Given that the results of Study 3 showed that the task was learnable, generalized to new exemplars, and generalized to new orientations, in Studies 4 and 5 we sought to examine how aspects of the training task affect learning outcomes. Within the domain of perceptual learning, there is decidedly mixed evidence regarding whether explicit feedback is necessary to drive improvements in performance. In Studies 2 and 3, the feedback screen had two key components: (1) correct/incorrect feedback (i.e., whether or not the participant responded correctly), and (2) information regarding exactly where the appendix was in the image (the arrow pointing at the appendix). Although previous work examining the importance of feedback in driving perceptual learning has typically contrasted conditions with feedback against conditions without any feedback, here, given the highly complex nature of the stimuli, it was unclear whether the correct/incorrect feedback alone would be sufficient to drive learning over the time frame of the study. Indeed, in all models of learning, the purpose of feedback is to drive changes in estimates, models, templates, etc. In the case of simple tasks where participants are asked to make decisions about simple unitary dimensions, it is clear how correct/incorrect feedback alone could be sufficient (or in some cases not even necessary, such as if participants have an internal model that could produce an internal error signal) to drive learning. Yet, in the complex radiological stimulus space, we surmised that in the absence of understanding why choices were correct or incorrect, or where the relevant information was in the stimulus, learning may be significantly slowed or even eliminated, at least to the extent that we could measure this over five sessions of training. In Study 4, we thus sought to determine how necessary the second form of feedback that was provided in Studies 2 and 3, where participants were shown where the appendix had been in the previous image after each trial, is for the learning process.