Reflections on reflections... on the nature of intuition, analysis and the construct validity of the Cognitive Style Index

Hayes, Allinson, Hudson, and Keasey (2003) maintain that the theoretical arguments and empirical evidence presented by Hodgkinson and Sadler-Smith (2003) in support of an alternative two-dimensional conception of the Allinson-Hayes Cognitive Style Index (CSI) are insufficiently robust to challenge the original unidimensional formulation and that, accordingly, users of the CSI should continue to score the instrument as a single bipolar scale. In reply, we show that their reasoning is flawed, being based on a series of misinterpretations of the nature and purpose both of the item parcelling procedure and the exploratory factor analysis technique we employed. Furthermore, their defence fails to take account of the extensive confirmatory factor analytic evidence we presented in support of our conclusions.