Home Storybook Reading in Primary or Second Language With Preschool Children: Evidence of Equal Effectiveness for Second-Language Vocabulary Acquisition

ERCT Criteria Breakdown

• Level 1 Criteria

  • C

    Class-level RCT

    • Children were randomly assigned to treatment conditions individually within mixed classes, not by entire class or school, and no tutoring-style exception is invoked.
    • "Children, blocked by language group (Hmong, Spanish), were randomly assigned to one of two treatment groups." (p. 109)
    • Relevant Quotes: 1) "Participants were 44 preschool children, from low-socioeconomic status families, who were enrolled in one of two morning classes or one of two afternoon classes in a state-funded preschool located in Sacramento, California, USA." (p. 107) 2) "Two classes each of 11 children attended the morning program, and two classes each of 11 children attended the afternoon program. The two morning and afternoon classes shared the same space and were mixed during all activities." (p. 107) 3) "Children, blocked by language group (Hmong, Spanish), were randomly assigned to one of two treatment groups." (p. 109) 4) "A classroom-instruction-only control was not included because the theoretical focus of the study was on examining the language of storybook reading at home and its influence on second-language (English) classroom vocabulary learning. Classroom instruction in English was a constant." (p. 110) Detailed Analysis: The paper explicitly states that individual children (blocked only by primary-language group) were randomly assigned to treatment, not entire classes or schools. The four classes at the single participating preschool "shared the same space and were mixed during all activities," meaning children in different treatment arms attended the same physical classroom groupings throughout the day. The only condition that differed between groups was the language of the storybooks sent home for family reading; classroom instruction in English was constant for everyone, which limits (but does not eliminate) classic within-class contamination concerns for the school-based component. The ERCT exception for personal tutoring/one-to-one teaching could arguably be considered, since home storybook reading is inherently delivered individually by each family to their own child and cannot literally be assigned "by class." However, the paper does not frame the intervention as a tutoring or one-to-one instructional program, and the described exception is intended for interventions that are themselves individual instructional programs (e.g., private tutoring), not for home-based reading assignments layered on top of a shared classroom curriculum. Absent an explicit invocation of that exception by the authors, and given the explicit statement of individual-level random assignment, this criterion is judged not met. Criterion C is not met because randomisation was conducted at the individual-child level within classes that were mixed together, and no valid exception is clearly established.
  • E

    Exam-based Assessment

    • The study administered the standardised, nationally normed PPVT-III (and Spanish TVIP-H) alongside its custom storybook measure, satisfying the standardised exam-based assessment requirement.
    • "On the Peabody Picture Vocabulary Test (PPVT-III; 1997), children pointed to one of four word illustrations on a page after the researcher spoke the target vocabulary item. ... This was a standardized test with a national norming sample." (p. 108)
    • Relevant Quotes: 1) "Pictures illustrating six key vocabulary words (related to main episodes or characters) were selected from each of six books ... Words were selected based on the research team's judgment that many of them would be unknown to participating children." (p. 107) 2) "On the Peabody Picture Vocabulary Test (PPVT-III; 1997) ... This was a standardized test with a national norming sample. The technical data included in the examiner's manual indicated that there were 700 children in the 2- to 5-year-old norming sample..." (p. 108) 3) "The Spanish Test de Vocabulario en Imágenes Peabody (TVIP-H; 1986) is structured and administered in the same manner as the English Peabody Picture Vocabulary Test-Revised (PPVT-R) ... The norming sample reported in the examiner's manual included monolingual Spanish-speaking children from México City, Mexico (n = 1,219), and San Juan, Puerto Rico (n = 1,488)." (p. 108) 4) "On the Preschool IDEA Oral Language Proficiency Test (Pre-IPT: English; 1988)... This is one of two commonly used tests for assessing preschool English oral proficiency in the state of California." (p. 108) Detailed Analysis: The study's primary, study-specific storybook-vocabulary measure (six target words per book, picture-pointing task) was custom-designed by the research team and is not itself a standardised exam. However, the study also administered the PPVT-III, a widely used, nationally normed standardised vocabulary assessment, along with its Spanish counterpart (TVIP-H) and the Pre-IPT, a commonly used state-recognised oral proficiency measure. The ERCT standard requires that "a standard exam-based assessment is used," without requiring that it be the sole or primary outcome measure. Because a genuinely standardised, nationally-normed instrument (PPVT-III) with documented reliability and validity was used to assess English vocabulary alongside the custom measure, the letter of this criterion is satisfied, though the central outcome of the paper rests more heavily on the non-standardised storybook-vocabulary task. Criterion E is met because the study incorporated the standardised, nationally normed PPVT-III (and Spanish TVIP-H) as outcome measures in addition to its custom-built storybook-vocabulary task.
  • T

    Term Duration

    • The intervention and its retention testing spanned roughly 12-16 weeks from start to final assessment, approximating the lower bound of a full academic term.
    • "There were two 6-week sessions of home combined with classroom storybook reading." (abstract, p. 103)
    • Relevant Quotes: 1) "There were two 6-week sessions of home combined with classroom storybook reading." (abstract, p. 103) 2) "For each of 2 six-week sessions, children took a different book home every Friday and returned it the following Friday." (p. 109) 3) "The third testing on the vocabulary for books 1-6 served as a retention measure of words introduced in the first six weeks of the program after an approximate 10-week interval." (p. 110) 4) "Retention of vocabulary words from books 1-6 over the two-and-a-half-month period between the posttesting of the words and the retention testing that occurred after all the data for books 7-12 had been collected was analyzed by comparing these two scores." (p. 114) Detailed Analysis: The core intervention consisted of two sequential six-week (12 weeks total) home- plus-classroom storybook-reading sessions, with the immediate posttest for books 7-12 occurring roughly 12 weeks (about 3 months) after the intervention began. In addition, the retention test for the books 1-6 vocabulary was administered "an approximate 10-week interval" after the end of that session's posttest (itself at week 6), placing that measurement roughly 15-16 weeks (about 3.5-4 months) after the original intervention start. This places tracking for at least part of the outcome data at or just beyond the lower end of the "approximately 3-4 months" academic-term definition used by the ERCT standard. Criterion T is met because, taking into account the retention-interval testing, the overall period from intervention onset to final outcome measurement reaches approximately one academic term.
  • D

    Documented Control Group

    • The comparator (English-language home-reading) arm's demographics, size, and baseline performance are extensively documented and confirmed equivalent to the primary-language arm at pretest.
    • "Preliminary analyses of the children's scores on the PPVT-III, the vocabulary pretest for storybooks 1-6, and the Pre-IPT indicated that there were no significant differences between the groups assigned to storybooks in English and those assigned to storybooks in their primary languages." (p. 111)
    • Relevant Quotes: 1) "Participants were 44 preschool children, from low-socioeconomic status families, who were enrolled in one of two morning classes or one of two afternoon classes in a state-funded preschool located in Sacramento, California, USA. Primary languages of the children were Hmong (48%), Spanish (32%), and English (20%)." (p. 107) 2) "Data are reported for 33 of the original 35 children whose primary language was Spanish or Hmong ... Of these 33, 17 children were female and 16 were male. Twenty children spoke Hmong as their primary language and 13 spoke Spanish. The mean age of participating children in months was M = 52.13 (SD = 3.65)." (p. 107) 3) "Preliminary analyses of the children's scores on the PPVT-III, the vocabulary pretest for storybooks 1-6, and the Pre-IPT indicated that there were no significant differences between the groups assigned to storybooks in English and those assigned to storybooks in their primary languages in the first of the 2 six-week storybook sessions." (p. 111) 4) Table 2 (p. 112) reports means and standard deviations for storybook vocabulary, PPVT-III, and Pre-IPT scores separately for each treatment and primary-language group at every testing point. Detailed Analysis: This is a two-arm crossover comparative-effectiveness trial rather than a treatment-vs-untreated design, so the English-language home-reading arm functions as the comparison/control condition against which the primary-language arm is evaluated. This comparator group's size, demographic composition, and baseline performance on all outcome measures are reported in detail (Table 2), and the authors explicitly confirm baseline equivalence between arms. This level of documentation satisfies the requirement for a well-documented control/comparison group. Criterion D is met because the comparison arm is thoroughly documented, including demographics, size, and confirmed baseline equivalence.

• Level 2 Criteria

  • S

    School-level RCT

    • All children were drawn from a single preschool site, with randomisation at the individual-child level rather than across schools.
    • "Participants were 44 preschool children ... enrolled in one of two morning classes or one of two afternoon classes in a state-funded preschool located in Sacramento, California, USA." (p. 107)
    • Relevant Quotes: 1) "Participants were 44 preschool children ... enrolled in one of two morning classes or one of two afternoon classes in a state-funded preschool located in Sacramento, California, USA." (p. 107) 2) "Children, blocked by language group (Hmong, Spanish), were randomly assigned to one of two treatment groups." (p. 109) Detailed Analysis: The study took place at a single preschool site. Randomisation of the language-of-home-reading treatment was performed at the individual-child level (blocked by primary language), not across multiple schools or sites. No school-level or site-level random assignment occurred anywhere in the design. Criterion S is not met because randomisation and implementation occurred within one single preschool site at the individual-child level.
  • I

    Independent Conduct

    • The sole author designed the intervention, materials, and measures, and personally trained the classroom teachers who delivered it; no independent evaluation team was involved.
    • "Two trained undergraduate students served as teachers. These two teachers were trained by (a) observing the author perform two lessons and (b) reviewing, discussing, and practicing two lessons and receiving corrective feedback prior to program implementation." (p. 111)
    • Relevant Quotes: 1) "Two trained undergraduate students served as teachers. These two teachers were trained by (a) observing the author perform two lessons and (b) reviewing, discussing, and practicing two lessons and receiving corrective feedback prior to program implementation." (p. 111) 2) "Scripted lesson plans were used for the introduction of the storybooks and the vocabulary instruction to ensure fidelity of implementation." (p. 111) 3) "Theresa A. Roberts teaches in the Department of Child Development, California State University, Sacramento, USA." (p. 126, author note) Detailed Analysis: This is a single-author study. The author developed the theoretical framework, designed the intervention materials, wrote the scripted lessons, and personally demonstrated the lessons to the two undergraduate teachers who delivered the classroom component. There is no mention of an independent, third-party evaluation team conducting data collection or analysis separately from the person who designed the intervention. Criterion I is not met because the same person who designed the intervention also trained the implementers and appears to have overseen the study without independent oversight.
  • Y

    Year Duration

    • Because criterion T is not met, criterion Y cannot be met; independently, the 12-week program is far short of 75% of an academic year.
    • "There were two 6-week sessions of home combined with classroom storybook reading." (Abstract)
    • Relevant Quotes: 1) "There were two 6-week sessions of home combined with classroom storybook reading." (Abstract) 2) Table 1 (p. 110) shows the sequence of pretests, posttests, and program implementation spanning only weeks 1-12 of the study. Detailed Analysis: Per the ERCT standard's cross-criterion rule, because criterion T (Term Duration) is not met, criterion Y is automatically not met. Independently, the total tracked interval is approximately 12 weeks (about 2.8 months), which is nowhere near 75% of a full academic year (roughly 9-10 months). Criterion Y is not met both because criterion T failed and because the actual duration (12 weeks) is far short of the year-long requirement.
  • B

    Balanced Control Group

    • Both arms received identical storybooks, home-reading schedules, and classroom instruction; only the language of the home-reading books differed, so no extra resources were given to either group.
    • "A crossover design was used where one half of the children took home books 1-6 in their primary language, while the other half took home the same books in English." (p. 109)
    • Relevant Quotes: 1) "A crossover design was used where one half of the children took home books 1-6 in their primary language, while the other half took home the same books in English." (p. 109) 2) "During the week that followed the at-home reading of each storybook, two classroom lessons on the same storybook were conducted in English." (p. 110) 3) "Classroom instruction in English was a constant." (p. 110) Detailed Analysis: Both treatment arms received the same 12 storybooks, the same weekly home-reading schedule, identical classroom lessons in English (explicitly held constant across groups), and identical family- caregiver training events. The only manipulated variable was the language in which the home-reading books were printed; no additional time, budget, staffing, or materials were allocated to either arm over the other. Applying the ERCT criterion-B decision procedure: extra resources are not present for either arm (branch 1 of the decision tree), so the balance requirement is trivially satisfied. Criterion B is met because the two arms received identical dosage and resources, differing only in the language of the home-reading storybooks.

• Level 3 Criteria

  • R

    Reproduced

    • No replication of this study is mentioned in the paper, and a search did not identify an independent peer-reviewed replication of this specific home storybook-reading crossover design.
    • Relevant Quotes: No quotes regarding replication appear in the paper. Detailed Analysis: The article contains no mention of a prior or planned independent replication of its own design or findings. A citation-record search (Semantic Scholar's indexed citations of this article, conducted during this verification pass) shows the paper has been cited by numerous later studies on bilingual/dual-language shared book reading, including Grover, Rydland, Gustafsson, and Snow (2020), "Shared Book Reading in Preschool Supports Bilingual Children's Second-Language Learning: A Cluster-Randomized Trial" (a Norwegian classroom-based trial); Walsh, Sanchez, and Burnham (2016) on questioning styles in Head Start dual language learners; Pollard-Durodola et al. (2016) on content-related shared book reading for dual-language learners; and Brouillard, Dube, and Byers-Heinlein (2020) on book formats for bilingual preschoolers. None of these studies replicate this specific paper's home-language crossover design (Hmong/Spanish primary- versus English-language storybooks for home reading, paired with constant English classroom instruction); they test different interventions, formats, or populations by different research teams, and none present themselves as a reproduction of this study's specific design and findings. Criterion R is not met because no independent replication of this specific study design was found.
  • A

    All-subject Exams

    • Because criterion E is not met, criterion A cannot be met; the study also only assessed vocabulary, not other subject areas.
    • "Writing, reading, science, and math centers were not available on a regular basis." (p. 107)
    • Relevant Quotes: 1) "Writing, reading, science, and math centers were not available on a regular basis." (p. 107) 2) "The major purpose of the present study was to examine how both primary-language (L1) and second-language (L2) home storybook reading influenced second-language (English) vocabulary development." (p. 104) Detailed Analysis: Per the ERCT standard's cross-criterion rule, because criterion E (Exam-based Assessment) is not met, criterion A is automatically not met. Independently, the study exclusively assessed vocabulary/language outcomes; the preschool's own curriculum did not include regular math or science instruction, and no other subject-area outcomes were measured. Criterion A is not met because criterion E failed and because only a single outcome domain (vocabulary) was assessed.
  • G

    Graduation Tracking

    • Because criterion Y is not met, criterion G cannot be met; the study ends at the 12-week posttest with no tracking toward graduation and no identified follow-up publication.
    • Relevant Quotes: No quotes regarding graduation tracking or long-term follow-up appear in the paper; the results and discussion sections describe only the 12-week program and its immediate posttest and retention check. Detailed Analysis: Per the ERCT standard's cross-criterion rule, because criterion Y (Year Duration) is not met, criterion G is automatically not met. Independently, the paper reports outcomes only through the end of the 12-week program (plus an embedded retention check within that same window); there is no indication that children were followed until graduation from preschool or any later school stage. A search for follow-up publications by Theresa A. Roberts tracking this same Sacramento (Noralto) preschool cohort of Hmong- and Spanish-speaking children was conducted during this verification pass (ERIC database records for this author); it did not identify any such longitudinal follow-up. Roberts' other indexed publications concern different topics and different or cross-sectional samples (e.g., alphabet-letter instruction methods, phonemic awareness and word reading in English-language learners), not a graduation-tracking extension of this specific cohort. Criterion G is not met because criterion Y failed and no graduation-tracking or follow-up evidence was found in the paper or in later publications.
  • P

    Pre-Registered

    • The paper contains no statement of a pre-registered protocol, registry identifier, or registration date.
    • Relevant Quotes: No quotes referencing a study registry, protocol pre-registration, or registration date appear anywhere in the manuscript, including the Method and Notes sections. Detailed Analysis: There is no mention anywhere in the paper of a pre-registered protocol, a registry platform, or a registration date preceding data collection. The manuscript was submitted in March 2007 and accepted in November 2007, predating the routine use of study pre-registration in education research. A search of trial/protocol registries conducted during this verification pass did not locate any pre-registration record associated with this study. Criterion P is not met because no pre-registration evidence exists in the paper or elsewhere.