We make thousands of decisions every day—where to walk, who to greet, what to eat—and though much research has found that choices made today may affect those we make tomorrow, no clear connection had explained why.
Now, work published in the journal eLife from University of Pennsylvania psychologist Alan Stocker and doctoral candidate Long Luu, shows that we’re influenced by our previous decisions in an effort to avoid cognitive dissonance. In other words, through the choices we make, we attempt to eschew inconsistencies and contradictions. The findings pave the way for deeper investigation into our decision-making processes.
Stocker, an associate professor in the Department of Psychology and Luu, a fifth-year graduate student, drew this conclusion by using a newly developed model to test the hypothesis that attempting to remain what’s called “self-consistent” leads to biases when people make future decisions. “The model we used assumes that a person’s perception is influenced by both his or her sensory evidence and earlier choices,” Stocker says.
The scientists carried out a set of three psychophysical experiments with six male and four female participants, who each went through several initial training sessions before completing 40 trials for each of 15 “stimulus orientations” and three control conditions.
In the first experiment, subjects viewed a white dot and two black marks for a little more than one second to establish the “boundary,” then viewed the “stimulus,” an array of white line segments, for about half a second. After seeing these components, they had to judge whether the overall orientation of the visual pattern was clockwise or counterclockwise of a reference, then recall the pattern’s actual orientation from memory.
The second experiment tested the extent to which subjects’ orientation estimates depended on their knowledge of the stimulus in the first test. The third, in an effort to get at whether participants treated their own judgements as being correct, asked them to recall the color of the stimulus—in this case, either red or green—rather than its orientation. An additional cohort of 10 people repeated the second and third experiments.
“Our tests showed that the bias pattern depended on the subjects’ prior knowledge of the stimulus, and that the subjects treated their decision as if it were absolutely correct,” Luu says. “The results suggest that our decisions can substantially change our immediate memory of what we have just perceptually experienced, in an attempt to make our experience consistent with our preceding decisions.”
Luu adds that the current model links to well-known cognitive phenomena such as confirmation bias and suggests that, in decision-making generally, the brain focuses more on remaining self-consistent than on remembering precise details of the past.
“This indicates that humans subconsciously condition themselves to be self-consistent in how they remember the past, and our model allows us to precisely quantify and predict these bias effects,” he says. “Further research is now needed to understand how this self-consistency bias is reflected in the neural processes underlying our decision-making activities.”
Alan Stocker is an associate professor at the Computational Perception and Cognition Laboratory in the psychology department in the School of Arts and Sciences. Long Luu is a graduate student in the CPC Laboratory.