Predicting state anxiety: A comparison of multidimensional and unidimensional trait approaches

Endler has proposed that trait anxiety is multidimensional, with individual differences varying across several dimensions of stressing situations. Despite empirical support for the independence of some dimensions, some researchers have employed a single score on a multidimensional measure of trait anxiety, summed from the dimension scale scores, as a singular measure of generalized trait anxiety. This violates the theoretical underpinnings of the approach. This study examined whether a total score on a multidimensional measure of trait anxiety could supplement the ability of individual dimension scale scores in predicting state anxiety responses to dimension-congruent stress. Thirty-two female college students who scored at differing levels of dimensional and general anxiety were required to perform two laboratory stress tasks under conditions of stress representing different dimensions of trait anxiety. The results provide support for a multidimensional approach to general trait anxiety but indicate some value of a total score in supplementing predictions based on individual scale scores. The data provide a cautionary note that a total score on a multidimensional measure is of little value unless full consideration is given to the profile of dimension scale scores.