The practical significance of early intervention and developmental research was examined by computing measures of effect size for 237 statistical tests from 59 early intervention studies. Analysis of the data revealed that the average (median) treatment effect across the studies was in the range Cohen (1977) defines as a medium effect size (d index =.68). The distribution of treatment effect sizes was positively skewed with a large portion of treatment effects in the medium and small categories. The interpretation of measures of magnitude strength is discussed in relation to statistical significance testing. The argument is made that the evolving nature of early intervention research requires the use of analytic procedures sensitive to the degree of change expected in field-based dependent variables. Reporting measures of effect size along with traditional quantitative results will enhance the interpretation of statistical findings in early intervention and developmental research.
ASJC Scopus subject areas
- Pediatrics, Perinatology, and Child Health
- Developmental and Educational Psychology
- Psychiatry and Mental health