Differential item functioning impact in a modified version of the Roland–Morris Disability Questionnaire
ABSTRACT To evaluate a modified version of the Roland-Morris Disability Questionnaire for differential item functioning (DIF) related to several covariates.
DIF occurs in an item when, after controlling for the underlying trait measured by the test, the probability of endorsing the item varies across groups.
Secondary data analysis of two studies of participants with back pain (total n = 875). We used a hybrid item response theory/ logistic regression approach for detecting DIF. We obtained scores that accounted for DIF. We evaluated the impact of DIF on individual and group scores, and compared scores that ignored or accounted for DIF in terms of the strength of association with SF-36 subscale scores.
DIF was found in 18/23 items. Salient scale-level differential functioning was found related to age, education, and employment. Overall 24 participants (3%) had salient scale-level differential functioning. Mean scores across demographic groups differed minimally when accounting for DIF. The strength of association of scores with SF-36 scores was similar for scores that ignored and scores that accounted for DIF.
The modified version of the Roland-Morris Disability Questionnaire appears to have largely negligible DIF related to the covariates assessed here.
[Show abstract] [Hide abstract]
ABSTRACT: Many constructs are measured using multi-item data collection instruments. Differential item functioning (DIF) occurs when construct-irrelevant covariates interfere with the relationship between construct levels and item responses. DIF assessment is an active area of research, and several techniques are available to identify and account for DIF in cross-sectional settings. Many studies include data collected from individuals over time; yet appropriate methods for identifying and accounting for items with DIF in these settings are not widely available. We present an approach to this problem and apply it to longitudinal Modified Mini-Mental State Examination (3MS) data from English speakers in the Canadian Study of Health and Aging. We analyzed 3MS items for DIF with respect to sex, birth cohort and education. First, we focused on cross-sectional data from a subset of Canadian Study of Health and Aging participants who had complete data at all three data collection periods. We performed cross-sectional DIF analyses at each time point using an iterative hybrid ordinal logistic regression/item response theory (OLR/IRT) framework. We found that item-level findings differed at the three time points. We then developed and applied an approach to detecting and accounting for DIF using longitudinal data in which covariation within individuals over time is accounted for by clustering on person. We applied this approach to data for the "entire" dataset of English speaking participants including people who later dropped out or died. Accounting for longitudinal DIF modestly attenuated differences between groups defined by educational attainment. We conclude with a discussion of further directions for this line of research.Psychological Test and Assessment Modeling 04/2013; 55(2):127-147.
[Show abstract] [Hide abstract]
ABSTRACT: To assess psychometric properties of the Western Ontario and MacMasters Universities Osteoarthritis Index (WOMAC)-pf, an osteoarthritis (OA)-specific questionnaire used to establish the level of physical functioning in patients with late-onset sequels of poliomyelitis (LOSP) and Parkinson's disease (PD). Unidimensionality (using principal component analyses [PCA] and Rasch analyses) was separately established for three diagnostic groups: OA (n = 288), LOSP (n = 168) and PD (n = 200). Additionally, differential item functioning (DIF) among the three diagnostic groups was assessed using ordinal regression (Polytomous Universal Model) analyses. Baseline data were used from an ongoing cohort study of these three patient populations. Unidimensionality was adequate, with all items loading on the first principal component. The Rasch analyses revealed that item fit was generally good. Uniform and non-uniform DIF were found to be present among the three diagnostic groups in three and one of the 17 physical functioning subscale (WOMAC-pf) items, respectively. The WOMAC-pf is a unidimensional measure of physical functioning in patients with LOSP and PD, in addition to its established use in OA. [IMPLICATIONS FOR REHABILITATION:• Disability in physical functioning related to mobility(walking, stair climbing, etc.) is a common feature of many chronic diseases, including osteoarthritis, late-onset sequels of poliomyelitis and Parkinson's disease.• In this study, the Western Ontario and MacMasters Universities Osteoarthritis Index-pf was shown to bea useful and adequate tool to assess physical functioning in these patient groups.• The study highlighted that, despite differences in diagnosis,history and course of the disease, patients with different conditions experience similar disabilities in their physical functioning.]Disability and Rehabilitation 02/2012; 34(17):1423-30. DOI:10.3109/09638288.2011.645110 · 1.84 Impact Factor
[Show abstract] [Hide abstract]
ABSTRACT: Differential item functioning (DIF) occurs when a test item has different statistical properties in subgroups, controlling for the underlying ability measured by the test. DIF assessment is necessary when evaluating measurement bias in tests used across different language groups. However, other factors such as educational attainment can differ across language groups, and DIF due to these other factors may also exist. How to conduct DIF analyses in the presence of multiple, correlated factors remains largely unexplored. This study assessed DIF related to Spanish versus English language in a 44-item object naming test. Data come from a community-based sample of 1,755 Spanish- and English-speaking older adults. We compared simultaneous accounting, a new strategy for handling differences in educational attainment across language groups, with existing methods. Compared to other methods, simultaneously accounting for language- and education-related DIF yielded salient differences in some object naming scores, particularly for Spanish speakers with at least 9 years of education. Accounting for factors that vary across language groups can be important when assessing language DIF. The use of simultaneous accounting will be relevant to other cross-cultural studies in cognition and in other fields, including health-related quality of life.04/2011; 2(1):19-25. DOI:10.4081/ar.2011.e4