Psychometric analysis of the Oral Health Behavior Social Support Short Form scales (OHBSS-SF) among Mexican-American adults

Background Social support is an important psychosocial resource that can facilitate engaging in healthy behaviors, including oral health behaviors. Nine scales assessing Oral Health Behavior Social Support (OHBSS) for brushing, flossing and seeking dental care were developed and validated in the English and Spanish languages. The full-length scales included 37 required items to assess social support from three sources: family, health providers, and others/friends. The OHBSS-SF scales cover three oral health behaviors, use six items each to assess social support for brushing, flossing and dental care, and retain the nine-scale structure. Response options range from one (never) to four (always), with higher scores reflecting more frequent receipt of social support. This paper describes the item reduction process/outcomes and presents model fit indices and psychometric properties of the 17-item OHBSS-Short Form (OHBSS-SF) scales. Methods A large Mexican-American young adult sample completed an online survey in English or Spanish (N = 502). Participants completed the OHBSS-SF along with several other validated scales assessing general social support, dental fear, and acculturation. The survey also queried oral health behaviors of interest and socio-demographic characteristics. A subset of participants (n = 56) also completed the OHBSS-SF items a second time two to six weeks later to assess test-retest reliability. Model fit indices and psychometric properties for each of the nine OHBSS-SF scales were calculated for the full sample and subgroups stratified by the three study design characteristics of language (English/Spanish), sex (male/female), and marital status (single/married). Results The OHBSS-SF scales correlated very highly with the full-length versions (r = 0.95–0.98), and all OHBSS-SF scales exhibited acceptable model fit indices. The OHBSS-SF demonstrated stability in the full sample and in Spanish, with good to excellent intra-class correlations, supporting test-retest reliability. Patterns of moderate correlations supported convergent validity with validated general social support measures. Patterns of weak correlations suggested divergent validity between OHBSS-SF and other non-social support measures. Conclusions The OHBSS-SF demonstrated acceptable psychometric properties. A few significant differences by subgroup characteristics were identified. The OHBSS-SF scales reduce respondent burden and are a useful tool to ascertain frequency of receipt of oral health behavior social support from family, health providers, and other/friend sources.