130 Chapter 6 For supported joint engagement, a RM ANOVA revealed no significant interaction effect between time and group. Due to violations of assumptions for RM ANOVAs on coordinated and symbol-infused joint engagement outcomes, Chi-square tests were employed instead. These tests indicated no significant differences between the BEAR and CAU groups for coordinated joint engagement [X2=.982 (2), p =.612] and symbol-infused joint engagement [X2 =.573 (2), p = .751] from baseline (T1) to endpoint (T2). Furthermore, Chi-square tests showed no significant differences from baseline (T1) to follow-up (T3) for both coordinated joint engagement [X2=1.95 (2), p = .337] and symbol-infused joint engagement [X2 =1.53 (2), p = .466]. Sensitivity analyses conducted on complete cases (details can be found in the Supplementary Materials) confirmed the results of the RM ANOVAs and Chi-square tests for all measures of joint engagement. In summary, the results suggest that the BEAR intervention does not lead to a significant increase in the amount of time children spend in joint engagement, nor does it significantly enhance various types of joint engagement. Secondary outcome measures Child RM ANOVAs did not reveal a significant interaction effect (time x group) on any of the secondary child outcomes. However, a significant time effect was observed for the N-CDI 4words understood (F1,12 = 7.27, p = .019, η2 = .377) and N-CDI words said (F1,12 = 11.42, p = .005, η2 = .488) in both the BEAR and CAU group. This suggests that both word comprehension and word production of the child (as rated by parents) significantly increased over time when comparing baseline (T1) to follow-up (T3). Additionally, the analysis of N-CDI words said revealed a significant group effect (F1 = 8.20, p = .014, η2 = .406), indicating that the children in the CAU group articulated more words compared to children in the BEAR group. Conversely, no significant differences were found in expressive language level and use, as measured by the JERI, between groups from baseline (T1) to endpoint (T2) [X2= 1.43 (2), p = .498] and from baseline (T1) to follow-up (T3) [X2=4.04 (2), p = .132]. Examining the total adaptive developmental age as measured by the Vineland Screener, the RM ANOVA revealed a significant effect of time (F1,53 = 74.19, p < .001, η2 = .583) from baseline (T1) to follow-up (T3). This indicates that total adaptive developmental age significantly increased over time, irrespective of the treatment group. Sensitivity analyses conducted on complete cases (see Supplementary Materials S1) confirmed these findings. 4 The results of RM ANOVA on the N-CDI words said and words understood are available in the Supplementary Materials, as missing data for the N-CDI was not imputed.
RkJQdWJsaXNoZXIy MjY0ODMw