VALIDATION AND PSYCHOMETRIC EVALUATION OF A 5-ITEM MEASURE OF PERCEIVED SOCIAL SUPPORT
Author(s)
McCarrier KP1, Bushnell DM1, Martin ML1, Paczkowski R2, Nelson DR2, Buesching D31Health Research Associates, Inc., Mountlake Terrace, WA, USA, 2Eli Lilly and Company, Indianapolis, IN, USA, 3Eli Lilly and Company, Indianapolis , IN, USA
OBJECTIVES: Previous researchers (Ren, 1999) have used a 5-item shortened version of the Medical Outcomes Study Social Support Scale (MOS-SSS; Sherborne, 1991), but the measurement properties of this brief measure have not been established. We filled this gap by evaluating the psychometric properties of a 5-item perceived social support instrument (the SS-5) and testing equivalence between paper and web-based administration modes. METHODS: The SS-5 and other questionnaire measures were administered to a sample of adults recruited through newspaper and web-based advertisements in 8 US cities. Participants were randomized to complete the SS-5 on either paper or computerized format followed by a one-week retest. The intraclass correlation coefficient (ICC) was calculated to assess reproducibility of the SS-5 and equivalence between administration modes. Cronbach’s alpha was calculated to assess internal consistency. To assess known groups’ validity, differences in SS-5 scores between marital status/living arrangement groups and (separately), trichotomous groupings of a validated measure of dispositional optimism, the Life Orientation Test (LOT-R), were assessed with analysis of variance (ANOVA) models. RESULTS: Of the 258 participants that completed the baseline assessment, 251 (97%) completed the retest. Mean age of participants was 48.6 years, 61% were female, and 71% were Caucasian. Participants included 44% married or living with a partner; 24% had never been married. The mean SS-5 score was 17.3±4.9, and the ICC between paper and web-based was 0.89. Reproducibility was strong (ICC=0.92), and the instrument was internally consistent (alpha=0.88). The SS-5 significantly discriminated between participants living with a partner and those living alone (means: 19.1 vs. 15.9; p<0.001) and tertiles of the LOT-R. The correlation of SS-5 scores to the full MOS-SSS was 0.91. CONCLUSIONS: The SS-5 was observed to have adequate reproducibility and internal consistency; and demonstrated appropriate known groups validity. Equivalence between paper and web-based administration was demonstrated.
Conference/Value in Health Info
2011-05, ISPOR 2011, Baltimore, MD, USA
Value in Health, Vol. 14, No. 3 (May 2011)
Code
PRM16
Topic
Patient-Centered Research
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
Multiple Diseases