Item analysis of cancer patient responses to the Multidimensional Affect and Pain Survey demonstrates high inter-item consistency and discriminability and determines the content of a short form

Abstract

To construct a short form (SMAPS-CP), item analysis was used to select 30 items, 1 from each of the 30 subclusters in the 101-item Multidimensional Affect and Pain Survey (101-MAPS). Representation of each of the subclusters ensured that the structure of the MAPS dendrogram obtained by cluster analysis was maintained in SMAPS-CP. Responses of outpatients with cancer to the 101-MAPS were treated by item analysis to obtain measures of interitem consistency (criterion between .40 and .80) and discriminability (criterion above .35) for each of the 101-MAPS questions. Both of these criteria for acceptance were met by 53 of the 57 questions in Supercluster I, Somatosensory Pain, by 25 of the 26 questions in Supercluster II, Emotional Pain, and by all 18 of the questions in Supercluster III, Well-being. The item within each cluster that best met the item analysis criteria was selected for the 30-item SMAPS-CP questionnaire. Perspective Item analysis of responses by cancer patients to a 101-item questionnaire was used to develop a short questionnaire that can be used to identify and treat both physical and emotional levels of pain without the burden of a long questionnaire.