Effects of Explicit Vocabulary Interventions for Preschoolers: An Exploratory Application of the Percent of Goal Obtained Effect Size Metric.

Abstract

Systematic reviews of literature are routinely conducted to identify practices that are effective in addressing educational and clinical problems. One complication, however, is how best to combine data from both group experimental design (GED) studies and single-case experimental design (SCED) studies. Percent of Goal Obtained (PoGO) has been developed as a metric to express the size of the effect relative to the distance to a goal, which could have broad applicability. This study sought to augment this descriptive index with estimates of standard errors, which are needed to use PoGO as an effect size metric in meta-analyses of SCED and GED studies. This study investigated the application of PoGO and standard errors to both SCED and GED studies examining a common intervention approach used with a single population. Sixteen articles investigating explicit vocabulary instruction applied to pre-K and kindergarten students were identified. PoGO and standard errors were calculated for variations of explicit vocabulary interventions. Evaluated interventions included six studies using exclusively an SCED, nine studies using a GED, and one that used both. PoGO was calculated for each treatment condition when applicable (i.e., alternating treatments designs). Standard errors and confidence interval limits also were calculated. PoGO effect size values ranged from 14.4% to 93.6%. PoGO for single-case experiments was 49.2% with a standard error of 7.26, and for group experiments, it was 30.8% with a standard error of 3.71. Despite variation in the percentage of goal obtained across studies, the high degree of overlap in PoGO and standard errors between single-case and group experiments provides an indication that systematic reviews can apply this effect size metric to combine information obtained across experimental designs.