Abstract
The four low-level matching tests (length, size, orientation, and position of gap) in the Birmingham Object Recognition Battery (BORB) (Riddoch & Humphreys, 1993) are commonly used to assess basic visual perception in neuropsychological participants. These tests have been valuable measures, but one of their limitations is that they require same/different judgments, and participants with normal perception sometimes have trouble determining what counts as same versus different. In addition, the tests are paper-based so they must be done in person and the length of the stimulus presentations cannot be easily controlled. Here we present a new basic visual processing test that involves forced-choice decisions and is completed on a computer: the Hanover Early Vision Assessment (HEVA). Each task has 24 trials, and in each trial, participants are briefly presented with three choices and asked to select the odd-one-out. The five different tasks require participants to make judgments about circle size, ellipse size, angular size, line length, and dot distance. Total testing duration is 15 to 20 minutes. Control subjects achieved an average of 74.6% correct (SD=10.1%, n=117) and the test-retest correlation for testing sessions separated by a week is high (r=.83, p<0.001, n=33). The HEVA correlates moderately with the BORB’s low-level matching tests (r=0.42 , p=0.02, n=29). It was not significantly correlated with the Leuven Perceptual Organization Test (r=0.33, p=0.09, n=29) (Torfs, Vancleef et al., 2014) which tests for intermediate and higher-level vision, or the Verbal Paired-Associates Memory Test (r=0.19, p=0.33, n=29) (Woolley et al., 2008), a test of word pair memory. Together, the results indicate the HEVA is a reliable and valid measure of low-level vision. We will make the test freely available for research purposes.