Abstract
A systematic analysis and comparison was conducted of the psychometric properties of standard (paper-and-pencil) and computerized sociometry and peer assessments. The standard assessment took place with 817 children (47.5% boys) in 34 fifth- and sixth-grade classrooms of 28 elementary schools. The computerized assessment took place with 893 children (52.3% boys) in 35 fifth- and sixth-grade classrooms of 30 elementary schools. The computerized method yielded more nominations given and received and fewer missing data. Correlations among behavioural and relational constructs followed similar patterns in both methods, but were generally stronger in the computerized method. Voter agreement was the same in both methods. The computerized method resulted in higher internal consistency of sociometric and peer assessment constructs. The methods were validated by predicting preference and popularity from behavioural and relational characteristics. The two methods yielded the same pattern of results. The advantages and disadvantages of both assessment methods were discussed.
