How do three biases quietly distort our judgment?

Each bias distorts a different stage of thinking. Confirmation bias affects how people look for and weigh evidence. Hindsight bias affects how people remember and judge the past once an outcome is known. Motivated reasoning affects which conclusions people reach when they want a particular answer. The three overlap, and researchers do not always agree on where one ends and another begins. [1] [2] [3]

In Nickerson's review, confirmation bias means seeking or interpreting evidence in ways that favor existing beliefs, expectations or a hypothesis in hand. He stresses that it is usually unwitting. A lawyer or debater builds a one-sided case on purpose, whereas confirmation bias is mostly unaware selectivity. It can occur when people are defending a cherished belief and also when they have no stake in the answer. [1]

The distortions take several forms. People may consider only one favored explanation, so data can never count for an alternative. They tend to seek the kind of information they would expect if their hypothesis were true. They also give more weight to supportive information and are more likely to explain away or discredit contrary information, rather than ignore it outright. A simple example is favoring a candidate and then seeking out and remembering favorable news about that candidate. [1] [4]

Hindsight bias is the feeling of having known it all along: an event seems more predictable after it is known than it was beforehand. A review in Perspectives on Psychological Science describes three possible aspects: distorted memory, beliefs about how likely the event objectively was, and beliefs about one's own predictive ability. It traces the bias to selective recall of information consistent with the outcome, to the ease of understanding a known outcome being mistaken for prior likelihood, and to a need to see the world as orderly and to avoid blame. [2]

The consequences include fixing on a single causal story while neglecting other reasonable explanations, and overconfidence in one's judgments. The review reports that prompting people to consider alternative causal explanations can reduce the bias. The bias matters in practice: in negligence cases, jurors must judge a defendant's conduct, and the law requires them to ignore outcome information to avoid it. [2] [5]

Motivated reasoning is described as interpreting new evidence in ways that support, or lead to, the outcome one wishes were true, even when the evidence does not truly support it. One popular summary also notes that people scrutinize arguments for conclusions they do not believe in more than those they favor. [6]

Hahn and Harris argue that the terms have become tangled. In their account, confirmation bias began with Wason's finding that many people test a rule by seeking confirming cases, a search strategy that needs no desire to confirm anything. The term then expanded to cover many phenomena that overlap with motivated reasoning, even though the original finding involved no motivation. They also say robust evidence that wanting an outcome inflates its judged probability, called wishful thinking, has been hard to find under controlled laboratory conditions. Real-world cases they discuss seem explainable as an unbiased evaluation of a biased body of evidence. Nickerson, by contrast, treats motivated and unmotivated forms as both well documented under one heading. The results here do not show which view most specialists hold. [3] [1]