When is an agreement between two methods for assessing coping styles "good''? A reply to Mendolia

In her comment on our article (Egloff & Hock, 1997) on the association of coping style classifications based on (a) dispositional vigilance and cognitive avoidance and (b) trait anxiety and social desirability, Mendolia (1999) argued that our data demonstrated only weak agreement between both assessment approaches. In this reply, we demonstrate by crosstabulations of the dichotomized split-half scales that even Mendolia's criteria of "fair to good" interrater agreement between the methods (Cohen's kappa > 0.40) is not reached within each assessment approach alone. This result may be due to unreliability of measurement and problems inherent in median split procedures. (C) 1999 Elsevier Science Ltd. All rights reserved.