Development and Psychometric Properties of the Solitude Behavior Scale

Objective: To develop a Solitude Behavior Scale and examine its reliabilities and validities in this study.Methods: The scale consisted of Positive Solitude,Seclusiveness,Social Avoidance and Loneliness subscales.College students(N=534),high-school students(N=252),schizophrenic patients(N=55) and controls(N=55) were administrated.Results: ①The discriminations of most items were satisfactory(above 0.3).②The Cronbach α of the four subscales were between 0.762 and 0.863;the stabilites of the four subscales were from 0.654 to 0.760.③The result of EFA for college sample showed the four-factor model explained 44.25% variance and the construct of factors was very distinct.The result of CFA for high school students sample showed x2/df=1.79,RMSEA=0.05,NFI=0.87,NNFI=0.93,CFI=0.94.④There were significantly positive correlations(r from 0.318 to 0.821) among Seclusiveness,Social Avoidance and Loneliness subscales with Social Avoidance,Distress Scale and UCLA Loneliness Scale.⑤Compared with the control sample,the schizophrenic patients obtained significantly lower scores in Positive Solitude subscale,but significantly higher scores in the three subscales of Seclusiveness,Social Avoidance and Soneliness.Conclusion: The Solitude Behavior Scale has good psychometric properties.