To investigate the psychometric properties of instruments for the assessment of self-reported functional vision performance and health-related quality of life
with intellectual disabilities (IDs).
Two instruments [Autoquestionnaire Enfant Image (AUQUEI), LV Prasad-Functional Vision Questionnaire
(LVP-FVQ)] designed for the assessment of functional vision and health-related quality of life
were adapted and administered to 168 school children
with ID, aged 8 to 18 years. Rasch analysis
was used to determine the appropriateness of the rating scales of these instruments and to identify any redundant items.
Redundant items were excluded based on descriptive statistics and Rasch analysis
, leaving 17 of 23 items in the revised AUQUEI and 16 of 22 in the LVP-FVQ. The AUQUEI items showed disordered thresholds on the rating scale. A modified step calibration (collapsed from four categories to three categories) resulted in ordered response thresholds for all items. The adjusted instrument produced an overall fit to the model (mean item infit = 1.06, SD = 0.32; mean item outfit = 1.11, SD = 0.35), indicating good construct validity. After Rasch analysis
, the AUQUEI showed good content validity (person separation = 2.18; item reliability = 0.99; Cronbach α = 0.89). Increased similarity of person and item means and SDs on the logit scale after modification would indicate that the instrument was more applicable to the target population in its modified form. In contrast, the LVP-FVQ had a low person separation (1.35), suggesting that a more appropriate instrument is needed for assessment of vision-related quality of life
The psychometric properties of two instruments were explored using Rasch analysis
. By rescaling and reduction of items, the instruments were modified for use in a population of children
with at least mild to moderate ID. However, an alternative instrument is needed for the assessment of vision-related quality of life
in intellectually disabled children
with normal vision or mild visual abnormalities.