Stability of Sociometric Responses in Classrooms

Abstract
Since most studies which employ sociometric measures have either ignored the sensitivity of these measures to changes in group structure or have failed to indicate the relative sensitivity of the different types of measures across time and at different age levels, a way to categorize sociometric responses according to the usefulness in reflecting changes both across time and at various age levels was hypothesized. A total of 607 students in grades four through eight responded to a sociometric questionnaire. These responses were analyzed by obtaining test-retest correlations for choices given, choices received, and reciprocal choices. The results indicate that reciprocal choices are less affected by age and time changes than all other ways of interpreting sociometric data.