Improving Reliability of Peer-Nomination with Young Children

Abstract

It is now well recognized that poorly accepted children run a greater risk of serious adjustment problems later in life. The poor-to-fair temporal reliability of the peer-nominations found among young children was an obstacle to early assessment. Although two reliable alternative techniques have been proposed, these have important drawbacks. The present study assessed the reliability of a third alternative, namely, asking children to make more nominations. Three to six positive and negative nominations were obtained from 97 kindergarten and 100 prekindergarten children. Although data proved unreliable at the prekindergarten level, data from kindergarten children were quite reliable and indicated that four choices yielded higher reliabilities than three-choice nominations. More nominations did not further improve reliability. Percentages of subjects classified in the various sociometric status following Coie and Dodge's 1983 criteria also suggested that four choices reproduced optimally the classification generated by the traditional nomination technique. Implications of these results are discussed.