An experimental analysis of maze discrimination.

Abstract

73 [male] and [female] rats were trained in a single-unit alley-type maze. Half were trained by the correction method (animals permitted to correct their mistake on each trial) and half by the non-correction method (correction not permitted[long dash]no reward received on trials in which animal initially chooses the blind). By the correction method increasing the length either of the blind or of the true path or both seemed to increase the difficulty of the problem (but not reliably). The non-correction method was superior to the correction method. A "two-factor" theory is put forth to account for this result and related findings by others.