Purchase this article with an account.
Olukemi Adeyemo, Pamela E Jeter, Duane Geruschat, Gislin Dagnelie; Comparing the item measures of an Ultra-low vision (ULV) questionnaire with those of a ULV assessment toolkit.. Invest. Ophthalmol. Vis. Sci. 2016;57(12):1979.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare the order and distribution of the item measures in an ultra-low vision (ULV) questionnaire with those of the corresponding tasks in a ULV assessment toolkit.
The item measures of a 23-item ULV questionnaire (ULV-23) measured in an 80-subject sample were calculated using Winsteps©. The item measures for the 17-item ULV assessment toolkit (ULV-ADL) measured in a separate 24-subject sample were calculated from the percent correct scores of the activities. The order and the item-measure distribution of the 17 items common to both instruments were compared. Some of the tasks on the ULV-ADL were modified to allow more replicable scenarios to be generated, this is expected to be reflected in minor shifts in item ranking.
The item-measure distribution of the ULV questionnaire items ranged from -2.50 (less difficult to complete) to +2.01(more difficult to complete). The item measures for the ULV assessment toolkit ranged from -2.26 to +1.81. Both distribution have good centering of the measures. 23.5% of the ULV-ADL items were ranked less difficult than the corresponding ULV-23 items, while 53% of the items were ranked as more difficult. The correlation of the two item measure distributions is 0.78. Expected slight shifts in ranking of items were observed but the distribution of item measures of the ULV-ADL were still within the same range as that of the ULV-23; showing the overall rank order of items in both instruments are comparable.
Despite being collected in different ULV subject samples, and despite necessary modifications of the ULV-23 items to create the corresponding ULV-ADL tasks, the distributions are highly correlated. This confirms that the ULV-23 may act as a proxy to the ULV-ADL, and vice versa. This may be particularly useful in low vision clinics and clinical trials with ULV participants, where time constraints may not allow administration of both instruments.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
This PDF is available to Subscribers Only