Metaphors of Distance, Size and Temperature in Sociometry of Small Social Groups: A Generalizability Theory Approach

We have developed a new interactive instrument that uses psychological distance (represented by visual distance) together with the metaphors of size and temperature to assess liking and influence among primary school students. To test its concurrent validity, we compare its results with an established sociometric instrument. Furthermore, we examine the characteristics of the obtained ratings, the main sources of variability of the observed ratings, and the reliability of person-level results. Our results indicate that the ratings derived from the interactive visual metaphors of distance, size, and temperature show adequate reliability and strongly correlate with a traditional sociometric inventory that uses rating scales. The novelty of our approach lies in the use of generalizability theory, which allows for the decomposition of obtained ratings into class, rater, rated person, dyadic, and idiosyncratic effects. This allows for a more nuanced analysis of the relationships between members of social groups, such as school classes.