All data will be screened for normality, homoscedasticity, outliers, multicollinearity, and linearity in SPSS V. 22 prior to model specification. Pearson correlations will then be performed to examine relations between parent–reported sleep problems and daytime behaviors (severity of ASD symptoms and challenging behaviors).
Following the strategies outlined by Aiken and West (1991), subjective sleep data (parent–reported sleep problems) will be analyzed using simultaneous multiple regression models. Specifically, these regression models will examine CSHQ sleep problem subscales (Sleep Onset Delay, Bedtime Resistance, Sleep Duration, Night Wakings, and Sleep Disordered Breathing) as predictors of daytime behaviors. Control variables will include child age, gender, family sociodemographic assets, and level of ASD functioning (as indexed by the Social Communication Questionnaire and/or the child’s ABA center). Separate regression models will be specified for each daytime behavior outcome including: negative affect, aggression, self–injury, and repetitive behavior. Given the estimated sample size (target n = 40), these analyses will only detect robust effects (standardized regression effect size of .31 or higher) with power of .80.
The proposed study will also use …show more content…
This will provide the opportunity to examine how well sleep predicts behavior both within (level 1) and between (level 2) subjects (McCrae et al., 2008). Level 1 analyses will address questions such as: On nights in which a child has poor sleep, does he/she also have increased challenging behaviors during the subsequent day? Level 2 will examine questions such as: Do children who are generally poor sleepers also have increased challenging behaviors in the ABA
Following the strategies outlined by Aiken and West (1991), subjective sleep data (parent–reported sleep problems) will be analyzed using simultaneous multiple regression models. Specifically, these regression models will examine CSHQ sleep problem subscales (Sleep Onset Delay, Bedtime Resistance, Sleep Duration, Night Wakings, and Sleep Disordered Breathing) as predictors of daytime behaviors. Control variables will include child age, gender, family sociodemographic assets, and level of ASD functioning (as indexed by the Social Communication Questionnaire and/or the child’s ABA center). Separate regression models will be specified for each daytime behavior outcome including: negative affect, aggression, self–injury, and repetitive behavior. Given the estimated sample size (target n = 40), these analyses will only detect robust effects (standardized regression effect size of .31 or higher) with power of .80.
The proposed study will also use …show more content…
This will provide the opportunity to examine how well sleep predicts behavior both within (level 1) and between (level 2) subjects (McCrae et al., 2008). Level 1 analyses will address questions such as: On nights in which a child has poor sleep, does he/she also have increased challenging behaviors during the subsequent day? Level 2 will examine questions such as: Do children who are generally poor sleepers also have increased challenging behaviors in the ABA