The study was conducted in 59 preschool classrooms in Ohio and Virginia.
Preschool classrooms were drawn from a variety of settings, all of which prioritized academically
at-risk children for enrollment. The classrooms included: 23 affiliated with Head Start, 19
subsidized prekindergarten programs, 12 independent programs that accepted vouchers, and
five early childhood special education programs. Lead preschool teachers were recruited to
participate in the study at information sessions presented at early childhood organizations. Two
waves of teachers (84 total) were randomly assigned as part of Project STAR to one of three
conditions: a high-dose shared reading with print referencing condition, a low-dose shared reading
with print referencing condition, and a comparison condition. Justice et al. (2010) compared
the high-dose shared reading with print referencing (intervention) condition (31 classrooms) to
the comparison condition (28 classrooms), for a total sample size of 59 classrooms at baseline.
A random subsample of children for whom consent was obtained was included in the study,
for a total baseline sample of 379 children (201 intervention and 178 comparison). The baseline
sample of children was 54% female; 42% White, 37% African American, and 8% Hispanic. The analytic sample included 58 classrooms—30 intervention and 28 comparison. The analytic sample
included 278 children for the analysis of impacts on language development (151 intervention,
127 comparison) and 288 children for the analysis of impacts on alphabetics (159 intervention,
The intervention condition, high-dose shared reading with print referencing, lasted 30 weeks with
four whole-class reading sessions per week and no more than one session per day. Materials
included a set of 30 books; a schedule for reading; and a description of the scope, sequence, and
frequency of print-related targets to be addressed during each read-aloud. There were 15 defined
print-knowledge objectives for each book. Each week, teachers would read the prescribed book,
using verbal (e.g., questioning) and nonverbal (e.g., tracking print) references to address the print-knowledge targets for the book. After reading, each book was placed in the classroom library and
not used for instruction or class reading during the study period. The low-dose shared reading
with print referencing condition involved the same shared reading interactions and materials as the
high-dose condition, but adults read with children twice per week instead of four times per week.
No studies report immediate posttest results for the low-dose condition; however, 1-year and
2-year follow-up results are reported in Piasta et al. (2012) and are summarized in Appendix D.
Teachers in the comparison group also conducted a whole-class book-reading session four
times weekly for 30 weeks. These teachers received the same set of 30 children’s books and
the same schedule for reading as those in the intervention group. They were instructed to simply
read the books as they normally would. After reading, each book on the reading list was placed
in the classroom library and not used for instruction or class reading during the study period.
To measure the alphabetics domain at posttest, a composite measure of print awareness
was constructed from three standardized tests: the PWPA Test (Justice & Ezell, 2001; Justice,
Bowles, & Skibbe, 2006), a structured task that examines children’s print concepts, and two
subtests of PALS–PreK (Invernizzi, Sullivan, Meier, & Swank, 2004)—the Upper-Case Alphabet
Recognition Subtest, which asks children to name upper-case letters, and the Name Writing
Subtest, which asks children to draw a self-portrait and then write their names on it. To measure
the language development domain at posttest, the authors created a composite score
based on three subtests of the CELF-P:2 (Wiig, Secord, & Semel, 2004): Sentence Structure,
Word Structure, and Expressive Vocabulary. These subtests collectively measure language in
the areas of vocabulary, syntax, and morphology and require approximately 15–20 minutes to
administer. Assessments were administered to children in fall and spring of the school year.
At follow-up, 1 and 2 years after the end of the intervention, the authors used two standardized
measures in the alphabetics domain: Woodcock-Johnson-III (WJ-III) Letter-Word Identification
subtest and WJ-III Spelling subtest. At follow-up, the authors used two standardized
measures in the comprehension domain: WJ-III Passage Comprehension subtest and Peabody
Picture Vocabulary Test-4 (PPVT-4). For a more detailed description of these outcome
measures, see Appendix B.
Support for implementation
Teachers in the shared reading with print-referencing intervention group received explicit
directions and materials at the start of the academic year on how to implement a 30-week
read-aloud program in their classrooms using a print-referencing style. Intervention teachers
received 8 hours of professional development prior to the start of the school year, two feedback
letters based on videos of their read-alouds, and another 3 hours of professional development