Abstract
Although secondary analyses have been spotlighted in Japan in recent years, most researches consist of analyses of available datasets. This paper analyzes data available only in the form of published figures and tables. We used the figures and tables from Hunter's book, “Community Power Structure (1953),” from which we tried to reconstruct network survey data. These figures and tables, however, contained inconsistencies. We then applied a unique restoration method that accommodated those inconsistent cases. We analyzed the restored data to verify Hunter's proposition that individuals' choices of others are not random. Hunter originally showed this by examining the form of the figures. Our analysis using the restored data, on the other hand, verified it by testing it statistically. In addition, we examined whether the observed tendency in choosing others could occur by chance. Our results indicated that such possibility is low.
