We (Hardy, Woodman, & Carrington, 2004) recently proposed an innovative segmental quadrant analysis for exploring the role of self-confidence within a higher-order catastrophe model framework. Using this exploratory analysis, we found initial support for the main hypothesis, namely that the maximum interaction effect size between cognitive anxiety and somatic anxiety would be located at a lower level along the somatic anxiety continuum for conditions of low self-Cong compared to conditions of high self-confidence. In the present issue of this journal, Tenenbaum and Becker (2005) offer a critique of this study. In formulating their critique they have employed four principal approaches: (a) a largely indiscriminate critique of catastrophe model research as a whole; (b) a more specific critique of the method and analysis employed in our study; (c) a misrepresentation of our own work and that of previous authors; and (d) abundant confusion and irrelevancy. We address each of these issues in turn.
The authors are with the School of Sport, Health, and Exercise Sciences, George Bldg., Holyhead Road, University of Wales, Bangor, Gwynedd LL57 2DG, U.K.