Purchase this article with an account.
Michael B Bartlett, Gina Mitzel, Song Zhang, Yi-Zhong Wang; Comparison of myVisionTrack® Vision Monitor Performance with 3-Alternative Forced-Choice (3AFC) and 4AFC Testing Paradigms for Assessing Shape Discrimination Hyperacuity. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):3892. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The myVisionTrack® (mVTTM), a mobile shape discrimination hyperacuity (SDH) test, was originally developed as a spatial 3-alternative forced-choice (3AFC) test (Wang, et al. IOVS 54:5501, 2013). Newer phones and tablets have bigger screens which allow for additional stimulus patterns to be displayed simultaneously. Increasing the number of choices will reduce the chance level and so decrease the likelihood of overestimating the performance in a psychophysical test. To test this hypothesis we compared the 4AFC and 3AFC test paradigms among normal subjects and patients with maculopathy in this study.
A cross-sectional study was conducted with 86 subjects (40 with normal vision and 46 with various types of maculopathy). The 3AFC and 4AFC SDH tests were implemented on an iPod touch and an iPad, respectively. The tests were controlled by a 2-down, 1-up staircase procedure with 4 reversals. SDH was estimated using a maximum likelihood fitting procedure. Results obtained from 162 eyes of these 86 subjects with BCVA 20/100 or better were included for linear regression and Bland-Altman analysis to assess the agreement of the self-testing results obtained with these two paradigms.
The linear regression of the SDH obtained with the 4AFC mVTTM versus those obtained with 3AFC mVTTM showed that the results of these two paradigms are highly correlated (r = 0.87, p<0.0001). The slope of linear regression is 0.94 (95% confidence interval, 0.85 - 1.02), including slope one, suggesting no significant difference in SDH measurements by these two testing paradigms. The Bland-Altman plot of the difference of 4AFC and 3AFC measurements versus their means showed the mean difference is 0.057 logMAR, indicating that SDH measured by 4AFC paradigm is slightly worse than that by 3AFC paradigm. This bias is significantly different from zero because the 95% confidence interval (0.033 to 0.081 logMAR) of the mean difference doesn’t include zero.
The performance of mVTTM employing a 4AFC testing paradigm is comparable to that of mVTTM using a 3AFC testing paradigm. The slight bias of the 4AFC measurements compared to 3AFC supports the hypothesis that the 4AFC paradigm reduces chance level (lucky guesses), so that it reduces the likelihood of overestimating patients’ ability to detect distortion in a shape discrimination task.
This PDF is available to Subscribers Only