized cognitive index, the Preschool Inventory (PSI) (Caldwell, 1976), and a
teacher rating scale, the Child Behavior Rating Scale (CBRS; Bronson, Good-
son, Layzer, & Love, 1990). In the spring, behavioral observations in the class-
room environment were done using the Bronson Social and Task Skills Profile
(BSTSP), 1991 revision (Bronson, 1991). This study examines data from the
spring data collection. Table 2 lists the child-level variables included in the
analyses.
The PSI is a brief, 32-item assessment of cognitive achievement designed to
be administered individually to children in the age range of 3 to 6 years old in
either English or Spanish. It was developed originally to provide Head Start with
a practical measure of preschool achievement. Correlations between PSI test
scores and the Stanford Binet Intelligence test ranged (in the five age groups)
7/24/2019 Relations Between Teacher and Classroom Activity Variables
7/30
OBSERVATIONAL STUDY OF CHAPTER
1
PREKINDERGARTEN PROGRAMS 259
TABLE 2
Child-Level Variables Included in the Analysis of Data From the Chapter 1 Study
Variable Categories and Variables
Standardized Cognitive Indicator
(PSI)
Score in the fall
Score in the spring
Teacher Rating (CBRS)
Mean competence rating in the fall
Mean competence rating in the spring
Classroom Behavior Observations
(BSTSP)
Mastery variables
Time in mastery activities (proportion of all observation time spent in a task with
a goal)
Use of organizing and planning strategies (rate of use of organizing and planning
strategies during tasks with goals)
Tasks completed successfully (proportion of tasks tried that are finished success-
fully)
Time distracted in mastery (proportion of mastery time spent distracted)
Social variables
Time in social activities (proportion of all observation time spent in social
interaction)
Time in organized social interaction (proportion of social time spent at the
organized and cooperative level)
Use of organizing strategies with peers (rate of use of strategies which organize
interactions with peers during time in social activities)
Use of accomodating strategies with peers (rate of use of social accomodating
strategies, like sharing and taking turns, with peers during time in social
activities)
Success in influencing peers (proportion of attempts to influence peers that were
successful)
Use of verbal means to influence peers (proportion of attempts to influence peers
that were verbal)
Self-management variables
Time uninvolved (proportion of all observation time not involved in any focused
activity)
Resists rules or teachers (rate of resisting rules or teachers during all observation
time)
from .39 to .65, with .44 being the correlation for the entire standardization
sample. Split-half reliability, estimated with the Spearman-Brown formula, was
reported as .95 on an earlier version of the PSI. The measure was selected
because it had been used in numerous national early childhood studies in recent
years. It yields an overall summary score that was included in the analyses for
this study.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
8/30
260
BRONSON, TIVNAN, AND SEPPANEN
The CBRS is a 32-item behavioral rating scale designed to assess categories
similar to the BSTSP. An earlier version of the CBRS was created by RMC
Research Corporation for use in a Head Start evaluation (Meleen, Love, &
Nauta, 1988). The CBRS is designed to evaluate a childs task behavior and
social behavior with peers and adults. The 32 behaviors are rated on a 5-point
scale ranging from 1 (never) to 5 (usually/always) to indicate how frequently
they occur. It was used in the Giant-Step evaluation (Abt Associates, 1988) in
which items on task orientation and strategies were more strongly correlated with
the PSI than were the adult and peer interaction items. There was a fall-spring
correlation of .67 for ratings of 364 Giant-Step children by their teachers. Inter-
nal consistency (Cronbachs alpha) was .96 overall for the Giant-Step sample. A
mean rating for all the items was computed to create the summary score used in
these analyses.
The BSTSP is an observation measure that was adapted from an earlier
version of the instrument (Bronson, 1985) for use in the Chapter 1 substudy.
The BSTSP uses structured categories to record a childs performance in plan-
ning and organizing mastery activities (puzzles, matching and sorting tasks,
number and letter tasks, etc.) and carrying them out successfully, in interacting
cooperatively and successfully with others, and controlling and directing oneself
appropriately in the school setting.
Social and mastery task behaviors are recorded in structured categories by an
observer who had been trained to a criterion of 90 interobserver agreement in
every category. A modified time sampling method is used in which half of the
observations are begun at the beginning of a social interaction and half are begun
at the beginning of a memory task. Whenever possible, observations are carried
out when children have some choice and control over the selection and direction
of their activities. In this study, observations were carried out in the natural
setting of the classroom with each child being observed for four to six lo-min
periods over 1 to 2 weeks. Technical information from the use of the BSTSP in
several studies is available (Bronson, 1985, 1994).
Because the focus of this study was on documenting relations between
classroom-level variables and the childs classroom experience, data from 12
BSTSP child-behavior variables were included in the analyses. Variables in three
areas (mastery, social, and self-management) were selected on the basis of their
usefulness in other studies (Bronson, 1994; Bronson, Pierson, & Tivnan, 1984;
Hauser-Cram, Bronson, & Upshur, 1993; Pierson et al., 1983). These are listed
in Table 2 and described more fully in Appendix A.
Classroom Level Measures. A number of instruments measuring the class-
room environment, classroom activities, and teacher interactions with children
were included in the descriptive component of the Observational Study of Early
?See Seppanen, Bronson, and Metzger (1992) for a description of the adaptation procedures.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
9/30
OBSERVATIONAL STUDY OF CHAPTER
1
PREKINDERGARTEN PROGRAMS
261
Childhood Programs, which studied 119 early childhood classrooms (Abt Asso-
ciates, 1991). Seven of these instruments were used in the Chapter 1 substudy
(Seppanen et al., 1993). This article includes data from variables selected from
two of the instruments, chosen on the basis of their relevance to specific educa-
tional practices. Table 3 lists the eight classroom-level variables selected for
inclusion in the study.
In the Abt Associates Adult-Focused Observation (Abt Associates, 1991), an
observer records the behavior of the classroom teacher in structured categories on
a time-sampled basis. Observers were trained to a criterion of 85 interobserver
agreement in each category. Recorded information includes the type of involve-
ment with children (teaching, management, etc.), the number of children with
whom he or she is involved, specific verbalizations, the content of teaching, and
techniques for communicating. The instrument was developed for the Observa-
tional Study of Early Childhood Programs and no psychometric analyses have
yet been done. Two variables from this instrument were included in the analyses
(see Table 3).
The Preschool ClasSroom Snapshot (adapted by Abt Associates, 1991, from
Stallings, 1975) provides a means for recording, in 24 categories, the types of
activities and groupings that are in evidence in a classroom at a given moment.
Observers were trained to a criterion of 85 percent interobserver agreement in
each of these categories. Multiple snapshots were recorded in classrooms at
TABLE 3
Classroom-Level Variables included in the Analysis of Data From the Chapter 1
Study
Instruments and Variable Categories
Variables From Abt Associates
Adult-Focused Observationa
Proportion of time teacher spends teaching
Proportion of teaching focused on cognitive concepts
Variables From Preschool Classroom
Snapshotb
Proportion of goal-directed activities (a composite of math and language, science and
natural world, sewing, woodworking, cooking, block construction, table games and
puzzles, looking at books)
Proportion of math and language activities
Proportion of activities involving block construction
Proportion of sand and water play
Proportion of dramatic and fantasy play
Average child : staff ratio in observed snapshot instances
avariables derived from the Abt Associates Adult-Focused Observation instrument
are based on brief time samples of adult behaviors. bvariables derived from the
Preschool Classroom Snapshot are proportions of snapshot instances in which the
event occurred.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
10/30
262
BRONSON, TIVNAN, AND SEPPANEN
regular time intervals and aggregated to produce activity and grouping scores.
Versions of the instrument were used in the National Day-Care Study (Stallings &
Porter, 1980) and by the New York City Board of Education (BOE) to describe
environments in BOE preschool and kindergarten programs (Jarvis, 1989). No
additional technical information is available. Six variables from this instrument
were included in the analyses for this study (see Table 3).
The classroom-level instruments, and their use in the descriptive part of the
Observational Study of Early Childhood Programs, are described in detail in
Layzer, Goodson, Layzer, and Levin (1991) and in Layzer et al. (1993).
Rationale for Classroom Level Variable Selection
The two instruments previously described included many more classroom vari-
ables (see Layzer, Goodson, & Moss, 1993) than the subset selected for inclu-
sion in this study (Table 3). The two teacher variables from the Abt Associates
Adult-Focused Instrument were selected because they showed ways teachers
spend time in classrooms that seemed likely to be related to childrens experi-
ences and cognitive or social outcomes. The five classroom activity variables
from the Preschool Classroom Snapshot were selected to reflect the variety of
activities typically found in classrooms for young children. The sixth variable
from the snapshot instrument, child:staff ratio, was selected because variations in
ratio have been associated with observed behavior differences in children in
previous studies (Prescott, 1973; Ruopp, Travers, Glantz, & Coelen, 1979; Van-
dell & Powers, 1983). All eight variables provide information on specific educa-
tional practices that could be influenced by public policy.
Data Collection Procedures
This study focuses on data from the child-level and classroom-level measures
obtained during Spring 1991.
Collection of ChiM Level Data. The data collection team at each site con-
sisted of child assessors who were trained to complete either the cognitive mea-
sures or the BSTSP, data collectors who were trained to complete the full battery
of classroom-level measures, and a field coordinator. The BSTSP observations
were completed by child assessors who spent approximately 2 weeks in each
class on days in which PSI administrations or classroom-level observations were
not scheduled. During each data collection period, teachers completed the CBRS
rating forms.
The quality of data collected was monitored by both the field coordinator and
the data-analysis team. Observer reliability for the BSTSP was monitored twice
in the field to be certain that the observers maintained the standard of 90
interobserver agreement in each category achieved during training.
Collection of Classroom Level Data. The classroom-level data collection was
completed over a period of approximately 1 week per class. Each data collector
7/24/2019 Relations Between Teacher and Classroom Activity Variables
11/30
OBSERVATIONAL STUDY OF CHAPTER
1
PREKINDERGARTEN PROGRAMS
263
completed the full battery of classroom-level observations for a particular class-
room and also interviewed classroom staff.
The quality of classroom-level observation and interview data was monitored
primarily by the field coordinator at each of the sites. The initial training of the
observers was reinforced subsequently in weekly review meetings among field
coordinators and data collectors. Field coordinators collected field reliability data
for the time-sample observation measure (Abt Associates Adult Focused Obser-
vation) by observing and coding with every observer at least once during the data
collection period.
RESULTS
Observed Variable Scores
Chil l Level Vari ables. Table 4 lists the means and standard deviations of the
child-level variables. The PSI mean score was 20.3 (SD = 4.9) and the CBRS
mean score was 3.8 (SD = .68).
The BSTSP classroom observation mastery variables showed children in-
volved in mastery tasks during approximately 38 of their overall observation
time and distracted during these tasks about 4 of the time.6 While working on
tasks, children used organizing and planning strategies (gathering appropriate
materials, grouping materials according to a criterion, sequencing activities ap-
propriately) at a mean rate of 1.6 per min. They completed tasks successfully
almost 83 of the time.
Under the structured observation conditions that maximized time in mastery
or social interactions, approximately 50 of the childrens time over all observa-
tions was spent in social interactions, with 38 of this time (19 of total
observation time) spent in social interactions with roles, rules, cooperative inter-
action, or true conversation (organized interaction). Involvement in social and
mastery activities could occur at the same time (a child could be working on a
puzzle and talking to a peer), so the time in social activities does not indicate that
the child was engaged exclusively in social interaction or a mastery task. During
interactions with peers, children used organizing strategies (Lets play store,
You be the mother) at the rate of approximately .6 instances per min (or about
once every 2 min) and accommodating strategies (You can share mine, We
could take turns using it) at the rate of about .15 times per min (about once
every 7 min). Children were successful in their attempts to influence peers 82
of the time and used verbal means to influence them 81 of the time.
Variables reflecting lack of adequate self-management (time uninvolved and
resisting rules or the teacher) were relatively low in frequency. Children in this
6Time allocated to either mastery or social behaviors is partially an artifact of the data collection
procedures that require that half the observations begin in a social interaction and half begin in a
mastery task. Therefore, time allocations do not necessarily reflect how a childs behavior is distrib-
uted in time during the ordinary course of a classroom day.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
12/30
264
BRONSON, TIVNAN, AND SEPPANEN
TABLE 4
Means and Standard Deviations of the Child-Level Variables=
Child-Level Variable
M
SD
PSI
Spring score
CBRS (Teacher Rating)
Spring summary score
BSTSP (Classroom Observations)
Mastery variables
Proportion time in mastery activities
Rate organizing and planning strategies
Proportion tasks completed successfully
Proportion time distracted in mastery
BSTSP (Classroom Observations)
Social variables
Proportion time in social activities
Proportion social time at organized level
Rate uses organizing strategies
Rate uses accomodating strategies
Proportion success in influencing peers
Proportion use of verbal means to influence
20.3 4.9
3.8 .68
37.6 13.5
1.6 1.0
82.7 22.2
4.0 5.9
49.8
16.4
38.3
29.1
.58 .54
.15
.20
81.5
13.6
81 .O
13.7
BSTSP (Classroom Observations)
Self-management variables
Proportion time uninvolved
Rate resists rules or teacher
1.7 3.0
.05
.I3
aN = 586 children.
sample were uninvolved less than 2 of the time and resisted rules or the teacher
at the rate of .05 instances per min.
Classroom Level Var iables.
Table 5 shows the mean classroom scores on
each of the eight classroom-level variables selected for this study.
Information from the Abt Associated Adult-Focused Instrument indicated that
39 of all teacher-child interactions were focused specifically on teaching (indi-
viduals or groups) and 44 of the teaching interactions involved teaching cogni-
tive concepts (language, reading, math, science). Mean percentages of activities
from the Preschool Classroom Snapshot records indicated that children were
involved in goal-directed activities 34 of the time. On average, 7 of all
observed activities involved math and language, 4 involved block construction
Goal-directed activities
included math and language activities, science and natural world activ-
ities, block construction, sewing, cooking, woodworking, table games and puzzles, and looking at
books or pictures.
7/24/2019 Relations Between Teacher and Classroom Activity Variables
13/30
OBSERVATIONAL STUDY OF CHAPTER
1
PREKINDERGARTEN PROGRAMS
265
TABLE 5
Means and Standard Deviations of Classroom-Level Variables8
Classroom-Level Variables
M SD
Variables From Abt Associates
Adult-Focused Observationb
Time teaching
Teaching focused on cognitive concepts
Variables From Preschool Classroom
Snapshotc
Goal-directed activities
Math and language activities
Block construction activities
Sand and water play activities
Dramatic and fantasy play activities
Average child : staff ratio
39 16
44 17
34
7
4
4
10
9
13
a
5
5
5
3
aN = 54 classrooms. bProportions based on time samples. CPro-
portions based on presence or absence of event in total number of class-
room snapshots.
4
involved sand and water play, and 10 involved dramatic or fantasy play.
The average enrollment across all the sample classrooms was approximately 19
(SD = 4.3), but the average attendance was only 16 so the observed child:staff
ratio was about nine children per adult in the classroom.
Table 6 gives the intercorrelations among the eight classroom-level variables.
Over all classrooms, teacher time spent in teaching was positively related to
teaching focused on cognitive concepts (r = .40, p < .Ol) and to the proportion
of snapshot instances in which goal-directed activities were observed (r = .36,
p
7/24/2019 Relations Between Teacher and Classroom Activity Variables
14/30
266
BRONSON, TIVNAN, AND SEPPANEN
TABLE 6
Correlations Among the Classroom-Level Variablesa
Variables 1 2 3 4 5 6 7 6
Time teaching
Teaching cognitive .40**
concepts
Goal-directed
.36** .33*
activities
Math and language
.27
.40** .63***
activities
Block construction
.I3
-.02
.I7 -.13
Sand and water -.12 -.Ol -.38**
-.30* -.I1
activities
Dramatic and -.07 -.I3
-.I4 -.45***
.26 .23
fantasy activities
Child
:
staff ratio .Ol -.20
.07
.24
-.08 -.33* -.15
aN = 54.
p