TY - JOUR
T1 - The refractive status and vision profile
T2 - Rasch analysis of subscale validity
AU - Gothwal, Vijaya
AU - Wright, Thomas
AU - Elliott, David
AU - Pesudovs, Konrad
PY - 2010/11/1
Y1 - 2010/11/1
N2 - PURPOSE: To determine whether the eight subscales of the Refractive Status and Vision Profile (RSVP) questionnaire provide valid measurement. METHODS: Two hundred patients, recruited from a refractive surgery clinic and general optometric practice, self-administered the 42-item RSVP questionnaire. Psychometric properties investigated included measurement of a single construct (unidimensionality), item fit to construct, measurement precision (represented by person separation), targeting of item difficulty to patient's quality of life (QOL), and differential item functioning. RESULTS: Only two subscales ("concern" and "driving") showed adequate person separation, indicating that they could discriminate patients' QOL. However, both demonstrated poor targeting (7.02 logits for "driving" and 1.11 logits for "concern"). One-third of items in each subscale suffered from differential item functioning. CONCLUSIONS: None of the RSVP subscales are valid for assessment of QOL in patients with refractive error, thereby indicating the RSVP should be considered as a single questionnaire without subscales.
AB - PURPOSE: To determine whether the eight subscales of the Refractive Status and Vision Profile (RSVP) questionnaire provide valid measurement. METHODS: Two hundred patients, recruited from a refractive surgery clinic and general optometric practice, self-administered the 42-item RSVP questionnaire. Psychometric properties investigated included measurement of a single construct (unidimensionality), item fit to construct, measurement precision (represented by person separation), targeting of item difficulty to patient's quality of life (QOL), and differential item functioning. RESULTS: Only two subscales ("concern" and "driving") showed adequate person separation, indicating that they could discriminate patients' QOL. However, both demonstrated poor targeting (7.02 logits for "driving" and 1.11 logits for "concern"). One-third of items in each subscale suffered from differential item functioning. CONCLUSIONS: None of the RSVP subscales are valid for assessment of QOL in patients with refractive error, thereby indicating the RSVP should be considered as a single questionnaire without subscales.
UR - http://www.scopus.com/inward/record.url?scp=78449242192&partnerID=8YFLogxK
U2 - 10.3928/1081597X-20100512-01
DO - 10.3928/1081597X-20100512-01
M3 - Article
VL - 26
SP - 912
EP - 915
JO - Journal of Refractive Surgery
JF - Journal of Refractive Surgery
SN - 1081-597X
IS - 11
ER -