OBJECTIVE: To develop a perceived social support scale for the Japanese. METHODS: Participants consisted of a total of 2,150 residents who responded to a perceived social support scale questionnaire in four communities involved in the Jichi Medical School Cohort Study at the study baseline in 1992-1994 (Analysis I) and 380 residents of Akaike town who responded to a questionnaire survey with the revised scale in 1999 (Analysis II). The first analysis was performed to confirm cross validity of the perceived social support scale measuring the availability of functional support from the spouse, family, and friends. The factor structures estimated for each district were compared with each other. The second analysis was set to revise the scale, modified in terms of a 4-point scoring format. The correlation between each item and a social desirability scale score, Cronbach alpha, and correlation coefficients between the revised scale scores and the original scale scores were calculated. RESULTS: Cross validity was confirmed based on the similarity of the factor structures of the responses obtained from the four different communities. Four selected indices supported the constancy of the factor structures across the communities. For the revised scale, two items were removed from the original scale-items measuring spousal support because of their statistically significant correlations with the social desirability scale score. The first principal component accounted for 57.0% of variance for spousal support, 68.2% for family support, and 67.0% for friends support and uni-dimensionality of the scales was confirmed. Alpha coefficients of the scales were .89, .95, and .94, respectively. Social support levels were stable over a mean period of 6.2 years. CONCLUSIONS: On an empirical basis, we have been able to develop a perceived social support scale. Homogeneous data from three different sources were obtained which were free from social desirability response bias and thus applicable for Japanese adults.
OBJECTIVE: To develop a perceived social support scale for the Japanese. METHODS:Participants consisted of a total of 2,150 residents who responded to a perceived social support scale questionnaire in four communities involved in the Jichi Medical School Cohort Study at the study baseline in 1992-1994 (Analysis I) and 380 residents of Akaike town who responded to a questionnaire survey with the revised scale in 1999 (Analysis II). The first analysis was performed to confirm cross validity of the perceived social support scale measuring the availability of functional support from the spouse, family, and friends. The factor structures estimated for each district were compared with each other. The second analysis was set to revise the scale, modified in terms of a 4-point scoring format. The correlation between each item and a social desirability scale score, Cronbach alpha, and correlation coefficients between the revised scale scores and the original scale scores were calculated. RESULTS: Cross validity was confirmed based on the similarity of the factor structures of the responses obtained from the four different communities. Four selected indices supported the constancy of the factor structures across the communities. For the revised scale, two items were removed from the original scale-items measuring spousal support because of their statistically significant correlations with the social desirability scale score. The first principal component accounted for 57.0% of variance for spousal support, 68.2% for family support, and 67.0% for friends support and uni-dimensionality of the scales was confirmed. Alpha coefficients of the scales were .89, .95, and .94, respectively. Social support levels were stable over a mean period of 6.2 years. CONCLUSIONS: On an empirical basis, we have been able to develop a perceived social support scale. Homogeneous data from three different sources were obtained which were free from social desirability response bias and thus applicable for Japanese adults.