Conditional Discrimination without Explicit Reinforcement

In a typical conditional discrimination procedure, the participant is first presented with a sample stimulus, followed by two or more comparison stimuli. Having chosen one of the comparison stimuli, the participant gets feedback on whether the choice was correct or incorrect. In the following study we asked whether it was possible to establish conditional discriminations without any programmed consequences, and furthermore, would the participant also respond correctly when presented with a test where the order of the conditional discrimination relations are for example reversed. The training and testing was presented on a paper and pencil format. There were 36 rows, each with 4 columns (1 sample stimulus in the leftmost column and three comparison stimuli in the other four columns). Sixteen college students participated. The students did not receive any programmed consequences on their choices during training or test. The criterion for passing the test was 100% correct on all tested relations. The results showed that 15 participants responded correctly during training, while none responded according to the experimenter defined criterion in the test.