Vinsamlegast notið þetta auðkenni þegar þið vitnið til verksins eða tengið í það: http://hdl.handle.net/1946/27773
It has been suggested that the brain tries to predict the most probable outcome in each circumstance using prior predictions to evaluate posterior outcome. That way the brain works as a probabilistic machine that avoids prediction error. This raises the problem: if the brain tries to avoid surprises, its optimal arousal level would be in a dark room where nothing ever happens. A recent proposal by Chetverikov and Kristjánsson tries to overcome this problem by adding that prediction is also guided by affect. Different perceptual outcomes yield negative or positive affect that influence futures choices. The proposal further suggests that cognition is driven by the chance of learning rather than the most probable outcome. Here, I tested under what conditions humans prefer novelty to familiarity. The hypothesis is that when all stimuli have a high prior likelihood, people are biased towards novelty. Furthermore, it suggests that when all stimuli do not have a high prior likelihood, people prefer familiarity. I tested whether manipulation of the context surrounding the stimuli affected which stimuli were chosen. The results did not support the proposal that participants choose differently depending on the surrounding context. The obtained results can be explained by the mere exposure effect.