What You See Isn’t Always What May Be: Confirmation Bias

By Sally Raskoff

Confirmation bias is a fascinating dynamic. What we see may not be what we judge it to be. What we think we are seeing may just be what we expect it to be.

A new study by sociologists Aliya Saperstein and Andrew M. Penner highlights how social status cues, mixed with gender, may change judgments and perceptions about racial group membership.

The study finds that classification of racial identity changes based on social status cues and is mediated by gender. For example, a woman who went on welfare and a man who served prison time were less likely to be identified as white. Gender isn’t always a factor; people who appear to be poor are also less likely to be identified as white.

When we see people who fit into our society’s cultural stereotypes about ”types” of people, we may then perceive their race to be something different than others might perceive, and different from the racial category with which the very person self-identifies.

Confirmation bias works in insidious ways as it silently confirms the stereotypes (or biases) that surround us. Wealthy, educated, and successful people are assumed to be white while poor, uneducated, and deviant people are assumed to not be white.

One of the fascinating aspects of the recent study is its ability to illustrate how these are simple perceptions that people make, based on the information they have about the person. The racial identification of the observers varied over time based on what life events occurred for the subject. No matter how the person as first identified – or how the individual identified themselves – later identifications varied based on those social status cues and resonate with societal stereotypes.

In sociology classes, we often show a series of images of people and ask the class to identify the racial category of those people. Among the many purposes of this activity is to help students realize that one cannot identify race based on appearance – and that race is a social construct, not a physiological one.

A favorite image that I use is that of Walter White. (Not the character from the TV show Breaking Bad.) Most students do not know who he is and the photo shows a well-dressed man with a fair complexion and light hair. Most students identify him as white. (See image at right.)

He, of course, was not “white” but was of mixed ancestry and identified as “negro.” He had an accomplished career, including important work for desegregation and civil rights.

What happens when we encounter people or images that do not mesh with societal stereotypes? They are hard to believe, we might explain them as the exception to the rule, or we might not even recognize what we see.

My favorite example of this is the photo of Marilyn Monroe reading a copy of James Joyce’ Ulysses. Marilyn Monroe, often used as the icon of femininity, reading a book, and a difficult one at that? Most people who see this photo assume that it is a staged or posed photo, that she just picked it up and was trying to figure it out, or they don’t notice the book or care what book it is.

What do you see when you think of doctors, astronauts, engineers, and professors? How about nurses, administrative assistants, teachers, and sex workers? Are those images gendered and perhaps even “raced”? Do you think of mostly white men inhabiting the first list and women, white or non-white, for the second list?

The concept of confirmation bias links well with tools such as the Implicit Association Test (IAT); you can take a demo of the test yourself. The IAT and Harvard’s Project Implicit, can be used to identify how deeply our societal biases run. What other kinds confirmation do you think people commonly hold?

Comments

[Katie Flemming]

My understanding of the confirmation bias is that people test hypotheses by looking for information that support it, rather than information that contradicts it. I feel as though you don't really apply this to the research. Can you elaborate on how the two are connected?