The purpose of this meta-analysis was to synthesize findings from a total of 19 single subject
studies on interventions that enhance basic mathematics proficiency (number concepts and the four
arithmetic operation) of students at risk for academic failure and those with mild disabilities. Using
both Standard Mean Difference(SMD) and Percentage of Non-overlapping Data(PND), we examined
the impact of six intervention components(student grade level, number of sessions, measurement tool,
instructional strategy, disability, class type). Results showed that, when PND method was used, there
was a significant difference depending on the number of sessions dedicated to the intervention
phase(more than 20 sessions > less than 20 sessions). A significant difference was also found
between the effect size of direct instruction and that of other instructional strategies. Difference by
class type was found only from SMD method.
Findings of this study revealed several discrepancies between the results from PND method and
those from SMD method. When significant difference was found from both of the methods, still the
effect size in each method using post-hoc test differs one another. Such finding suggests that when
we conduct meta-analytic research to identify evidence based intervention, we need to consider more
than one method to calculate effect size of single subject studies.