ow

tsto

ctiochs' usf fahanioninstruction is strong. Similar efforts could help parents play a further role in

ildhooelp thealingand leas about

domains, a developmental progression of skills (e.g., Adams, 1990; later preschool years to give greater emphasis on print forms and the

Journal of Applied Developmental Psychology 31 (2010) 195201

Contents lists available at ScienceDirect

Journal of Applied DeveCommittee on the Prevention of Reading Difculties in YoungChildren, 1998; Whitehurst & Lonigan, 1998). Precursor skills beginwith the ability to understand and produce language in the infant andtoddler years, skills practiced in shared reading with an adult whofocuses predominately on word meaning and book conventions suchas front and back, top and bottom. A second strand of skills relates todecoding print and sound units including the abilities to name lettershapes, make associations between sounds and alphabet letters, andseparate spoken words into constituent sounds (Adams, 1990;National Research Council, 2001; van Kleeck, 1998; Whitehurst &

correspondence between forms and meanings.A number of interventions have been designed to encourage

parents' reading with preschool children. Typically, they share thegoals of increasing the frequency of parentchild reading andimproving children's vocabulary, syntactic skills, and their knowledgeof the conventions of print and books. One of the most extensivelystudied book reading intervention programs is called dialogic reading.Unlike conventional reading in which the adult reads the text andoccasionally asks for contributions from the child, dialogic reading ishighly interactive. The child is encouraged to take an active role inLonigan, 1998). Whitehurst and Lonigan (1gent literacy skills and processes as o

Corresponding author. Maternal and Child Health PPublic Health, University of Washington, Seattle, WA 981685 9852.

E-mail address: colleenh@u.washington.edu (C.E. Hu1 Kathryn Payne is now at Seattle Public Schools' Hea

0193-3973/$ see front matter 2010 Elsevier Inc. Aldoi:10.1016/j.appdev.2010.02.002the processes leading toMost depict multiple

g domains and, within

preliteracy instruction. She advocates emphasizing print meaning andform sequentially, beginning with a focus on the meaning of wordsand print in the toddler and early preschool years and shifting in theindependent reading are relatively new.domains of knowledge, interactions amonCommunity-based research

Introduction

Parents, policy makers and early chreading with young children can hReading with young children is appassumed link between being read tomodels fromwhich to test assumptiond researchers agree thatem prepare for school.largely because of therning to read. Empirical

vocabulary and general world knowledge) and insideout skills(e.g., knowledge of the rules for translating writing into meaning).Like Whitehurst and Lonigan (1998), Adams (1990) suggests thesedomains of knowledge and skills are inseparable but are not fullyintegrated in the preliteracy period. van Kleeck (1998) proposes twostages of preliteracy development and posits their implications for998) group these emer-utsidein skills (e.g.,

telling the storythe plot and teacEpstein, 1994; Ware tied to the ch2- and 3-year-olteaches adults tevents picturedvocabularies. Th

rogram Box 357230, School of95, United States. Tel.: +1 206

ebner).d Start Program.

l rights reserved. 2010 Elsevier Inc. All rights reserved.

Parentchild interaction children's emergent literacy development.Language developmentEmergent literacy

with more active participatreading style through briefHome support for emergent literacy: Follimplementation of dialogic reading

Colleen E. Huebner a,b,c,, Kathryn Payne a,1

a Department of Health Services, University of Washington, United Statesb Department of Family and Child Nursing, University of Washington, United Statesc Department of Pediatric Dentistry, University of Washington, United States

a b s t r a c ta r t i c l e i n f o

Article history:Received 25 September 2007Received in revised form 4 February 2010Accepted 8 February 2010Available online 26 March 2010

Keywords:Reading

This study tested if parenvocabulary skills continuedparticipants received instruno prior instruction. Parentgroup differences in parentchild's age, and frequency odialogic reading behaviors t-up of a community-based

taught to use an interactive (dialogic) reading style to promote earlyread this way as their children grew older. Approximately half the 78

n in dialogic reading when their child was age 2 or 3 years, the other half hadild reading evaluated more than 2 years after instruction showed signicante of dialogic reading techniques. Analysis controlling for maternal education,mily reading found parents with prior instruction used on average 90% moreparents without instruction. Use of dialogic reading behaviors was associatedof the child in the reading session. Evidence of the ability to change parents'

lopmental Psychologywhile the adult coaches the child's understanding ofhes new vocabulary (Arnold, Lonigan, Whitehurst, &hitehurst et al., 1988). Dialogic reading techniquesild's developmental level. The program for younger,d children is taught to adults in two parts. The rsto ask simple questions about objects, actions, andon a page, to build a child's expressive and receptivee techniques taught next, typically 12 months after

1) parents and children who, more than two years previously, hadparticipated in an eight-week dialogic reading intervention (Huebner &

196 C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201the initial instruction, use open-ended questions and expansions toelaborate the child's comments and questions. These behaviors fostersimple grammatical skills and the ability to use language to describeand explain (Cole, Maddox, & Lim, 2006; DeBaryshe, 1993).Throughout, adults are encouraged to praise the child's participationin telling the story. The program for 4- and 5-year olds uses similarevocative techniques to encourage conversation about the picturesand words in the book and adds an emphasis on the understanding ofconcepts and narrative structure through adults' use of open-endedquestions and distancing prompts that ask children to relate aspectsof the story to their own experiences (Lonigan, 2006). Althoughdialogic readingmethods are relatively easy to learn andwell-liked bycaregivers of young children (Blom-Hoffman, O'Neil-Pirozzi, &Cutting, 2006; Huebner 2000a), without instruction the behaviorsoccur infrequently during shared reading (Dickinson & Keebler, 1989;Dickinson & Smith, 1994; Hammett, van Kleeck, & Huberty, 2003;Huebner, 2000b; Huebner & Meltzoff, 2005).

Dialogic reading programs with preschool children have beentested in 1-to-1 interventions with parents and children (Arnold et al.,1994; Huebner, 2000b; Whitehurst et al., 1988), with small groups ofchildren and a teacher in Head Start, and in other classroom settings(Huebner, 2006; Lonigan, Anthony, Bloomeld, Dyer, & Samwel,1999; Wasik, Bond, & Hindman, 2006; Whitehurst et al., 1994). Theintervention is implemented during a preschool year or, in commu-nity-based trials, taught to parents as a 4- to 8-week interventionprogram. In most studies, outcome data have been collected 6 weeksto 6 months following the end of the intervention period. Taken insum, these studies have shown positive effects of dialogic reading onexpressive language skills (Hargrave & Senechal, 2000; Huebner,2000b; Lonigan et al., 1999; Wasik & Bond, 2001; Whitehurst et al.,1994; Zevenbergen, Whitehurst, & Zevenbergen, 2003), use ofevaluative devices and reference to internal states of characters inthe story (Zevenbergen et al., 2003), and receptive vocabulary(Huebner 2000b; Wasik & Bond, 2001; Whitehurst et al., 1988). Ameta-analysis of 16 studies found signicantly stronger effects onexpressive than on receptive vocabulary (Mol, Bus, de Jong, & Smeets,2008).

The benets of dialogic reading, alongwith its low cost, ease of use,and high acceptability have led to large-scale dissemination throughpublic libraries (i.e., the Public Library Association's Early LiteracyInitiative), early childhood education programs, and informal com-munity-based programs (e.g., Hear and Say Reading with Toddlers,http://www.bainbridgeislandrotary.org/). As dialogic reading gainspopularity, it becomes increasingly important to know the extent towhich its effects on children and adults persist and its potential tosupport developmental advances in children's expressive languageskills.

To date, only one study has examined long-term effects of adialogic reading intervention. In this randomized controlled trial byWhitehurst and colleagues (Whitehurst et al., 1999), Head Startteachers and parents were taught dialogic reading techniques. HeadStart classrooms were randomized to receive their usual curriculumor dialogic reading plus a phonemic awareness curriculum in theclassroom. The phonemic awareness curriculumwas designed to helpchildren distinguish consonant sounds within spoken words. Thechildren, ages 3 and 4 years, participated in dialogic reading sessionsin small groups in the classroom and at home with their parents.Parents did not receive instruction in the phonemic awarenessactivities. Follow-up assessments at the end of kindergarten, rst,and second grade, focused on the children only. Standardized tests ofvocabulary and other emergent literacy skills (i.e., knowing the namesof alphabet letters, print concepts, early writing) showed signicantbenets for children in the dialogic reading condition at the end of theHead Start year and one year later, at the end of kindergarten. Positiveeffects of the intervention did not generalize to children's reading

scores in the rst or second grade. The authors speculate that theMeltzoff, 2005), and 2) a comparison group recruited from thecommunity at the same time as the intervention group but whoreceived no intervention.

Method

Participants

Setting and eligibility criteria. The intervention and follow-up studiestookplace in amixed-income, rural county inWesternWashingtonwithan annual birth rate of approximately 225 live births per year. Thesetting was selected in part because in this community relatively fewyoung preschool age children are enrolled in early education programs.A home-based intervention was both feasible and had the support ofpolicy makers, public health professionals, and the local libraries. Thesample includedparents contactedrstwhen their childrenwere ages 2or 3 years. At that time, one group (n = 125) participated in anintervention study that tested three methods of instruction in dialogicreading and found all methods were effective (Huebner & Meltzoff,2005). While the intervention study was concluding, a second set ofparents (n = 40) were recruited to create a comparison group for thisfollow-up study. Parents indicated their interest in the future study bygiving written permission to contact them when their children wereolder.

Participants eligible for the follow-up study were 108 parents whoagreed to future contact and whose children would be at age 4 yearsduring the period 8/20047/2006. Parents were invited by postcard ortelephone call to participate. Of the 108 who were eligible, 18 parentsdeclined, 4 children were excluded because they had begunkindergarten, and 8 families had moved out of the study area. Thenal sample included 78 parents and their children; 41 hadparticipated in the previous dialogic reading intervention and 37parentchild dyads had no prior experience with dialogic reading (seeattenuation of group differences was due to the modularity ofemergent literacy skills. Consistent with the models of the develop-ment of emergent literacy skills discussed above, they argue thatsuccess at independent reading depends on extensive vocabulary andprint knowledge (skills that can be developed in the early preschoolyears by dialogic reading) and increasingly complex decoding skills.Because the follow-up assessments did not include assessments of thefrequency or quality of parentchild reading interactions, it is notpossible to know if advantages might have been greater for the subsetof children whose families continued a habit of interactive reading.

Knowing if a brief and early dialogic reading program has long-term effects on the quality of shared reading is essential informationfor community-based efforts intended to promote young children'sschool and reading readiness. Given growing evidence of theinterdependent yet separable components of emergent literacy, asequence of developmentally-timed, parent-focused interventionsmay be warranted to support language and preliteracy skills over thepreschool years.

The present study is a long-term follow-up of parents who receivedinstruction in dialogic reading when their children were 2 or 3 years ofage. The goal was to learn if parents taught the techniques of dialogicreading when their children were young continued to read this way astheir children grew older. Based on previous research that found mostparents enjoyed dialogic reading (Huebner 2000a), we hypothesizedparents would maintain a dialogic style, demonstrated by moreinteractive reading behaviors. We hypothesized also that parents' useof dialogic reading behaviors would be associated with greaterparticipation by the child in sharing the story. To test these hypotheses,this study compared patterns of parentchild reading of two groups:Fig. 1).

197C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201Description of the sample. The majority of parents were Anglo-American (94% of mothers) with a family income derived from wages

Fig. 1. Flow diagram of study participants.(94%) rather than public assistance. Five mothers were less than30 years of age, the youngest was 26 years. Additional characteristics

Table 1Characteristics of study participants (N =78).

No instruction indialogic reading

n = 37

Prior instruction indialogic reading

n = 41

Percent (n) Percent (n)

Parent's relationship to childMother 96% (35) 90% (37)Father 2% (1) 5% (2)Guardian 2% (1) 5% (2)

Mother's educationHigh school graduationor less

0% (0) 15% (6)

One or more years college 100% (37) 85% (35)Child's age54 to 59 months 38% (14) 29% (12)60 to 64 months 62% (23) 71% (29)

Child's sex (male) 51% (19) 51% (21)Family read to childlast week12 days 5% (2) 7% (3)36 days 22% (8) 41% (17)Everyday 73% (27) 52% (21)

Family taught letters, wordsor numbers last weekNo days 5% (2) 10% (4)12 days 43% (16) 29% (12)3 or more days 52% (19) 61% (25)

In daycare or preschool0 to 9 h per week 46% (17) 41% (17)10 to 19 h per week 38% (14) 44% (18)20 to 40 h per week 16% (6) 15% (6)of the study participants are presented in Table 1. Tests for differencesbetween parents with and without prior instruction in dialogicreading found one statistically signicant difference: mothers withoutprior instruction had on average one additional year of formalschooling (p b .01). For participants in the prior intervention study,the average time between prior instruction in dialogic reading and thefollow-up assessment was 33.36 months.

At the time of the follow-up, the children were on average59.40 monthsof age (4 years, 11.4 months).Due todelays in scheduling,ve childrenwere older than 60 months. Because the study ndings didnot change appreciably when theywere excluded from the analyses, allchildren were included to increase precision of the estimates. Childrenwithout prior experience in dialogic reading were on average 15 daysyounger than those with experience. Although this difference did notreach statistical signicance, child's age was included as a covariate inthe analyses because age is a known correlate of early language skill.

Procedure

Parents and children completed one in-person assessment at theirhomeor at one of the twopublic libraries in the county. The assessmentswere conducted by one of three college-educated community residentstrained in the data collection protocol and unaware of parents' priorexperience with dialogic reading. The assessment took approximately45 min to complete and included: 1) a brief questionnaire about familydemographics and home literacy activities, and 2) an audio recording ofparentchild reading.

The reading session occurred at the end of the assessment visit.Parents were instructed to read with their children for 5min. Allparents read the same book in which a teacher and children play acounting game. The book was chosen because of its familiar topic (inthe U.S., approximately two-thirds of 4-year-old children haveexperience with center-based early child care or preschool; U.S.Department of Education, 2007), its simple plot, and description ofactions and objects. If parents nished the book before 5min elapsed,they re-read it or another story about bugs.

Measures

Sociodemographics and child characteristics. Sociodemographicinformation collected by written questionnaire included the parent'sage, education, ethnicity, marital status, and source of family income.Information about the children included age, sex, and the number ofhours spent each week in child care or early childhood educationprograms.

Children's literacy experiences. Included in the questionnaire wereitems that asked about reading frequency and literacy-orientedexperiences in the home. The items, taken from the NationalHousehold Education Surveys for the year 2001 (U.S. Department ofEducation, n.d.), included the following two questions: how manytimes have you or someone in your family read to your child in thepast week? (0 days, 12 days, 36 days, everyday) and in the pastweek, have you or someone in your family taught (him/her) letters,words, or numbers? (0 days, 12 days, 3+days).

Parentchild reading style. Audio-recordings of the parentchildreading sessions were reviewed and coded to determine the extent towhich parents used the techniques of dialogic reading, eitherspontaneously or as a result of prior training, to engage their childin the story. Coders were not aware of parents' prior experience withdialogic reading.

The coding scheme for the dialogic reading behaviors was similarto that used in previous studies (i.e., Huebner 2000b; Huebner &Meltzoff, 2005). It is a frequency count of behaviors that typifydialogic reading with 2- and 3-year olds (i.e., use of what questions,questions about function or attributes, repetition, labeling, imitative

directives, prompts to say more, making connections to the child's

week. Nearly two-thirds (62%) reported a family member read withtheir child everyday. Reading frequency did not differ signicantlybetween the two study groups. Direct teaching of letters, words, ornumbers during reading or in other family interactions was reportedto occur less often than shared reading. For the group as a whole, 8%reported this occurred not at all in the prior week, 36% reported doingso on 1 or 2 days, and 54% reported doing so on 3 or more days. Thedifference between the study groups was not statistically signicant.

Exposure to preschool or daycare was considered anotherpotential source of literacy experience. At the time of the follow-upassessment, when the study children were on average nearly 5 yearsof age, 85% were enrolled in a daycare or preschool program. Forty-four percent of the children attended daycare or preschool 0 to 9 h perweek, 41% attended 10 to 19h per week, and 15% attended 20 or morehours per week. The average hours per week spent in daycare orpreschool did not differ signicantly between study groups.

Parents' use of dialogic reading behaviors during shared reading

Table 2 presents data on the frequency of dialogic readingbehaviors used by parents in each study group. Averages are reported

198 C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201experiences, praise, open-ended questions, and expansions) as well asbehaviors adults are asked to minimize during dialogic reading (i.e.,reading without including the child, use of yes/no questions, andpointing questions).

Dialogic reading behaviors were computed as rates in 5min ofreading. Reading samples of less than 5min duration were adjustedupwards to yield ve-minute averages. The average sample lengthbefore adjustment was 4.4min (SD=0.6); the shortest sample was2.9min.

Inter-rater agreement between a coder and an expert (CEH) forparents' use of dialogic reading behaviors was computed based on 21%of the audio records. The agreement statistic used for the dialogicreading behaviors was intraclass correlation using a 2-way randomeffects model and an absolute agreement denition. Intraclasscorrelations (ICC) were computed for 10 of 13 parent codes. Theaverage ICC was .86. ICCs were not computed for parent behaviorsthat occurred less than twice in the entire sample, these were:imitative directives, questions about function or attributes, andpointing requests.

Two indicators were used to summarize the child's participation inshared reading: the frequency of multi-word phrases and a free-speechmeasure of children's syntactic maturity based onmean lengthof utterance in words. We chose to report the longest ve utterancesbased on words because in this study the amount and clarity of childspeech varied greatly. ICCs for the child behaviors were .99 for multi-word utterances and .95 for the mean length of the longest veutterances in words.

Design

The design was a 2-group quasi-experimental study of the effect ofprior instruction in dialogic reading on parentchild reading beha-viors. Group differences were examined using regression analysis.Regression modeling allowed us to assess the effect of priorinstruction on the quality of shared reading after controlling forother variables that might affect this activity, specically: mother'seducation, child's age, and the frequency of family reading. The rstmodel, our primary analysis, tested the effect of prior instruction onparents' use of dialogic reading behaviors at the follow-up assess-ment. Because this outcome is a count variable that included multiplevalues of zero, a zero-inated negative binomial model was chosen.

Two additional regression models were used to examine associa-tions between parents' use of dialogic reading behaviors and thechildren's participation in the reading session. Two dependentvariables, the total number of multi-word utterances and the averagelength in words of the longest ve utterances, were assessed.Inspection showed signicant skew in the distribution of the numberof multi-word utterances. Four scores exceeded the 75th percentile ofthe distribution bymore than 1.5 times the inter-quartile range. Theseoutliers were replaced with less extreme scores by a procedure calledsemi-Winsorization (Barnett & Lewis, 1994) that uses a 1.5 Inter-quartile Range Rule to reduce the effect of outliers on the shape of thedistribution. The adjusted variable, a count variable of multi-wordutterances with four observations of a value of zero, was modeledwith a zero-inated negative binomial distribution. The averagelength of the longest ve utterances in words was modeled usinglinear regression. Statistical analyses were performed using Stata(Version 10.0, Statacorp, College Station, TX).

Results

Frequency of children's literacy experiences

Data describing the children's literacy experiences are presentedin Table 1. For the sample as a whole, reading was a relatively

common family activity. All parents reported reading in the priorfor the ten behaviors parents are instructed to increase in dialogicreading and three behaviors parents are instructed to decrease. Forthe sample as a whole, the least frequent dialogic behavior was toprompt the child to elaborate on a comment or gesture; the mostfrequent dialogic behavior was to ask a what-question. Two dialogicreading behaviors were not observed to occur at all: use of imitativedirectives (e.g., can you say) and questions about the function orattribute of a pictured object. Reading sentences from the bookverbatim was the most frequent behavior of all.

Association between prior instruction in dialogic reading and parents'reading behaviors

A dialogic reading (DR) summary score was constructed to test theassociation between prior instruction in dialogic reading and parents'use of these behaviors at 33 months, on average, post instruction. Thesummary score was the sum of all ten behaviors parents are taught toincrease in dialogic reading. For the sample as a whole, the variable,total number of DR behaviors ranged from 0 to 60 behaviors in the5-min reading session. The average number of dialogic behaviors washigher among parents with prior instruction (M=11.86, SD=12.10)than among parents without it (M= 6.05, SD= 6.23). The effect size

Table 2Item-level descriptives of dialogic reading (DR) behaviors during 5 min of sharedreading (N = 78).

No instruction indialogic reading

n = 37

Prior instruction indialogic reading

n = 41

Mean (SD) Mean (SD)

Adult behaviors to increase in DRWhat questions 2.44 (2.98) 3.10 (3.86)Attribute questions 0 .03 (.21)Repeats child .78 (1.54) 1.62 (2.39)Labeling .42 (1.03) .66 (1.14)Imitation directives 0 0Prompts to say more .12 (.54) .05 (.31)Connects to experience .24 (.60) .71 (1.28)Praise 1.08 (1.49) 2.78 (3.67)Open-ended questions .92 (1.93) 2.68 (4.26)Expansions .04 (.23) .21 (.54)

Adult behaviors to decrease inDRReading without thechild involved

69.63 (14.11) 58.48 (19.02)

Yes/no questions 3.98 (4.15) 5.29 (3.88)Pointing questions .03 (.19) .03 (.16)

corresponding to this difference was .57, interpreted as a moderateeffect size where .20 is considered small and .80 is large (Cohen,1998). A multivariable model was used to test the difference betweengroups after controlling for mother's education (in years), child'sage (in months), and frequency of family reading in the prior week(12 days, 36 days, or everyday). Maternal education was missingfor 3 subjects so the analysis was limited to a sample size of 75. Thenal model is presented in Table 3 as Model 1. Parents with priorinstruction were observed to use 90%more dialogic reading behaviorsduring shared reading than were parents without prior instruction.The z-test associated with receipt of prior instruction (yes = 1, no =0), after adjustment for the covariates, was statistically signicant(z = 3.04, p b .05).

Associations between parents' use of dialogic reading behaviors andchild's participation in shared reading

Two regression models were constructed to test if parents' use ofdialogic reading behaviors, regardless of their experience with the priorintervention, was associated with greater verbal participation of thechild in the reading session. Children's participation was examined intwoways: the number ofmulti-word utterances and the average lengthof the longest ve sentences in words. The number of multi-word

199C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201utterances ranged from 0 to 25 utterances in the 5-min reading session.Four extreme scores were reduced to values of 16.13 using thetechniques of semi-Winsorization (Barnett & Lewis, 1994) describedabove. A zero-inated negative binomial model was chosen for theanalysis because the distribution of scores included four zeros and theVoung test indicated the chosenmodel t the datawell. The nalmodelis presented in Table 3 as Model 2. There was a strong associationbetween the number of dialogic reading behaviors used by the parentand the number of multi-word utterances of the child after adjustmentfor three covariates: mother's education in years, child's age, andfrequency of family reading. The z-test associated with parents' dialogicreading behaviors was highly signicant (z= 4.12, p b .000).

The analysis of the dependent variable, the average length oflongest 5 utterances by the child included 73 children. The reason forthe reduction in sample size from 75 to 73 was two children did notspeak at all during the reading session and thus did not provide anadequate sample from which to calculate this variable. The meansentence length of the longest utterances for the 73 children was 3.78words (M=3.78, SD=1.31). The rangewas from 1.40 to 6.80words. A

Table 3Regression analyses: Associations between instruction in dialogic reading (DR),parents' use of DR, and child reading behaviors 2 years later.

Model 1a totalDR behaviorsby parent

Model 2a totalmulti-wordutterances by

child

Model 3b avg.length of longest5 utterances by

child

Variables IRR 95% CI IRR 95% CI B 95% CI

Mother's education (years) 1.08 .98, 1.19 .95 .88, 1.01 .11 .27, .05Age of child (months) 1.16 .99, 1.35 1.02 .92, 1.13 .06 .15, .27Frequency of familyreading (week)

1.04 .76, 1.43 1.01 .80, 1.28 .32 .20, .84

Parent's prior instructionin DR (yes)

1.90 1.26, 2.88

Parents' use of DRbehaviors (total)

1.03 1.01, 1.04 .04 .01, .07

Note. Number of observations used in Model 1 was 75; mother's education was missingfor 3 subjects. Model 3 is based on 73 observations because 2 children did not speakduring the reading session.

a Zero-inated negative binomial model for count data; values are incidence rateratios and associated 95% condence levels.

b Linear model; values are unstandardized regression coefcients and associated 95%

condence levels.linear model was used to test the association between parents' use ofdialogic reading behaviors and this indicator of the child's level ofparticipation. The model, with adjustment for mother's education inyears, child's age, and frequency of family reading, is presented as inTable 3 as Model 3. The overall model was statistically signicant (F (4,68) = 2.73, p b .05) and the variable of interest, the total number of DRbehaviors used by the parent, was the only independent variable toreach signicance (t= 2.62, p b .01).

Discussion

This study provides the rst evidence that brief instruction ininteractive reading has an enduring effect on parents' reading style.Parents taught to use dialogic reading behaviors when their childrenwere ages 2 or 3 years continued to use this reading style more than2 years later. The frequency of dialogic behaviors among those withprior instruction was nearly double that of parents with no priorinstruction. For the sample as a whole, use of dialogic readingbehaviors was associated with greater participation by the child intelling the story. Two-party discourse, here prompted by familiarity orcuriosity with the story, exposes children to linguistically complexand cognitively challenging experiences that teach the meaning ofnew words and concepts and are correlated with subsequentvocabulary and reading achievement (Beals, 1997).

Studies that implement dialogic reading as a universal preventiveintervention have been criticized for not delivering it only to children athighest risk and thereby greatest need.We disagree. In thismoderately-advantaged sample, there was substantial range in children's preschoolexperience and in home behaviors to foster language and otheremergent literacy skills. Although shared reading was reported to becommon, only slightly more than half (56%) of parents reported theirchild was taught within the family about letters, words, or numbers onthree or more days in the previous week. This percentage is lower thanthe 77% reported nationally for children ages 3 through 5 years (U.S.Department of Education, 2006). It is worth emphasizing that thechildren in this study were, on average, nearly 5-years old and mostwould not be identied as at risk for difculty learning to read due tosocio-economic disadvantage. County-level data for the study site fortheyears 20052009 indicate only 68 to 71%of third-grade childrenmetthe State's reading goals as measured by a standardized test andrequired by theNoChild Left Behind legislation of 2000. The countydataare consistentwith national statistics that show in theU.S., an estimatedone in three children has difculty learning to read (Adams, 1990).Practical, universal recommendations about how to support earlylanguage and literacy skills are needed to benet children from allsocio-economic strata.

Three aspects of the studymay have limited its results. First, no earlylanguage measures were collected for the comparison group childrenand thereby it was not possible to control for potential differencesbetween the groups at the follow-up assessment. Second, it is notpossible to know if prior instruction in dialogic reading or concurrentuse of the techniques were associated with stronger vocabulary skills.

A third limitation is that characteristics of parents in the comparisongroup showed they were not equivalent to parents in the interventiongroup in ways that could be related to a stronger motivation to supportchildren's preliteracy development and stronger readiness skills. Forinstance, all parents recruited to the comparison group agreed toparticipate in the study,whereas thiswasnot true of thedialogic readinggroup. One reason for the difference may be timing. All comparisongroup parents were recruited at the end of the dialogic readingintervention study (which was in the eld for approximately1.5 years). For them, less time elapsed between the time they wererecruited and data collection; less time to move out of the area and lesstime to reconsider their interest. Also, parents eager to participate in afuture study of children's language abilities and reading readiness

without receiving any benet in the here-and-now, may be more

parentchild book readings: A meta-analysis. Early Education and Development, 19,726.

200 C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201invested in these skills than parents who agree to participate in aresearch study with immediate rewards (i.e., instruction in dialogicreading and 2 free books for their children). These sources of self-selection bias, if they exist, would favor the comparison group childrenand reduce the likelihood of observing differences between groupsrelated to early experiences in dialogic reading. The implication is thatthe signicant effects on parents' reading behaviors observed in thepresent study may in fact underestimate the difference between thestudy groups.

Parents must be recognized as essential partners in helping allchildren prepare for the demands of formal schooling and learning toread. Although preschool curricula exist to support these goals,preschool programs can't carry the responsibility alone. In the U.S.nearly one-third (31%) of 4-year olds are not in center-based childcareor early child education programs (U.S. Department of Education,2007). For those who are, there is substantial variability betweenprograms in terms of curricula and, within programs, in the delity ofprogram implementation (e.g., Whitehurst et al., 1999). The ndingsof this study show parents are an under-utilized resource. Parentsmay have received the message about the importance of reading withyoung children (in this study, 94% reported their child was read to 3 ormore times in the previous week), but they need more informationabout how to use shared reading for instructional purposesthroughout the preschool years. For instance, in the present study,parents' use of distancing behaviors (e.g., Sigel, 1982) to connect themeaning of the text to the child's own experiences were observed tooccur less than once on average in the 5-min reading session. Use ofopen-ended questions, another technique that can prompt extendeddiscourse and in turn build knowledge of narrative structure (Snow &Beals, 2006) and of print skills (Korat, 2009), occurred only 2 to 3times on average among dyads with prior experience in dialogicreading and less than once per session among those without priorinstruction.

Maximizing shared reading to support language and other emergentliteracy skills may require parents to learn and practice behaviors thatare entirely new. Parental belief systems play an instrumental role ineveryday caregiving behaviors including home literacy activities(Goodnow & Collins, 1990; DeBaryshe, 1995). In a study of middle toupper-middle class parents and their 5- and 6-year-old children,DeBaryshe and her colleagues (DeBaryshe, Binder, & Buell, 2000)found parents' beliefs about children's literacy development and theirown roles in the processwere associatedwith their shared reading styleandwith children's literacy skills. Studies of low-income Latino familiesreport similar ndings. Individual differences in parents' beliefs abouthow children learn to read were related to how parents' used literacymaterials provided by the children's school (Goldenberg 1987; Gold-enberg, Rees, & Gallimore, 1992). Future studies are needed to test ifparents' beliefsmediate theirwillingness to adopt an interactive readingstyle.

Parent involvement in reading readiness holds considerable promisefor helping all preschool children prepare to succeed in school. Theevidence that brief instruction in dialogic reading can change parents'reading behaviors and support early vocabulary development is strong.Similar efforts are needed to help parents play a further role in theirchildren's emergent literacy development. Parents are willing but needguidance in how to do this. One approach would be to build on thesuccess of dialogic reading as a parent-focused community-basedprogram (e.g., provided through public libraries, family centers, worksites) to create a follow-up program for parents of older children.Inviting parent participation in program design and implementationwill likely to strengthen its acceptance and effectiveness.

Acknowledgments

This research was supported in part by grants from the University

of Washington's Research Royalty Funds and from the HealthNational Research Council (2001). Curriculum and pedagogy: The what and the how ofearly childhood education. In B. T. Bowman, M. S. Donovan, & M. S. Burns (Eds.),Eager to learn: Educating our preschoolersCommission on Behavioral and SocialSciences and Education. (pp. 182232). Washington, D.C: National Academy Press.

Sigel, I. E. (1982). The relationship between parental distancing strategies and thechild's cognitive behavior. In L. M. Laosa & I.E. Sigel (Eds.), Families as learningResources and Services Administration, Maternal and Child HealthBureau. We acknowledge also Karmen Meier for her assistance withprogram management, and we thank the parents and children ofJefferson County, WA, whose participation made this study possible.

References

Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge:MIT Press.

Arnold, D. S., Lonigan, C. J., Whitehurst, G. J., & Epstein, J. N. (1994). Acceleratinglanguage development through picture book reading: Replication and extension toa videotape training format. Journal of Educational Psychology, 86, 235243.

Barnett, V., & Lewis, T. (1994). Outliers in statistical data (3rd ed., pp. 143190). WestSussex, England: Wiley and Sons.

Beals, D. E. (1997). Sources of support for learning new words in conversation:Evidence from mealtimes. Journal of Child Language, 24, 673694.

Blom-Hoffman, J., O'Neil-Pirozzi, T. M., & Cutting, J. (2006). Read together, talk together:The acceptability of teaching parents to use dialogic reading strategies viavideotaped instruction. Psychology in the Schools, 43, 7178.

Committee on the Prevention of Reading Difculties in Young Children (1998).Preventing reading difculties in young children. In C. E. Snow, M. S. Burns, & P.Grifn (Eds.), Commission on Behavioral and Social Sciences and Education, NationalResearch Council. Washington DC: National Academy Press.

Cohen, J. (1998). A power primer. In A. E. Kazdin (Ed.), Methodological issues andstrategies in clinical research (2nd ed., pp. 339348). Washington, DC: AmericanPsychological Association.

Cole, K. N., Maddox, M., & Lim, Y. S. (2006). Language is the key: Constructive interactionsaround books and play. In R. J. McCauley, & M. E. Fey (Eds.), Treatment of languagedisorders in children (pp. 149173). Baltimore, MD: Paul H. Brookes.

DeBaryshe, B. D. (1993). Joint picture-book reading correlates of early oral languageskill. Journal of Child Language, 20, 455461.

DeBaryshe, B. D. (1995). Maternal belief systems: Linchpin in the home readingprocess. Journal of Applied Developmental Psychology, 16, 120.

DeBaryshe, B. D., Binder, J. C., & Buell, M. J. (2000). Mothers' implicit theories of earlyliteracy instruction: Implications for children's reading and writing. Early ChildDevelopment and Care, 160, 119131.

Dickinson, D., & Keebler, R. (1989). Variation in preschool teachers' styles of readingbooks. Discourse Processes, 12, 353375.

Dickinson, D. K., & Smith, M. W. (1994). Long-term effects of preschool teachers' bookreadings on low-income children's vocabulary and story comprehension. ReadingResearch Quarterly, 29, 104122.

Goodnow, J. J., & Collins, W. A. (1990). Development according to parents: The natures,sources and consequences of parents' ideas. Hillsdale, NJ: Erlbaum.

Goldenberg, C. N. (1987). Low-income Hispanic parents' contributions to their rst-grade children's word-recognition skills. Anthropology & Education Quarterly, 18,149179.

Goldenberg, C., Rees, L., & Gallimore, R. (1992). Effects of school literacy materials onLatino children's home experiences and early reading achievement. AmericanJournal of Education, 100, 497536.

Hammett, L. A., van Kleeck, A., & Huberty, C. J. (2003). Patterns of parents' extratextualinteractions during book sharing with preschool children: A cluster analysis study.Reading Research Quarterly, 38, 442468.

Hargrave, A. C., & Senechal, M. (2000). A book reading intervention with preschoolchildren who have limited vocabularies: The benets of regular reading anddialogic reading. Early Childhood Research Quarterly, 15, 7590.

Huebner, C. E. (2000). Community-based support for preschool readiness amongchildren in poverty. Journal of Education for Students Placed At Risk, 5, 291314.

Huebner, C. E. (2000). Promoting toddlers' language development through community-based intervention. Journal of Applied Developmental Psychology, 21, 513535.

Huebner, C. E. (2006). Optimizing the effects of shared reading on early language skills.In A. van Kleeck (Ed.), Sharing books and stories to promote language and literacy(pp. 149177). San Diego: Plural.

Huebner, C. E., & Meltzoff, A. N. (2005). Intervention to change parent-child readingstyle: A comparison of instructional methods. Journal of Applied DevelopmentalPsychology, 26, 296313.

Korat, O. (2009). The effect of maternal teaching talk on children's emergent literacy asa function of type of activity and maternal education level. Journal of AppliedDevelopmental Psychology, 30, 3442.

Lonigan, C. J. (2006). Development, assessment, and promotion of preliteracy skills.Early Education and Development, 17, 91114.

Lonigan, C. J., Anthony, J. L., Bloomeld, B., Dyer, S. M., & Samwel, C. (1999). Effects oftwo preschool shared reading interventions on the emergent literacy skills ofchildren from low-income families. Journal of Early Intervention, 22, 306322.

Mol, S. E., Bus, A. G., de Jong, M. T., & Smeets, D. J. H. (2008). Added value of dialogicenvironments for children (pp. 4786). New York, NY: Plenum.

Snow, C. E., & Beals, D. E. (2006). Mealtime talk that supports literacy development.NewDirections for Child and Adolescent Development, 111, 5166.

U.S. Department of Education (2006). National Center for Education Statistics. Thecondition of education 2006 (NCES 2006-071) http://nces.ed.gov/programs/coe/2006/section4/table.asp?tableID=496 (Table 33-1). Accessed November 25,2009.

U.S. Department of Education (2007). National Center for Education Statistics, NationalHousehold Education Survey 2005 http://nces.ed.gov/programs/coe/2007/section1/table.asp?tableID=662 (Table 2-1). Accessed November 25, 2009.

van Kleeck, A. (1998). Preliteracy domains and stages: Laying the foundations forbeginning reading. Journal of Children's Communication Development, 20, 3351.

Wasik, B. A., & Bond, M. A. (2001). Beyond the pages of a book: Interactive book readingand language development in preschool classrooms. Journal of EducationalPsychology, 93, 243250.

Wasik, B. A., Bond, M. A., & Hindman, A. H. (2006). The effects of a language and literacyintervention on Head Start children and teachers. Journal of Educational Psychology,98, 6374.

Whitehurst, G. J., Epstein, J. N., Angell, A. D., Payne, A. C., Crone, D. A., & Fischel, J. E.(1994). Outcomes of an emergent literacy intervention in Head Start. Journal ofEducational Psychology, 86, 542555.

Whitehurst, G. J., Falco, F. L., Lonigan, C. J., Fischel, J. E., DeBaryshe, B. D., Valdez-Menchaca, M. C., & Cauleld, M. (1988). Accelerating language developmentthrough picture book reading. Developmental Psychology, 24, 552559.

Whitehurst, G. J., & Lonigan, C. L. (1998). Child development and emergent literacy.Child Development, 69, 848872.

Whitehurst, G. J., Zevenbergen, A. A., Crone, D. A., Schultz, M. D., Velting, O. N., & Fischel,J. E. (1999). Outcomes of an emergent literacy intervention from Head Startthrough second grade. Journal of Educational Psychology, 91, 261272.

Zevenbergen, A. A., Whitehurst, G. J., & Zevenbergen, J. A. (2003). Effects of a shared-reading intervention on the inclusion of evaluative devices in narratives of childrenfrom low-income families. Journal of Applied Developmental Psychology, 24, 115.

201C.E. Huebner, K. Payne / Journal of Applied Developmental Psychology 31 (2010) 195201

Home support for emergent literacy: Follow-up of a community-basedimplementation of dialogic readingIntroductionMethodParticipantsProcedureMeasuresDesign

ResultsFrequency of children's literacy experiencesParents' use of dialogic reading behaviors during shared readingAssociation between prior instruction in dialogic reading and parents' reading behaviorsAssociations between parents' use of dialogic reading behaviors and child's participation in sh.....

DiscussionAcknowledgmentsReferences