Differential Item Functioning and Person Fit on the PROMIS Pediatric Physical Functioning Items (303142)

Keywords: person fit, differential item functioning, item response theory

The Patient Reported Outcomes Measurement Information System (PROMIS) pediatric physical functioning item banks consist of 29 upper extremity items and 23 mobility items. A sample of 5,091 children and adolescents (mean age = 12 years old, range: 8--17; 49% male) completed the English-language version of the items. A sample of 605 children and adolescents (mean age = 12 years old, range: 8--17; 55% male; 96% Hispanic) completed the Spanish-language version of the items.

Differential item functioning (DIF) between two groups is present if the probability of selecting a particular response varies by group when controlling for the underlying level of the concept being measured. Person fit evaluates the extent to which an individual's pattern of item responses is consistent with the model. We evaluate language DIF for the PROMIS upper extremity and mobility item banks.

We found language DIF for four upper extremity items and eight mobility items. After excluding cases with significant person misfit, we found DIF for three of the four upper extremity items with DIF in the full sample and for nine mobility items (including seven of the items that had DIF in the full sample).

Excluding respondents displaying person misfit was expected to reduce DIF between English and Spanish-language responses to the item banks, but DIF was not reduced. Person misfit indicates a pattern of responding that differs from that of the typical respondent. However, if those with misfit have a variety of different response patterns that deviate from the model, then excluding those with aberrant responses would not necessarily translate into reducing DIF by a subgroup characteristic (in this case, survey language). Future analyses will examine the variability in patterns of person misfit to help elucidate the lack of impact on DIF in this study. In addition, we will explore the extent to which items with DIF differ from non-DIF items (e.g., readability level).