Thursday, January 20, 2011

Expectations Influence Perception

One major principle behind many of our psychological errors and biases is that expectations influence perception. That is, if we expect a given event to happen, we are more likely to perceive that event happening, whether it has really happened or not.

Take the placebo effect, for example. If an individual expects a pill (or other sham treatment) to relieve their headache, their headache will probably dissipate even if it contains no active ingredients. In general, the greater the expectation, the greater the perceived effect, and the effect disappears when the subjects are informed of the placebo. It's hard to overstate the power of this phenomenon. It works with everything from arthritis to bipolar disorder. A review of studies on cough medicine concluded that a staggering 85% of their effectiveness stemmed from the placebo effect, and only 15% was related to the medicine itself.

The effect also changes based on the nature of the placebo given: two sugar pills are better than one, big pills are better than small ones, more expensive pills are better than cheaper ones, brand name pills are better than generic ones, and saline injections are better than sugar pills. The effect also responds to cultural factors: red pills work better as stimulants while blue pills work better as depressants, because red is considered an "energetic" color while blue is thought of as "calm."

Pareidolia is the tendency to find significance in meaningless data. Our brains evolved to detect patterns such as animal calls and faces. But the cost of not identifying these patterns is far greater than the cost of accidentally finding patterns where there are none. For example, it's better to run away from a non-existent predator than not run away from a real one. (Pareidolia that results in perceiving living creatures is sometimes called hyperactive agent detection.)

The clustering illusion is a subset of this phenomenon: people who looked at a string of randomly generated Xs and Os thought it looked non-random. They found "streaks" of Xs and Os in the string – patterns in meaningless data. Pareidolia is also the reason we see clouds that look like bicycles or elephants, and find images of Jesus or Mary in pieces of burnt toast. Weperceive these patterns because we unconsciously expect to find patterns – not just because we deal with them every day (although that's important too), but also because this behavior is hard-wired into us.

Confirmation bias is another area where this principle comes into play: people tend to favor information that agrees with their existing beliefs. For example, suppose one believes that marijuana should be legalized. If one reads about one country that legalized it with negative results and another that did so with positive results, one will tend to ignore the former and focus on the latter. (The opposite would be true if one believes it shouldn't be legalized.) Because we believe that our existing viewpoints are correct, we expect to find information that confirms them, and thus that is largely what we perceive.

This bias is an especially important one because it can affect us at multiple stages. We will tend to gather information, interpret it, and even recall it, according to our expectations. Each step represents another opportunity to create an inaccurate, warped perception of the information we learn and use each day.

The fact that what we expect to see influences what we actually see has profound consequences for Christians and other religious people. For example, if a Christian expects to encounter demon possession, they're likely to perceive someone with a mental illness as possessed. And someone who believes in demon possession might perceive themselves as possessed and then act as such (remember the cultural aspect of the placebo effect mentioned above). This is especially true if there are people surrounding them who validate their behavior. A case like this is particularly powerful, because the expectations, perceptions and actions of two parties build on each other in a positive feedback loop. As another example, Christians often interpret their prayers as having been answered by God, even when the results could easily have come about by chance. They'll also tend to ignore cases where prayers aren't answered (more on how cognitive bias relates to prayer in a later post).

This is my basic model for how these ideas are often reinforced:
Culture → Expectations  Perceptions  Actions  Culture
Our culture and upbringing influence our expectations about the world. Those expectations influence our perceptions, which in turn determine in large part how we act. Those actions, finally, build upon our existing culture. Christians brought up thinking prayer works expect it to work, perceive it to work, and act by continuing to pray and telling others that it works. Those actions strengthen existing Christian beliefs about prayer, and the cycle begins again. It happens in every religion. In this respect, as in many others, Christianity is not special.

No comments:

Post a Comment