Why is it difficult to learn absolute judgment tasks?

The effects of using different presentation orders on absolute judgment were examined in four experiments. 120 college students identified horizontal bars of eight different lengths. Using the same set of stimuli, participants were trained under different presentation schedules with feedback and then tested with various common schedules with or without feedback. Participants trained with the small-differences schedule (small-step) performed well during the training, consistent with prior findings but generally poorer on subsequent tests. In contrast, the participants trained with the large-differences schedule (large-step), performed less well during the training but did not perform more poorly on the tests. The results suggest that mental processes such as abstraction of relative uniqueness and memory overload may be involved.