Differences in Responses to Computer-Based Color Symbol Conjoint Analysis Among Patients with Visual Impairment

Purpose: Conjoint analysis is a validated technique used to elicit preferences regarding relative importance of different choice attributes. This study evaluates variability of response time within a population of patients with visual impairment to test for significant differences in potential burden based on age and gender and to describe how the potential burden is related to the number of choices posed. Methods: Fifty-three patients with visual impairment were asked to indicate the most preferred alternative in a series of twelve choices regarding hypothetical rehabilitation regimens. Attributes of the regimens included the outcomes of difficulty with reading, transportation, and television viewing, the number of sessions, and personal expenses. An interviewer read a description of each attribute level for each choice while respondents saw a screen with two columns with colored bars. After data distribution examination, a generalized linear regression with log link (Poisson family) assessed whether sex, age, and the question number were associated with response time. Results: Average response time for all respondents was 21.6 seconds (SD 22.58). Men had shorter response times than women on average, however, men demonstrated a larger maximum response time and a wider range of response times than did women. Later questions took less time but the time required leveled off. Older adults took longer to respond. Conclusions: A significant response time difference exists between types of respondents. Future efforts to use conjoint analysis to elicit patient preferences for rehabilitation outcomes should be aware of the potential difference in burden for these groups of respondents.