Development and psychometric testing of Hensarling's Diabetes Family Support Scale

The purpose of this study was to develop Hensarling's Diabetes Family Support Scale (HDFSS) as a valid and reliable instrument to specifically measure perceived family support of adults with type 2 diabetes. The sample included 158 participants with type 2 diabetes who responded to three instruments: General Information Form (GIF), HDFSS, and Social Support in Chronic Illness Inventory (SSCII).

Psychometric testing focused on estimating reliability (i.e., internal consistency) and several types of validity: content, criterion-related (concurrent and predictive), convergent, and construct validity. Testing also resulted in item reduction, from 29 to 24 items. The Content Validity Index (CVI) of the 24-item HDFSS was 1.00. The HDFSS demonstrated estimates supporting internal consistency with an item-item correlation mean of .52, item-total score correlations between .49 and .87, and Cronbach's alpha of .96. Concurrent validity was tested by examining the correlation of the total HDFSS score with a 1-item rating of "perceived overall family support" (on a scale of 0 to 10) using Kendall's tau (τ). The resulting correlation was moderate (.58, p = .01), as predicted. Predictive validity was tested by examining the correlation of the total HDFSS score with the self-reported Hemoglobin A1c (HbA1c), resulting in a low (τ = −.048, p = .45) correlation; though low, this correlation was in the predicted direction. Convergent validity examined the relationship between the scores on the HDFSS and a similar instrument, Social Support in Chronic Illness Inventory (SSCII), a measure of satisfaction with family support by persons with chronic illness. The resulting correlation was moderate (τ = .52), slightly below the moderately high correlation (τ ≥ .55) predicted. The HDFSS demonstrated internal construct validity through the emergence of four theoretically clear dimensions: (a) empathetic support, (b) encouragement, (c) facilitative support and (d) participative support.

Results demonstrate that the HDFSS is a valid and reliable instrument when used with this sample. Further testing is needed with other samples to determine whether findings replicate. An alternate approach to assessing predictive validity is suggested. The HDFSS should serve as an efficient and psychometrically sound tool for both clinical and research applications.