Testing compensatory health beliefs in a UK population

Abstract

Compensatory health beliefs, beliefs that healthy behaviours can compensate or neutralise unhealthy behaviours, have been proposed as one way of understanding why people engage in health-risk behaviours (Knäuper, B., Rabiau, M., Cohen, O., & Patriciu, N. (2004). Compensatory health beliefs scale development and psychometric properties. Psychology and Health, 19, 607–624). However, measuring compensatory health beliefs has proved a challenge, with several recent studies being unable to replicate the psychometric properties of Knäuper et al.'s (2004) scales. The aims of this study were to: (1) test the factor structure of the compensatory health beliefs scale in the UK, (2) examine the predictive validity of the scale by testing the relationships between compensatory health beliefs and health behaviours over a six-month time interval and (3) assess the 6-month test–retest reliability of the scale. A total of 393 participants completed measures of compensatory health beliefs and health behaviours at two time points separated by six months. The findings were potentially problematic for research into compensatory health beliefs: the factor structure was not confirmed, there was little evidence of predictive validity, and test–retest reliability was poor. Further research is required to understand the operation of compensatory health beliefs and to develop the measurement of compensatory health beliefs.

Keywords:

• compensatory health beliefs
• health behaviours
• smoking
• exercise
• fruit and vegetable intake

Acknowledgements

A special thanks to Bärbel Knäuper for providing us with the 40-item compensatory health beliefs questionnaire. Also, thank you very much to Theda Radtke for providing us with an early copy of her paper. As well as to all the participants that took part in the online studies.

Notes

Note

1. The analyses were repeated without the students’ data, but this made no difference to the findings. Furthermore, more principal factor axis analyses were reiterated with an orthogonal rotation, similarly this made no difference to the findings.