Abstract
Recent evidence suggests that ideational flexibility shares a significant amount of variance with ideational fluency, and more importantly, that the unique variance of flexibility is unreliable and invalid in terms of inter-item and intertest correlations. However, these findings have been demonstrated only with nongifted individuals. The present study used partial-correlation procedures to evaluate the unique variance of flexibility in the divergent thinking of 230 gifted and nongifted children. Four groups (quartiles of a distribution of composite achievement test scores) were compared for reliability and convergent validity of their flexibility scores. Results confirmed there were significant differences among the four groups and that the unique variance of flexibility was only reliable and valid for the children in the quartile high in achievement.