**6. Results**

Table 1 shows the means and standard deviations by group for the October (pre) and December (post) assessments. Table 2 further describes the differences in performance by group. To determine if statistically significant differences occurred due to treatment, a repeated-measures analysis of variance (ANOVA) was conducted where the group composed the between-subject factor (BAU and FDL) and the BAS assessments for October and December formed the within-factor. Effect sizes for partial eta-square were interpreted using guidelines from Kirk (1996). Results from the multivariate test using Pillai's Trace showed that all students made statistically significant gains, *F*(1,68) = 67.37, *p* < 0.001, η2 = 0.498. The test for differences between groups was not significant, *F*(1,68) = 0.179, *p* = 0.674. However, the interaction test for simple effects (time-by-group) resulted in statistically significant results with large effects, *F*(1,68) = 12. 91, *p* = 0.001, η2 = 0.160. A closer look reveals a 0.14 difference in the pretest means of the two groups (treatment = 5.15 and control = 5.29) which an independent samples *t*-test showed to be a non-significant difference, *t*(68) = 0.202, *p* = 0.841. When the December means were compared for the BAU and FDL groups, a difference of 0.75 favoring the FDL group was found (6.61 − 5.86 = 0.75). This difference between the groups was more than six times greater in December than in October (75 + 14 = 89/14 = 6.4). While the BAU group showed a 13% increase between October and December (5.86 − 5.29 = 0.67), the FDL group showed a 28% increase (6.61 − 5.15 = 1.46), a result 2.2 times greater than the BAU group. Figure 1 graphs the results showing that while the two groups began at very similar starting points in October, the FDL group out-gained the BAU group.


**Table 1.** Means (sd) by group for October and December.

**Table 2.** Pre- and post-test differences by reader group.


**Figure 1.** Raw score post-test gains by group.
