Purchase this article with an account.
Gislin Dagnelie, Meesa Maeng; Independent recalibration of the 150 item ultra-low vision visual functioning questionnaire (ULV-VFQ). Invest. Ophthalmol. Vis. Sci. 2020;61(7):2668.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To obtain a comparative re-calibration of the ULV-VFW questionnaires by an independent assessor in a new population of individuals meeting the criteria for ultra-low vision (logMAR ≥ 1.9).
Clients and staff at the Chicago Lighthouse (CLH) were invited to join the study. All participants were interviewed by a single assessor (MM), to assure uniform administration. Results were Rasch-analyzed with Winsteps 3.92 and examined for dimensionality, DIF according to gender and demographic variables, and person and item measure Infit characteristics. Results were compared to the previous analysis based on an 80-person sample (Legacy).
100 participants (57% female, age range 21 – 97; median 58), with best corrected vision ranging from bare light perception (LP) to 20/1200 in the better eye were enrolled. Estimated median logMAR acuity in the better eye was 2.2. Responses on a 4-level scale (not difficult – impossible) or "not applicable" (=missing item) were Rasch analyzed with Winsteps 3.92. Reliability of the analysis was 0.99 for persons, 0.97 for items. The median and mean/SD person measures were -0.66 and -1.04±2.29 (CLH) vs. -0.67 and -0.86±2.13 (Legacy). Item measures showed a good range, but 21% of CLH person measures (vs. 15% Legacy) fell below the item measure range. 14 CLH person measures and 12 item measures showed infit z-scores > 3 SE, indicating underfit, as did 7 Legacy person measures. DIF analysis showed that up to 10% of ULV-VFQ items may need to be examined for ambiguity or deviation from unidimensionality.
The ULV-VFQ performed very well in this independent sample, but infit problems in both person and item measures warrant further investigation. We are continuing to collect data from an additional 100-person sample.
This is a 2020 ARVO Annual Meeting abstract.
This PDF is available to Subscribers Only