TY - JOUR
T1 - Differential Performance of SF-36 Items in Healthy Adults With and Without Functional Limitations
AU - Horner-Johnson, Willi
AU - Krahn, Gloria L.
AU - Suzuki, Rie
AU - Peterson, Jana J.
AU - Roid, Gale
AU - Hall, Trevor
PY - 2010/4/1
Y1 - 2010/4/1
N2 - Horner-Johnson W, Krahn GL, Suzuki R, Peterson JJ, Roid G, Hall T, the RRTC Expert Panel on Health Measurement. Differential performance of SF-36 items in healthy adults with and without functional limitations. Objective: To determine whether Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) items show differential item functioning among healthy adults with various types of functional limitations as compared with a healthy sample with no identified limitations. Design: Survey responses were analyzed by using partial correlations. Setting: General community. Participants: Participants (N=206) included (1) adults with spinal cord injury (SCI), (2) adults who were deaf or hard of hearing, (3) adults who were legally blind, (4) adults with psychiatric or emotional conditions, and (5) adults with no reported functional limitations. Participants were screened to ensure the absence of substantial health problems. Interventions: Not applicable. Main Outcome Measure: SF-36. Results: Partial correlations showed a significant negative correlation, indicating differential item functioning (ie, apparent bias) for people with SCI on all 10 SF-36 Physical Functioning items. For people who were blind, 5 items showed a significant negative correlation. Two items had significant negative correlations for the deaf/hard-of-hearing group. One item showed significant negative performance for people with mental health conditions. Conclusions: Our data indicated a possibility for measurement bias caused by the blending of health and function concepts in the SF-36.
AB - Horner-Johnson W, Krahn GL, Suzuki R, Peterson JJ, Roid G, Hall T, the RRTC Expert Panel on Health Measurement. Differential performance of SF-36 items in healthy adults with and without functional limitations. Objective: To determine whether Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) items show differential item functioning among healthy adults with various types of functional limitations as compared with a healthy sample with no identified limitations. Design: Survey responses were analyzed by using partial correlations. Setting: General community. Participants: Participants (N=206) included (1) adults with spinal cord injury (SCI), (2) adults who were deaf or hard of hearing, (3) adults who were legally blind, (4) adults with psychiatric or emotional conditions, and (5) adults with no reported functional limitations. Participants were screened to ensure the absence of substantial health problems. Interventions: Not applicable. Main Outcome Measure: SF-36. Results: Partial correlations showed a significant negative correlation, indicating differential item functioning (ie, apparent bias) for people with SCI on all 10 SF-36 Physical Functioning items. For people who were blind, 5 items showed a significant negative correlation. Two items had significant negative correlations for the deaf/hard-of-hearing group. One item showed significant negative performance for people with mental health conditions. Conclusions: Our data indicated a possibility for measurement bias caused by the blending of health and function concepts in the SF-36.
KW - Bias
KW - Disabled persons
KW - Health surveys
KW - Rehabilitation
UR - http://www.scopus.com/inward/record.url?scp=77950491739&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77950491739&partnerID=8YFLogxK
U2 - 10.1016/j.apmr.2009.12.015
DO - 10.1016/j.apmr.2009.12.015
M3 - Article
C2 - 20382289
AN - SCOPUS:77950491739
VL - 91
SP - 570
EP - 575
JO - Archives of Physical Medicine and Rehabilitation
JF - Archives of Physical Medicine and Rehabilitation
SN - 0003-9993
IS - 4
ER -