Learning how bias really works

PREJUDICED

Mahzarin Banaji is an unfamiliar name. Is it Jewish, I wonder. I learn that it belongs to a female professor at Harvard University. Her lecture at the National Institute of Advanced Studies, Bengaluru is on Blindspot: Hidden Biases of Good People. The topic seems fascinating. I decide to attend.

Merriam-Webster defines bias as “a tendency to believe that some people, ideas, etc., are better than others, that usually results in treating some other people unfairly”. As the lecture begins, I wonder whether by believing that a professor from Harvard is better, I was going to treat a professor from another University unfairly? No, I don’t think so.

As the lecture progresses, we learn of biases in places as varied as elite colleges, media organisations and musical operas. We learn, not surprisingly, that people associate theft and looting with the lower income group and think that psychology is more of a woman’s field and physics, a man’s field. Surely, the educated audience doesn’t have these biases? Therein lies the rub. “We may believe that we are not biased towards any particular group, but we most definitely have such preferences,” explains Mahzarin.

Sexist biases

More often than not, our biases result in actions that are detrimental to the other group. For instance, an employer might profess to having no gender bias, but will choose only a male candidate because of the unconscious belief that men are better at a particular job.
Mahzarin has been studying this area of social cognition for the past three decades and her work is considered transformational. But how do we prove that these biases exist when people deny them outright? In a series of interesting experiments, Mahzarin and her colleagues have demonstrated the widespread existence of biases in various categories — gender, race, ethnicity, sexual orientation and even political affiliation.

Implicit Association Test (IAT), which measures these biases, has become very popular among psychology researchers and even the general public. An astounding number of tests (over 15 million of them) have been taken by people interested in gaining greater awareness about their own unconscious preferences and beliefs. The IAT works like this. For example, to measure gender and career association, people are asked to press one key for male names and another for female names. The test proceeds to ask people to press one key for words associated with family and another for career. Most people find these two tasks easy enough; they perform correctly and fast. But when a female name and career word or a male name and family word are grouped together and people are asked to press a key, they (even career women) take longer and make more errors!

These implicit associations are biases that have been socially and experientially strengthened in most of us. So, if these biases are our blind spots, does that mean they can’t be changed? Mahzarin has this interesting take on the matter: “You may not know about your high blood pressure till you get yourself tested, but once you know, you would do something about it.”

She believes being aware of these biases is the first step in its correction. She recalls the stunned reaction of the CEO of a big company when he found that he had simply missed the obvious contributions of his women colleagues — information that Mahzarin had placed before him.

Subsequently, he vowed to do the needful to amend the situation. Personally, Mahzarin says that while going through candidates’ applications, she makes sure that non-essential details like name, alma mater and the like, which might bias her selection criteria, are not included in the forms.

Matter of perception

Exposure to a different set of data can influence perceptions too. Mahzarin mentions the case of a young boy whose mother was a physician and whose paediatrician was also a woman. This led him to ask, rather innocently, “Can boys become doctors, mama?”

Racial bias in the American population has been proved in various fields — from employment and education to even presidential elections. Does India, with its myriad regional, class, caste and other distinctions, have even more biases? “We can speculate that India has more scope for biases as it is subdivided into many more groups, but it is neither easy to prove nor useful,” says Mahzarin.

And what’s her take on reservations for women? Mahzarin says that reservations in the traditional sense are not useful in the present context. Even an excellent candidate who comes through the quota system will have her skills undervalued. Instead, Mahzarin recommends novel measures such as evaluating a candidate’s merit without being aware of the candidate’s gender.

She also strives to encourage more women in science by being a role model. “I have to fight the bias that psychology is not a science,” she says, laughing. “I tell people, understanding the mind is not rocket science. It’s a whole lot harder.” And oh, does the fact that Mahzarin is originally from Hyderabad and a Parsi bias the reader? Well, it’s time for some introspection!