I have read about a lot of psychology experiments where the subjects are not told what they are really being tested for. A typical experiment may, in addition to the control group, have one group do a vocabulary test using positive words and another group do a test using negative words. Then all groups might be asked to evaluate a piece of writing, with the hypothesis being that the evaluations will be more likely favorable or unfavorable based on the vocabulary words.

Based on what I have read, if I were ever a subject in a psychological experiment, I would keep trying to guess what they were trying to test for. For example, if I were in an experiment like the one I described, I might make an extra effort to not be influenced by the vocabulary words that were offered. This might not make for very accurate testing, regardless of whether I guessed correctly.