Browsing by Person "Smith, Adam B."

Now showing 1 - 3 of 3

Rasch analysis of the dimensional structure of the Hospital Anxiety and Depression Scale
(2006-12-14) Smith, Adam B.; Wright, E.P; Rush, Robert; Stark, Dan; Velikova, Galina; Selby, P.J.
The Hospital Anxiety and Depression Scale (HADS) has been used extensively in cancer patients to identify psychological distress. Reports of the factor structure and screening performance of the instrument vary. Rasch models allow an assessment of the structure of a questionnaire by identifying item fit. Removal of misfitting items may improve both the dimensionality and efficacy of screening questionnaires. A Rasch analysis of the HADS-T and subscales was used to explore the factor structure, dimensionality and screening efficacy. A total of 1855 patients completed a touchscreen version of the HADS, including 381 patients who had received a psychiatric interview (SCAN/PSE). These data were analysed using Rasch models, and the screening efficacy at identifying cases of psychological distress and anxiety and depression evaluated. The results demonstrated that the structure of the HADS-T and subscales was unidimensional. Three items from the HADS-T, and one from each of the subscales demonstrated misfit. Screening efficacy for the HADS-T and subscales was modest. However, removal of misfitting items had little impact on screening, demonstrating that items could potentially be omitted, if required. The item range covered a narrow spectrum of psychological distress, predominantly higher levels of distress. Additional items have to be added if screening for moderate to mild distress is to be improved for cancer patients.
Rasch fit statistics and sample size considerations for polytomous data
(2008-05-29) Smith, Adam B.; Rush, Robert; Fallowfield, Lesley J.; Velikova, Galina; Sharpe, M.
Background: Previous research on educational data has demonstrated that Rasch fit statistics (mean squares and t-statistics) are highly susceptible to sample size variation for dichotomously scored rating data, although little is known about this relationship for polytomous data. These statistics help inform researchers about how well items fit to a unidimensional latent trait, and are an important adjunct to modern psychometrics. Given the increasing use of Rasch models in health research the purpose of this study was therefore to explore the relationship between fit statistics and sample size for polytomous data. Methods: Data were collated from a heterogeneous sample of cancer patients (n = 4072) who had completed both the Patient Health Questionnaire – 9 and the Hospital Anxiety and Depression Scale. Ten samples were drawn with replacement for each of eight sample sizes (n = 25 to n = 3200). The Rating and Partial Credit Models were applied and the mean square and t-fit statistics (infit/outfit) derived for each model. Results: The results demonstrated that t-statistics were highly sensitive to sample size, whereas mean square statistics remained relatively stable for polytomous data. Conclusion: It was concluded that mean square statistics were relatively independent of sample size for polytomous data and that misfit to the model could be identified using published recommended ranges.
The initial development of an item bank to assess and screen for psychological distress in cancer patients
(John Wiley & Sons, 2007-11-10) Smith, Adam B.; Rush, Robert; Velikova, Galina; Wall, L.
Psychological distress is a common problem among cancer patients. Despite the large number of instruments that have been developed to assess distress, their utility remains disappointing. This study aimed to use Rasch models to develop an item-bank which would provide the basis for better means of assessing psychological distress in cancer patients. An item bank was developed from eight psychological distress questionnaires using Rasch analysis to link common items. Items from the questionnaires were added iteratively with common items as anchor points and misfitting items (infit mean square >1.3) removed, and unidimensionality assessed. A total of 4914 patients completed the questionnaires providing an initial pool of 83 items. Twenty items were removed resulting in a final pool of 63 items. Good fit was demonstrated and no additional factor structure was evident from the residuals. However, there was little overlap between item locations and person measures, since items mainly targeted higher levels of distress. The Rasch analysis allowed items to be pooled and generated a unidimensional instrument for measuring psychological distress in cancer patients. Additional items are required to more accurately assess patients across the whole continuum of psychological distress