Errata for Phillipson GP, Read JCA (2010)

Stereo correspondence is optimized for large viewing distances

On p. 4 of the paper, it states, “Subjects each performed at least four blocks of 150 trials (five repetitions of five different correlation levels, two conditions, three experiments).”

This statement is misleading, giving the impression that our subjects took much less data than they actually did (4 blocks of 5 repetitions resulting in just 20 trials for each data point). It is true that the smallest number of repetitions for any data-point was 20: Subject LAH performed 9 correlation conditions, and at 3 correlations performed only 20 trials for each condition/expt combination. However, at the remaining correlations, LAH performed 80 or 100 trials. Subject KCR took only 60 repetitions per data-point, but all other subjects took at least 80 at 5 different correlations. Thus KCR completed 1800 trials in total; the others completed at least 2400, and two did 3600. A better statement would be “Subjects each performed at least 60 repetitions of five different correlation levels, for each of two conditions and three experiments. All but one subject took 80 repetitions of five different correlation levels.”