Over the last two and one-half decades several researchers have reviewed and summarized the vast literature on Direct Instruction. One of the most widely used methods is meta-analysis, a technique that uses statistical techniques to combine the results of numerous studies that test the same hypothesis. Several researchers have examined how Direct Instruction affects students' achievement through meta-analysis and other ways of assessing the research base. They consistently find that DI is highly effective compared to other programs.

Hattie (2009)¹ examined meta-analyses of research studies relating to student achievement, and concluded that Direct Instruction is highly effective. No other curricular program showed such consistently strong effects with students of different ability levels, of different ages, and with different subject matters. Hattie summarized the results of four meta-analyses that included DI, incorporating 304 studies, 597 effects, and over 42,000 students.

Similar results appear when the analysis is limited to studies involving whole school reform models. Borman and associates (2003)² examined studies of 29 comprehensive school reform models. They found that much more evidence was available for the Direct Instruction model than for other interventions, with 49 studies and 182 different comparisons for the DI studies compared to an average of only four studies and 23 outcomes for other interventions – a ten-fold difference. DI was found to produce the strongest effects of all models examined.

Adams & Engelmann (1996)³ conducted a meta-analysis of the literature on the effectiveness of DI programs, and analyzed data from 34 different research studies. The average effect size across these studies was .87, again far larger than the standard criterion of educational significance.

Couglin (2011)⁴ limited this analysis to studies of Direct Instruction that employed a randomized control group design. Twenty studies with 95 separate comparisons were identified. Slightly more than half dealt with reading (n=11), five involved mathematics, two regarded language and 2 regarded other areas. Slightly over half of the studies involved general education students. The average effects size over all 95 comparisons was .66, and similar for general education (.69) and special education students (.71)

Another method to explore the cumulative body of evidence was used by Stockard (2011)⁵ in an examination of scores on state assessment tests, the “high stakes tests” required under the No Child Left Behind Act. Data were reported from 18 different sites and 60 different comparisons, and the average effect of being in a DI school was .50. The impact was similar across different grades, schools with different SES and racial-ethnic composition, and in different areas of the country. In other words, over time, schools that implemented DI had increases in their average state assessment scores that were one-half of a standard deviation greater than that of other schools in their state.

Citations:

1. Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. London and New York: Routledge.
2. Borman, G.D., Hewes, G.M., Overman, L.T., & Brown, S. (2003). Comprehensive school reform and achievement: A meta-analysis. Review of Educational Research 73(2), 125-230.
3. Adams, G., & Engelmann, S. (1996). Research on Direct Instruction: 25 Years beyond DISTAR. Seattle, WA: Educational Achievement Systems. (With this one might be important to add that Adams did the meta-analysis and Engelmann wasn't involved with that part of the work.)
4. Coughlin, C. (2011). Research on the effectiveness of Direct Instruction programs: An updated meta-analysis. Presented at the Annual Meetings of the Association for Behavior Analysis International.
5. Stockard, J. (2011) Merging the Accountability and Scientific Research Requirements of the No Child Left Behind Act: Using Cohort Control Groups. Quality and Quantity, Published online 11 December 2011.