Please take our brief survey

Blueprints Programs = Positive Youth Development

Return to Search Results

Promising Program Seal

Story Talk - Interactive Book Reading Program

Blueprints Program Rating: Promising

A reading strategy intended to promote the development of language and literacy skills in young children from low-income families.

  • Barbara A. Wasik, Ph.D.
  • Professor, PNC Chair in Early Childhood Education
  • Temple University
  • College of Education
  • 1301 Cecil B. Moore Ave.
  • Ritter Hall 437
  • Philadelphia, PA 19122
  • (215) 204-4982
  • barbara.wasik@temple.edu
  • Early Cognitive Development
  • Preschool Communication/Language Development

    Program Type

    • Early Childhood Education
    • Teacher Training

    Program Setting

    • School

    Continuum of Intervention

    • Universal Prevention (Entire Population)

    A reading strategy intended to promote the development of language and literacy skills in young children from low-income families.

      Population Demographics

      The program is targeted at teachers of preschool aged children (3 to 5 years) from low-income families. The program has been tested with predominantly African-American children attending Title I early learning centers and Head Start preschools.

      Age

      • Early Childhood (3-4) - Preschool

      Gender

      • Male and Female

      Race/Ethnicity

      • All Race/Ethnicity

      Race/Ethnicity Specific Findings

      • African American

      Race/Ethnicity/Gender Details

      In study 1 (Wasik, Bond and Hindman, 2006), the sample children were 99% African American. In study 2 (Wasik and Bond, 2001), the sample children were 94% African American (there is no description of the other 6% of children). Neither study provided details on the gender mix of their samples.

      Children raised in poverty often enter school with limited exposure to books and underdeveloped literacy and language skills. As such, the program aims to increase the vocabulary skills of children from low-income families.

      • Individual
      • School
      • Family
      Risk Factors
      • Family: Low socioeconomic status
      Protective Factors
      • School: Instructional Practice

      The Story Talk - Interactive Book Reading program is a reading strategy intended to promote the development of language and literacy skills in young children (preschool aged). Teachers are trained to use specific book reading and oral language strategies and methods to increase opportunities for language and vocabulary development in other areas of teaching. Training lasts 15 weeks while the skills and reading strategies are incorporated into the daily classroom interactions throughout the school year.

      The intervention model focuses on training teachers in book reading and oral language strategies. Teachers are provided with prop boxes that include books, concrete objects that represent target words in the books and lesson plans. The props and books are used as part of the book reading and oral language activities. Teacher training lasts 15 weeks while the skills and reading strategies are incorporated into the daily classroom activities throughout the school year. The specific training components are described below.

      Book Reading Training: Teachers are trained in three areas of the book reading module: (a) asking questions, (b) building vocabulary, and (c) making connections. In asking questions, teachers are trained to ask open-ended questions that elicit more than one-word (yes/no) answers from the children. In building vocabulary and making connections, the teachers are taught to introduce key words before reading the story and to show children a concrete object (from the prop box or in the classroom) that represents the key words.

      Teachers are trained to ask the children questions such as, "What is this?" or "What do you call this?". These questions are followed with conversation builders such as, "What can I do with this?" or "What do you know about this?". Teachers are also trained to ask questions as they read the story to promote discussion. After reading the story, teachers ask questions that allow the children to reflect on the story such as, "What was the part of the story that you liked best?". Teachers are provided with example open-ended questions that relate to the story; however, once comfortable with the story and the strategy, teachers develop their own questions that encourage the children to talk about the book.

      As noted earlier, teachers are given prop boxes that contain story books with target key words, concrete objects that relate to the books and key words, a book of pictures with the key words, and lesson plans with suggested art and center activities that relate to the specific topic or theme of the prop box. Teachers are advised to change the theme each week.

      Oral Language Training: The oral language training is designed to train teachers how to use conversational strategies that promote multiple opportunities to speak, to actively listen, and to use a varying vocabulary. There are three parts of the oral language module: (a) practicing and promoting active listening, (b) modeling rich language, and (c) providing feedback.

      During the active listening component, teachers are trained to listen to what children say by directing their attention to the children, to patiently wait for the child to speak, and to respond in a meaningful way. The teachers are taught to acknowledge what the child said and try to extend the child's language about the concept about which the child was talking. For example, if a child says, "I see a dog." The teacher can respond, "Yes, you see a big black dog. What is the dog doing?". Further, if the teacher uses active listening, this teaches children to also be active listeners.

      In the modeling rich language component, teachers are trained to expand their use of vocabulary and to provide elaborate explanations and descriptions of common activities and events. For example, instead of saying "the glue is on the table" the teacher would be trained to say something more descriptive, such as "the glue is on the round table next to the scissors."

      In the providing feedback component, teachers are trained in three explicit strategies promoting children's language. One strategy is using "informational talk," an elaborated, rich description of common classroom activities. For example, "You are putting the big rectangular block on the small square block." The next strategy is to expand on the children's language. For example, if a child says "I built a house." The teacher replies, "Yes, you built a house using 10 blocks." The last strategy is to ask questions that encourage the children to use their language skills. Similar to the open-ended questions in the book reading module, teachers are encouraged to ask children questions such as, "Tell me how you did that?" or "What if..." thereby encouraging conversations with children about relatable, everyday occurrences.

      The intervention was developed out of recognition that children living in poverty start school with limited pre-literacy skills, to include limited vocabularies, than their more affluent peers. One explanation for this disparity is that children in affluent families have more exposure to books and have more opportunity to interact with their parents or caregivers. With concern of limited language and literacy opportunities in low-income families, more emphasis has been placed on instruction in the classroom. Prior evidence from experimental testing of preschool interventions with children from low-income families suggests that providing opportunities for children to talk and develop language skills is an important aspect of high-quality programs.

      The specific content of the intervention is based on dialogic reading, a systematic shared booking reading approach for children and their parents (or one-on-one reading with an adult). Dialogic reading is a method of reading picture books in which children are provided with multiple opportunities to talk and engage in conversation while the adult becomes an active listener, asks questions, adds information, and promotes the child's use of descriptive language. Experimental tests of its use with children from middle-income families showed that the method can lead to improvement in children's vocabulary.

      • Skill Oriented

      Interactive book reading has been evaluated in two randomized controlled trials involving preschool classrooms made up of children from low-income families. The 2006 study involved 16 teachers and 207 children, while the 2001 study involved two teachers and 121 children. The sample children were largely African American, were between the ages of 3 and 4 years old, and were enrolled in either Title I or Head Start preschool classrooms. Teachers received training in book reading and oral language development strategies. The children's vocabulary skills were tested using the Peabody Picture Vocabulary Test (PPTV-III) and the Expressive One-Word Picture Vocabulary Test (EOWPVT-III) immediately before and after the program was implemented (9 months in the 2006 study and 15 weeks in the 2001 study). The teachers were also observed in their classrooms to measure their use of the taught book reading and oral language strategies during story time as well as imbedded in the regular classroom activities.

      In both the Head Start (Wasik, Bond and Hindman, 2006) and Title-I (Wasik and Bond, 2001) studies, children in intervention classes scored significantly higher on tests of vocabulary skills than did their peers in the control classes. Classroom observations found that intervention teachers in both studies were significantly and substantially more likely than control teachers to use the target words and to use strategies that promoted language development during both book reading and other classroom activities.

      At posttest, intervention classes scored significantly higher on tests of vocabulary skills than did their peers in the control classes.

      In the Head Start study (Wasik, Bond and Hindman, 2006):

      • Children in the intervention classrooms scored significantly higher on both measures of vocabulary at posttest than did their peers in the control classrooms.
      • A medium to large effect was found on the children's receptive vocabulary (PPVT-II effect size: d = 0.73).
      • A small to medium effect was found on children's expressive vocabulary (EOWPVT-III effect size: d = 0.44).

      In the Title-I study (Wasik and Bond, 2001):

      • Children in the intervention classrooms scored significantly higher on the measure of receptive vocabulary and on tests of their knowledge of key target words.

      Classroom observations found that intervention teachers in both studies were significantly and substantially more likely than control teachers to use the target words and to use strategies that promoted language development both during book reading and other classroom activities.

      In the Head Start study (Wasik, Bond and Hindman, 2006): PPVT-II effect size: d = 0.73; EOWPVT-III effect size: d = 0.44. Effect sizes were not provided in the Title I study (Wasik and Bond, 2001).

      The samples were small. The larger of the two studies (Wasik, Bond and Hindman, 2006) included just two Head Start centers from the same geographic area (16 teachers, 121 children). The other study included just two teachers. Also, the samples were nearly all African American (99% in the Head Start study - Wasik, Bond and Hindman, 2006 - and 94% in the Title-I study - Wasik and Bond, 2001). Further studies using larger samples and varied demographics are needed before it can be certain that the effects are generalizable.

      • Neither study included a long-term follow up. Follow up studies are needed to test the lasting impact on vocabulary and, subsequently, whether there is an impact on later reading skills as expected.
      • The research team - who developed the intervention - is heavily involved in delivery of the teacher training and in the assessment of the outcome measures (in particular the teacher-specific measures). This involvement in the assessment of child-specific measures is less of a threat as the standardized measures were used (PPVT and EOWPVT). Nonetheless, bias is possible when assessors are not blind to group status.
      • In the Head Start study (Wasik, Bond and Hindman, 2006), the information on the attrition of children is limited; however, the rates of missing data are low (range from 8% to 2%).
      • One study selected and randomized only two schools, and the other selected one school and randomized four teachers. The small number of schools may not represent the population of low-income children.
      • Study 1 randomized schools but analyzed classrooms and individuals. The classroom analysis found significant effects with an n of 16, but the proper n would be two.

      • Blueprints: Promising

      Wasik, B. A., Bond, M. A. & Hindman, A. (2006). The effects of a language and literacy intervention on Head Start children and teachers. Journal of Educational Psychology, 98(1), 63-74.

      Wasik, B. A. & Bond, M. A. (2001). Beyond the pages of a book: Interactive Book Reading and language development in preschool classrooms. Journal of Educational Psychology, 93(2), 243-250.

      Barbara A. Wasik, Ph.D.
      Professor, PNC Chair in Early Childhood Education
      Temple University
      College of Education
      1301 Cecil B. Moore Ave.
      Ritter Hall 437
      Philadelphia, PA 19122
      Phone: (215) 204-4982
      barbara.wasik@temple.edu

      Study 1

      Wasik, B. A., Bond, M. A. & Hindman, A. (2006). The effects of a language and literacy intervention on Head Start children and teachers. Journal of Educational Psychology, 98(1), 63-74.

      Wasik, B. A., Bond, M. A. and Hindman, A. (2006). The effects of a language and literacy intervention on Head Start children and teachers. Journal of Educational Psychology, 98 (1), 63-74.

      Evaluation Methodology

      Design

      The randomized study included two Head Start centers (geographical location unclear). One center was randomly selected to receive the Interactive Book Reading teacher training and the other was selected as the control center. Sixteen teachers from the two centers participated. The 10 teachers in the intervention center received the intensive book reading program which included the teacher training, books and prop boxes. The control center received a list of the books being used in the intervention center and a stipend to purchase those books and others titles. Checks of the control center purchases and book use showed that all but three of the titles on the intervention list were put to use in the control classrooms.

      The teacher training and skills implementation in the classroom took place over the course of an academic year (September to June). The training included instruction, skills modeling, observation of skills implementation, and a minimum of 2 hours per month of direct coaching. Pre-test measures were applied during September and post-test measures were applied at the end of May/beginning of June.

      All 16 teachers who started the study also completed the study. There is no mention in the study as to whether any children dropped out completely. There a note that missing data on the children lowered the sample size on certain analyses. Sample sizes ranged from 191 to 202 (92% to 98% of the participants at the start of the study).

      Sample Characteristics

      Sixteen teachers participated in the project, 10 in the intervention group and 6 in the control group. All teachers remained in the study from beginning to end. Of the 10 teachers in the intervention group, 4 teachers had their bachelor of arts degree, 3 teachers had their associate of arts degree from a 2-year community college, and 3 teachers had a child development associate certificate. One of the teachers with the associate of arts degree was working toward her bachelor of arts degree. The years of teaching experience ranged from 3 to 32, with an average of 15 years. In the control sites, all 6 teachers had their bachelor of arts degree, and the years of teaching ranged from 1 to 17 years, with an average of 7 years.

      Two hundred and seven children, 139 in the intervention and 68 in the control group, participated in the study. Overall and group attrition is not explicitly stated; however, the sample sizes for the analysis ranged from 191 to 202. The mean age for the fall was 3 years 10 months, with children’s ages ranging from 2 years 8 months to 4 years 10 months. The Head Start centers were located in high-poverty neighborhoods, and the income eligibility criteria to receive Head Start services range from $8,980 and below for a family with one child to $30,960 for a family with eight children. Ninety-nine percent of the children were African American.

      Measures

      Data was collected on children in September (start of teacher training) and May/June (at the end of the school year). Reliability information was not provided on either the child-specific or the teacher-specific measures. The child-specific measures, however, are commonly used measures of young children's vocabulary.

      Child-specific measures: All children were individually pretested during September (baseline) and posttested during the end of May and beginning of June on the Peabody Picture Vocabulary Test (PPTV-III) and the Expressive One-Word Picture Vocabulary Test (EOWPVT-III). It is not clear who administered the tests to the children but they are self-completion. In addition, all children were administered a measure of alphabet knowledge in which they were asked to identify the 26 letters of the alphabet. Alphabet knowledge was not a part of the intensive book reading training, but all teachers in both Head Start centers received training on alphabet instruction by their on-site education coordinators.

      Teacher-specific measures: The teacher-specific measures were observational in nature and were primarily concerned with the extent to which teachers applied the taught skills in their classrooms. All teacher measures were carried out by the research team (who also were responsible for providing the teacher training).

      • Book Reading Observations: All teachers in the intervention and control classrooms were observed reading a book to their class in September (to collect baseline data) and again at the end of May/beginning of June (as a posttest measure). In addition, a running record of the teacher's talking and questioning strategies was recorded. The goal was to determine whether questioning strategies that promoted conversations with young children were used during book reading and whether the nature of the questioning changed because of the intervention. The data were coded to determine the types of questions that were asked during the book reading experience and to determine whether there was a change in questioning strategies over time.
        • The data were coded into four general categories: (a) informational questions related to the book, (b) informational talk related to the book, which included teachers comments about the book, (c) managerial questions unrelated to the book and focused on children’s behavior and attention to the story, such as “Are you listening to what I said?” and (d) managerial talk unrelated to the book, such as “Sit down” and “Pay attention.
        • The questions were further categorized into open and closed questions. Open-ended questions are defined as questions that require more than a yes–no or a one-word response.
      • Skills Application in Other Contexts: All teachers in the intervention group were observed while engaging with children in activities other than book reading both in September (baseline) and again in May (posttest). Teachers were observed for 30–40 min during activities other than book reading, and the target behaviors were noted as being observed or not observed. The purpose of this observation was to determine the degree to which teachers were implementing conversation strategies in contexts other than book reading. The three conversation strategies that the teachers were trained on were active listening, providing feedback, and modeling rich language. These strategies were quantified in behavioral terms, and a checklist was developed that reflected the targeted behaviors.
        • For the active listening component, the target behaviors included the following: (a) teacher offered explicit praise to children for active listening, (b) teacher discussed the idea of active listening, (c) teacher used the flannel board to promote active listening, and (d) teacher provided opportunities for active listening to occur.
        • For the providing feedback component, the target behaviors included the following: (a) teacher asked children to tell more about their ideas, (b) teacher expanded on children’s language, (c) teacher described children’s activities in detail, and (d) teacher provided opportunities for children to respond to feedback.
        • For the modeling rich language component, the target behaviors were the following: (a) teacher encouraged children to describe the features and function of objects, (b) teacher encouraged the use of theme-related vocabulary, (c) teacher used open-ended questions, and (d) teacher encouraged children to speak in more than one-word responses.
      • Training Observations: Teacher observations were conducted after each of the six trainings and modeling sessions were also collected and scored. The six observations coincided with six areas of training: the three for oral language development (active listening, modeling rich language, and providing feedback) and the three for book reading (asking questions, building vocabulary, and making connections.) Because each training session focused on different strategies, the protocol was different for each of the six observations. Behaviors specific to each training were included in each observation protocol. During observations, the trainers coded whether the target behavior was demonstrated by the teacher.

      • Fidelity of Implementation: Teachers were also rated on their level of implementation with regard to the intervention strategies on a 3 point scale: 0 = not observed and not implemented, 1 = observed inconsistently and 2 = observed consistently. Ratings were assigned by two raters, who agreed on 29/30 observations, resulting in a 97% inter rater reliability score.

      Analysis

      One-way analysis of covariance (ANCOVA) was used to assess the impact on posttest scores for the child specific measures (PPTV-III & EOWPVT-III). Pretest scores were used as a covariate. Separate analyses were conducted using the classroom as the unit of analysis and using individual children as the unit of analysis. Note that randomization took place at the center level, yet, analysis was not conducted at this level (presumably because the sample size of the centers was too small; n=2). Cohen's d is also reported for the child specific measures.

      Exploratory analysis using ANCOVAs were also conducted to determine what, if any, alternative explanations could account for the difference found at posttest. These included tests of baseline equivalence, tests of fidelity of implementation, tests accounting for the age of the children, and, finally, a comparison of the intervention and control classes on the alphabet measure (which used a proxy for general teaching skills since all teachers received alphabet training independent from the intensive book reading training).

      The teacher-specific measures were assessed using ANCOVAs, but also via paired t -tests and correlation analysis (here too, the study does not state if the tests for significance were one- or two-tailed). It is not explicitly stated if intent-to-treat analysis was used but this would appear to be the case.

      Outcomes

      Fidelity of Implementation
      ANCOVAs were conducted using scores from classroom observations that reflected the degree to which intervention teachers faithfully implemented the program as they were instructed. However, the scores themselves were not stated, so the overall extent to which the teachers implemented the program as intended is unknown.

      Results of the ANCOVAs indicted that higher posttest scores corresponded to higher levels of implementation. For PPVT-III (using classroom as the unit of analysis), the partial correlation between level of implementation scores and posttest scores (using pretest scores as the covariate) was .69 (p< .04). For EOWPVT-III, the partial correlation was .16 (ns).

      Baseline Equivalence and Differential Attrition
      Analysis of pretest scores for both receptive (PPVT-III) and expressive language (EOWPVT-III) revealed that children in the intervention classrooms had scores that were very similar to those of children in the control classrooms. The pretest means and standard deviations are as follows:

      • PPVT-III: intervention - M= 82.02, SD= 12.66; control - M= 81.48, SD= 11.86
      • EOWPVT-III: intervention - M=79.76, SD= 11.33; control - M=79.46, SD= 12.20

      The data on the baseline observations of book reading conducted during the fall (prior to the intervention) revealed no significant differences between intervention teachers and control teachers in the amount of informational open questions, closed questions, or teacher talk that occurred before, during, or after a book was read. Additionally, an analysis of variance revealed no differences for intervention and control teachers in asking open-ended questions in the fall.

      There was no attrition by teacher. All 16 teachers who started the study also completed the study. There is no mention in the study as to whether any children dropped out completely. There is a note that missing data on the children lowered the sample size for certain analyses. Sample sizes ranged from 191 to 202 (92% to 98% of the participants at the start of the study). There is no break out of missing data by group, nor is there an analysis of how the children with missing data differ from the full sample.

      Posttest
      Compared to their peers in the control classrooms, children in the intervention classrooms scored significantly higher on both measures of vocabulary at posttest. This effect was significant regardless of whether the child or the classroom was the unit of analysis. The specific results are as follows:

      • PPVT-III (student unit of analysis): F (1,189)= 33.28, p< .001
      • PPVT-III (classroom unit of analysis): F (1,13)= 27.13, p< .01
      • PPVT-II effect size: d = 0.73
      • EOWPVT-III (student unit of analysis): F (1,197)= 18.08, p< .001
      • EOWPVT-III (classroom unit of analysis): F (1,13)= 15.24, p< .01
      • EOWPVT-III effect size: d = 0.44

      After the intervention in the spring, however, there was a significant main effect for condition, with teachers in the intervention group talking more in general during book reading compared with teachers in the control group (when all categories are summed). More importantly, teachers in the intervention group asked more open-ended questions than did teachers in the control group. While there was no difference at baseline, there was a significant difference after the intervention. Also, within-condition paired t-tests were conducted to examine changes over time in (a) the total amount of open-ended questions that were asked at any time (before, during, and after reading) and (b) all other questions asked at these times (closed questions and general questions). Whereas intervention teachers showed a significant increase in the amount of open-ended questions after the intervention, control teachers did not show a significant change in the amount of open-ended questions.

      Wasik, B. A. & Bond, M. A. (2001). Beyond the pages of a book: Interactive Book Reading and language development in preschool classrooms. Journal of Educational Psychology, 93 (2), 243-250.

      Evaluation Methodology

      Design

      The study took place in Baltimore, Maryland. Four teachers, each teaching two preschool classes per day (a.m. and p.m.) participated in the study. The teachers were randomly assigned to either the treatment group (two teachers) or the control group (two teachers). The intervention lasted 15 weeks (mid-October to mid-May). The treatment group teachers were trained in interactive book reading and book reading extension activities. Training included instruction in defining vocabulary words, providing opportunities for children to use words from the books, asking open-ended questions, and providing children with opportunities to talk and be heard. Because teachers would be conducting large-group interactive reading sessions, the teachers were also given instruction on how to help children listen to their peers. Teachers were provided with books to read to their students and props representing target vocabulary from the books.

      Control teachers received all the books treatment teachers did. These books were read as often in control classrooms as they were in treatment classrooms; however, control teachers did not receive the training that treatment teachers did.

      For the first four weeks of the intervention, an experienced teacher (and member of the research team) modeled the shared book reading techniques in each treatment classroom and assisted with reading extension activities. For the next 11 weeks, treatment teachers ran the program without assistance.

      All four teachers were included in the analysis.

      Initially, there were 64 children in the intervention group and 63 in the control group. Six children transferred from the school, leaving 61 in the intervention and 60 in the control group.

      Sample

      121 four-year-old children from low-income families served as the study sample for this investigation (61 intervention and 60 control). The children were drawn from four pre-school classes at a Title I early learning center in Baltimore, Maryland. 95% of students at this center were eligible for free or reduced lunch and 94% were Black. The gender mix of the children was not stated. Demographic data was not provided on the four participating teachers.

      Measures

      All children were pre-tested individually on the Peabody Picture Vocabulary Test (PPVT-III), a widely-used measure of receptive vocabulary skills. The study does not state who administered the assessments. Following the 15 week intervention, children were re-assessed on the PPVT-III. Children were also tested on their knowledge of target words on a receptive and an expressive measure that was devised for the study. In addition, teachers were observed by the research team during a 20-minute activity session in weeks nine and eleven, and their use of ten target words was recorded.

      Analysis

      The design of this study was 2 (condition: treatment, control) X 4 (teacher) X 2 (time of day: A.M. class, P.M. class), with teachers nested within condition (2 experimental, 2 control) and time of day nested within teachers (each taught an A.M. class and P.M. class). For PPVT-III the design included a repeated measure factor as well. The analysis was done at the level of randomization (classroom) as well as at the individual level.

      The low attrition suggests that the study was intent to treat, although this is not stated explicitly.

      Outcomes

      Implementation Fidelity
      Fidelity of implementation was not explicitly assessed, though the teachers were observed in their classrooms as part of the training and support received. During observations, researchers counted the number of times each teacher used any of the 10 target vocabulary words during the 20-min extension activities. A separate 2 (group: treatment, control) X 10 (each word) ANOVA was conducted for each of the two extension activities. There were significant main effects of group and word for each activity, as well as significant Group X Word interactions.

      Baseline Equivalence and Differential Attrition
      There was very little attrition. Neither teacher left the study, and complete data was available on both. Further, only 3 children from each group (5% overall attrition rate) transferred from the school before the study ended. The groups were equivalent at baseline on the primary outcome measure (PPVT-III).

      Post-test
      At post-test, the children had improved on the main outcome measure. Treatment classes scored significantly higher on the PPVT-III than did control classes. The analyses with classroom as the unit of analysis revealed the expected interaction for both the A.M. classes, F(l, 2) = 62.73,p < .016, and P.M. classes, F(l, 2) = 346.08, p < .001. The means were 73.66 (treatment) and 72.01 (control) for PPVT-III scores at the pretest and 81.30 (treatment) and 72.10 (control) for PPVT scores at the posttest. The comparable analyses with students as the unit of analysis produced the same Treatment X Trial interaction, F(l, 120) = 13.69, p < .001.

      Treatment classes also scored significantly higher on their knowledge of target vocabulary words. Classroom observations found that teachers in the treatment group were significantly and substantially more likely than control teachers to use the target words during related activities.