Objective: The objective of this study was to develop a revision of the Symptoms of Stress Inventory (SOSI), the Calgary SOSI (C-SOSI), which would be shorter, easier to administer and score, and have excellent factor structure, reliability, and validity for use with oncology patients.
Methods: The SOSI was administered to 344 cancer patients registered for a stress-management program. Exploratory factor analysis (EFA) was applied using three criteria based on communality, factor loading, and desired subscale size. Scores on the revised C-SOSI were correlated with scores on measures of quality of life, mood disturbance, sleep, and spirituality to begin investigation of convergent and discriminant validity.
Results: The EFA resulted in a 56-item scale (down from the original 94 items) with 8 subscales, each consisting of 6-9 items named: Depression, Anger, Muscle Tension, Cardiopulmonary Arousal, Sympathetic Arousal, Neurological/GI, Cognitive Disorganization, and Upper Respiratory Symptoms. Cronbach's alpha reliabilities for the subscales ranged from 0.80 to 0.95. Convergent and discriminant validity was supported by correlations with other measures as conceptually predicted.
Conclusions: The C-SOSI is a reliable tool with converging validity for assessing stress symptoms in an oncology population. Further validation work is recommended to support use in other patient or community groups.