Factors Associated With Outcomes in Type 1 Diabetes
Strengths of the present study include the use of multimethod, multi-informant data collection procedures, including parent and adolescent report of family organization, youth, and parent self-efficacy for diabetes and diabetes management, as well as a biological measure of metabolic control collected at two time points. Additionally, through the use of advanced statistical modeling techniques, various direct and indirect relations were evaluated simultaneously among the relevant variables. Although the current study may help elucidate the relations between family organization and metabolic control, then, there are a few limitations to note. First, the data are predominantly cross-sectional and cannot determine causality; it is possible that the direction of the studied relationships could be reversed or be considered bidirectional. Longitudinal assessment will be required in future research to address these issues. Second, in this study, the FES Organization subscale evidenced lower than desirable reliability when parent and youth report were considered separately; when parent and youth report were combined, however, internal consistency was found to be adequate. Additionally, the lower reliability in the current sample seems to be in line with that found by other researchers (Loveland-Cherry et al., 1989). Furthermore, as the model was analyzed using structural equation modeling in Mplus 6 (Muthen & Muthen, 1998–2010), the error variance due to reporter bias was removed and a more stable true score, as represented by the latent variable, was used in analyses (Kline, 2011; Llabre, 2010). Nevertheless, future research should consider inclusion of alternative measures of family organization to create a latent variable of this important construct. Diabetes-specific measures of family organization could also be valuable to develop in future research as the current study assessed general organization.
Strengths and Limitations for Future Directions
Strengths of the present study include the use of multimethod, multi-informant data collection procedures, including parent and adolescent report of family organization, youth, and parent self-efficacy for diabetes and diabetes management, as well as a biological measure of metabolic control collected at two time points. Additionally, through the use of advanced statistical modeling techniques, various direct and indirect relations were evaluated simultaneously among the relevant variables. Although the current study may help elucidate the relations between family organization and metabolic control, then, there are a few limitations to note. First, the data are predominantly cross-sectional and cannot determine causality; it is possible that the direction of the studied relationships could be reversed or be considered bidirectional. Longitudinal assessment will be required in future research to address these issues. Second, in this study, the FES Organization subscale evidenced lower than desirable reliability when parent and youth report were considered separately; when parent and youth report were combined, however, internal consistency was found to be adequate. Additionally, the lower reliability in the current sample seems to be in line with that found by other researchers (Loveland-Cherry et al., 1989). Furthermore, as the model was analyzed using structural equation modeling in Mplus 6 (Muthen & Muthen, 1998–2010), the error variance due to reporter bias was removed and a more stable true score, as represented by the latent variable, was used in analyses (Kline, 2011; Llabre, 2010). Nevertheless, future research should consider inclusion of alternative measures of family organization to create a latent variable of this important construct. Diabetes-specific measures of family organization could also be valuable to develop in future research as the current study assessed general organization.