Acquiring science and social studies knowledge in kindergarten through fourth grade: Conceptualization, design, implementation, and efficacy testing of content-area literacy instruction (CALI).

418 students (212 CALI / 206 control) across 40 classrooms in 6 Florida schools; K n=83, grade 1 n=109, grade 2 n=75, grade 3 n=76, grade 4 n=75

US K-4 general education classrooms in a Florida panhandle district; content-area literacy instruction delivered inside the existing literacy block.

Within-classroom individual randomization: students were matched on passage comprehension inside each classroom, then randomly assigned to CALI or business-as-usual literacy (Houghton Mifflin core, no focused science/social studies content). Dose: four 3-week units across the year, 4 days/week, ~30 min whole-class initiation or 15-20 min per small group. Attrition 8.9%, evenly distributed and non-differential on fall comprehension. CALI was developed by this team through design-based implementation research (2010-2012) and the efficacy trial was run by the same team, with CONDITION DELIVERED BY RESEARCH-TEAM-HIRED TEACHERS rather than regular classroom teachers — an efficacy ceiling, not an effectiveness estimate. Journal of Educational Psychology 109(3); Crossref registers the print date as 2017-04, online 2016-09-12.

The single cleanest paired aligned-vs-independent contrast in this tranche, and the gap is roughly TENFOLD. On researcher-developed, unit-aligned pre/post assessments (12 multiple-choice + 3 open-ended per unit, written for the CALI units), social studies knowledge g = 2.27 and science knowledge g = 2.10. On standardized Woodcock-Johnson-III measures, the effects appear ONLY in grade 4 — Picture Vocabulary d = 1.20, Oral Comprehension d = 0.47, Passage Comprehension d = 0.22 — and the paper states plainly: 'There was no effect of treatment — either positive or negative — in any other grades.' So four of the five grade cohorts show a null on standardized measures while showing g > 2 on the aligned ones. Treat the grade-4 standardized effects with suspicion too: they rest on ~75 students, and d = 1.20 on WJ-III Picture Vocabulary from four 3-week units is not a credible population parameter. The defensible reading is that CALI teaches the content it teaches, extremely well, and its transfer to independent literacy measures is at best small and confined to the oldest grade.

LOW. Within-classroom randomization removes family, school and teacher selection entirely; the threats here are measure alignment, developer leadership, and the small per-grade cells that generate the implausible grade-4 standardized effects.

Unreplicated as a trial of CALI itself. It is included in Hwang, Cabell & Joyner's (2022) meta-analysis of integrated literacy and content-area instruction, which reports the same measure-type gradient across 35 studies (comprehension: g = 0.54 researcher-developed vs 0.25 standardized; vocabulary: 0.86 researcher-developed vs a non-significant 0.64 standardized; content knowledge 0.94 vs a non-significant 0.68).