Assessing Exercise Involvement among Participants in Health and Fitness Centres

Abstract

We tested the reproducibility of the factor structure and the predictive efficacy of the modified involvement scale (MIS) among 260 Greek-speaking exercise participants aged from 15 to 71 years. Confirmatory factor analysis did not clearly support the hypothesized five-factor structure. Subsequent exploratory factor analyses revealed three factors, labelled positive outlook, social bonding and identity expression. Linear regression analyses illustrated that the emergent solution was an effective predictor of other constructs that have demonstrated utility for leisure service providers (i.e., psychological commitment, satisfaction and behavioural loyalty). While the hypothesized factor structure of the MIS was not supported by these data, the emergent three-factor solution bore some resemblance to solutions emerging from prior work. The performance of the MIS within the Greek exercise context can be attributed to the incongruence in meaning that Greek exercise participants ascribe to the activity and item wording. Further work on item wording is required to better refine the MIS for Greek recreationists. None the less, this work adds to a growing literature documenting the utility of enduring involvement for understanding leisure behaviour and demonstrating its potential for identifying loyal consumers.

Notes

1. The direct oblimin method of rotation was chosen for two reasons. First, because the five dimensions of the MIS are hypothesized to reflect a single higher-order construct (i.e., enduring involvement), it is reasonable to assume that these dimensions will be correlated. Secondly, there is considerable evidence indicating moderate to high positive correlations among the dimensions. While few authors report the factor correlations in their findings, our own experience with earlier involvement scales, in addition to evidence provided in Havitz and Dimanche's (1997) review, provides support for this assumption.