a

OCD

Using the transformed OCI scores to investigate the relationship between OCD and ASD, OCI-R scores were a significant predictor of AQ scoresβ =.42, SE =1.47, t (358)=8.71,p <.001, accounting for 17.3% of the variance in AQ,F (1,358)=75.86, p <0.001. OCI scores were also predictive of IU, β =.51, SE=.04, t (358)=11.08,p <.001, and accounted for 25.3% of the varianceF (1,358)=122.87, p <0.001. The significant relationship between both IU and SOR and ASC had already been established in previous models. The final step of the model predicted AQ with both IU and OCI-R scores, which accounted for 27.5% of the variance, F (2,357)=69.22, p <0.001, with both transformed OCI-R (β =.23, SE= 1.59 t (358)=4.39,p <.001), and IUS-12 (β =.38, SE=. 0.43t (358)=7.20, p <.001) scores remaining significant predictors, as demonstrated in Figure 3a. After conducting a Sobel test, it was revealed that there is a significant partial mediation of IU (z =7.64, p <0.001).
Secondly within OCD, Figure 3b represents the mediation analysis conducted to determine if SOR was a significant mediator. SOR was significantly predicted by transformed OCI scores, β =.51,SE= 3.92, t (358)=11.27, p <.001, and accounted for 26% of the variance,F (1,358)=127.07, p <0.001. In the final step of the model, both transformed OCI scores and SOR together predicted 30.7% of the variance in AQ scores, F( 2,357)=80.48,p <0.001, yet both OCI (β =.20, SE= 1.56,t (358)=3.88, p <.001) and SOR scores (β =.43, SE= .018, t (358)=8.39,p <.001) remained significant predictors. A significant Sobel test suggests that there is partial mediation in the model (z =8.13, p <0.001).