Criterion Problems in Rehabilitation Outcome Research: A Comment on Allison and Stephens (2004) Academic Article uri icon


  • The recent article by Allison and Stephens, ‘The assessment of pain beliefs and their role in predicting recovery from repetitive strain injury’, addressed an important issue for both rehabilitation counsellor practitioners as well as for rehabilitation researchers — the role of people's beliefs in influencing their behaviours and achievements while living with chronic conditions or recovering from traumatic injuries (Radnitz, 2000). Using chronic pain patients as study participants, the authors assessed pain beliefs and then correlated belief scores with three indices of ‘recovery’ (function improvement, pain decrease and total improvement). While the study as reported had a number of strengths, there were two aspects of the published article which concerned me as an academic who is heavily involved in teaching and conducting research in the area of understanding factors that influence rehabilitation ‘recovery’. One of my concerns has to do with the design of the study, and my other concern pertains to the conclusions drawn by the authors following the analysis of the (minimal) relationships between the study's key pain and functioning constructs. Both of these concerns are elaborated upon in the following paragraphs.My concern with the design of the study is only important because of the title of the paper and the conclusions that the low obtained correlations seemingly precipitated. The title of Allison and Stephens' paper contains the phrase ‘…predicting recovery …’. Yet the criterion measures used in the study are open to extreme criticism as adequate measures of ‘recovery’. Identifying appropriate criteria in applied research is rarely easy (Ronan & Prien, 1966; Smith, 1976) but ignoring traditional observable measures of rehabilitation progress (such as enhanced self management and increased assumption of normal role responsibilities) means that rehabilitation counsellors reading the article can learn little about the influence of pain beliefs on the reeular behaviours of their clients.

publication date

  • January 2005