Obstacles to Good Reasoning 2: Biases

The first principle [of critical thinking] is you must not fool yourself—and you are the easiest to fool! —Richard Feynman

[A]ll men are by nature provided of notable multiplying glasses, (that is men their Passions and Selfe-love,) through which, every little payment appeareth a great grievance; but are destitute of those prospective glasses, (namely Morall and Civill Science,) to see a farre off the miserises that hang over them, and cannot without such payments be avoyded. —Hobbes, Leviathan Ch. 19

Overview:

A bias is an inclination or a prejudice that can interfere with our reasoning. Biases are best understood as the result of psychological elements such as our attitudes, desires, fears, and motivations. They mess up our reasoning because of how our brains work. Merely having a bias doesn't mean you're going to make a mistake in reasoning. That depends on the kind and strength of the bias. A legitimate bias is simply having a point of view but not allowing that point of view to distort reasoning. An illegitimate bias is one that interferes with one's judgment or reasoning and causes the arguer poorly or dishonestly represent reasons and evidence or make weak logical connections between premises and the conclusion.

There are different strengths of bias: A vested interest is when the arguer stands to gain in some important way if their conclusion is true. I.e., there is a personal benefit to the arguer if their position turns out to be true or is believed to be true.
A conflict of interest is a vested interest on steroids: "A conflict of interest is a set of circumstances that creates a risk that professional judgement or actions regarding a primary interest will be unduly influenced by a secondary interest."1. Primary interest refers to the principal goals of the profession or activity, such as the protection of clients, the health of patients, the integrity of research, and the duties of public office. Secondary interest includes not only financial gain but also such motives as the desire for professional advancement and the wish to do favours for family and friends, but conflict of interest rules usually focus on financial relationships because they are relatively more objective, fungible, and quantifiable." (Wikipedia)

Biases

"Suppose" you happen to run a fruit stand for your summer job. At the beginning of the day you had 10 apples but you sold 6 over the course of the day. The sun is setting and you need to do inventory. You are a Millennial and as such are unable to do simple algebra in your head. You look down at the cellphone that hasn't left your hand all day and pull up the calculator app. You carefully punch in the keys 10-6=7. "What? That's weird," you think to yourself. "I'm not sure what the answer is but I know it isn't 7." You try again and you get the same result.

Fortunately for you a GenXer is passing by. "Hey, Grampa!" you call. "Could you help me for a moment? What's 10-6?" 4, replies the old man.

There are two general ways we can think about errors in reasoning. The first involves what occurred at the fruit stand. You were putting in all the right data but there was a processing error with the app. We can think of this first kind of error of having to do with how we think. The human brain has some glitches in it as a result of our evolutionary history. The other kind of error has to do with what we believe. In the above example such an error would be miscounting the number of apples you had at the end of the day. Even if you did the subtraction correctly, you'd end up with a false belief about how many apples you have. The line between these two kinds of errors isn't clearly defined. Some errors with share properties of both. Nevertheless, it's a useful way to think about the kinds of obstacles we encounter to good reasoning. This lesson will focus on the first kind of error: errors to do with the way our brains work; i.e., with the built in glitches in our brains.

Mommy, What's a Bias?
A bias is an "inclination or prejudice for or against" some fact or point of view. In arguments, this means that we are prone to giving undue favour or neglect to some fact or point of view. Everybody does this (except me, of course); it's part of being a human being. As philosopher Richard Feynman says, "the first principle [of critical thinking] is you must not fool yourself—and you are the easiest to fool"!

There is a wealth of evidence in psychology demonstrating that we begin with our position first then collect or reject evidence and reasons to support that pre-existing position. Our pre-existing position is usually grounded in emotion/preferences rather that "Reason."

The more emotional our investment in an issue, the greater the likelihood that some kind of bias has crept into our supporting arguments—in attributing either undue strength to a supporting assertion or in overlooking or dismissing contrary reasons or evidence. To quote another philosopher, David Hume, "reason is slave to the passions."

Example of how biases overwhelm our ability to reason (5:52-end):

Kinds of Biases

Biases Caused by Individual Interests and Group Identity

There are many ways to think about biases. One distinction is according to whether the bias is associated with individual interests or group interests. A bias that is associated with individual interests will likely cause the arguer to favor their own point of view. One extremely interesting kind of individual bias is called biased fairness. Biased fairness occurs when people are presented with the exact same situation but are situated in different positions. They think they're being objectively fair but their particular role in distorts reasoning.

E.g. 1
Situation from position A: If someone sues you and you win the case, should he pay your legal costs? 85% of respondents say, yes.
Situation from position B: If you sue someone and lose the case, should you pay his costs? Only 44% said yes. Notice that it's people's relative position in the situation that influences their intuitions of fairness rather than facts about the situation. Self-interest distorts intuitions of fairness.2

E.g. 2
In a real world case, researchers examined historical records of salary negations between teachers' unions and school boards. School board typically set salaries relative to comparable school districts. Researchers discovered that the source of disagreement often arose due to what counted as a comparable district. Teachers thought their district was comparable to higher paying districts and the school boards thought it was comparable to lower paying districts. Notice that both agreed that the fair thing to do would be to compare to other districts but their relative position in the situation led them to choose different sources of comparison.

Biased fairness shows how individual self-interests bias judgments of fairness.

Our relation to a group and how that group relates to important components of our identity also distorts our ability to reason well.
Group bias cause us to believe a position on a issue, not because of its merits but because of our affiliation with a group and that group's position on the issue.

E.g., 1.
In the first part of the experiment Democrat and Republican voters were presented with Democrat and Republican welfare policies. The Democrats preferred the Democrat polices and vice versa. In the second part of the study, the experimenter merely switched the label on who the policy belonged to. In this case Republicans preferred Democrat policies while Democrats preferred Republican policies! Group bias overwhelms our reasoning capacities and most people choose whatever their group chooses without considering the content.

Today, a glance at history shows the effect of group allegiance on policy support. For example, the carbon tax and Obamacare which Republicans today abhor were originally Republican policy proposals in the 90s. Democrat support for the War on Drugs was solidified when Bill Clinton rallied behind it. Similarly, Clinton's (at the time) popular Welfare Reform was originally a Republican policy which many Democrats had previously opposed.

Other well-known instances of group bias has to do with the relationship between political ideology and science. Suppose I know that you're a Liberal. Should knowing your political ideology allow me to predict you beliefs about bio-chemistry or particle physics? It would be really weird if it could since scientific beliefs and political beliefs don't seem to be logically connected in any way. And that just the point. In fact, I can fairly confidently predict your views on climate change, nuclear energy, evolution, and GMOs!

But why? Science and politics aren't logically connected. Perhaps not logically but they certainly are psychologically connected. In order to preserve identity and group membership people overwhelmingly adhere to whatever beliefs their political tribe adopts in respect to science. Their capacity to reason has been overwhelmed by a glitch in their hardware: The tendency to conform one's view with that of one's tribe rather than to evaluate the evidence impartially and systematically.

Anther interesting example in group bias:

President-elect Donald Trump's electoral win was followed by a dramatic spike in favorability for Putin by GOP participants in the poll. Democratic participants, meanwhile, now disapprove of him more than they did before November 8. In July 2014, Republicans viewed Putin with a -66 net favorability; and now, in December 2016, they view him with only a -10 negative favorability. During the same time frame, Democrats went from -54 to -62 net favorability. Additionally, in summer 2013, WikiLeaks—which released secret emails from Hillary Clinton's team during the campaign—was viewed negatively by Republicans by a 47-point margin. Today, the party's voters view the organization, founded by Julian Assange, favorably by a 27-point margin—a post-election swing by 74 points. Democrats have, unsurprisingly, moved the opposite direction in a 25-point downward swing.3

Political biases are so strong that they can even overwhelm our ability to do math! Click here for a write up.
And here's more on how your politics affects your non-political beliefs: Article
Even supposed objective fact-checkers can sometimes be overwhelmed by their biases: What's the Black youth unemployment rate according to Politifact?

Legitimate and Illegitimate Biases

We've established people (except me) have biases. Now what? Do we automatically reject everybody's arguments 'cuz they're biased? Nope.

We can make a distinction between legitimate and illegitimate biases. The distinction will depend mostly on how opposing reasons, evidence, and arguments are portrayed, and if there are any intentional important omissions or distortions.

A legitimate bias is simply favoring a point of view but not in a way that misrepresents the opposing position. It allows an impartial observer to fairly evaluate the proposed point of view.

As you might have guessed, an illegitimate bias is one in which the arguer poorly or dishonestly represents evidence and reasons, or if the bias leads to weak logical connections between premises and the conclusion. Any website or blog with a strong political bias in either direction will usually provide excellent samples of arguments with illegitimate biases. We can distinguish between two kinds of illegitimate biases in terms of their strength:

Vested Interests
A vested interest is when an arguer (or someone paying the arguer) stands to benefit from their point of view being accepted. When vested interests are involved there's a very high likelihood of illegitimate bias.

For example, when certain industries spend millions of dollars to pay lobbyists and "donate" to politicians, it's not unreasonable to suspect that their arguments for special treatment or tax and regulatory exemptions contain illegitimate biases.

Not all vested interests need be financial. One might be motivated by the desire for power, fame, revenge, attention, sex, etc.. or to get out of trouble/prove one's innocence.

We should be cautious of dismissing arguments out of hand just because the arguer has a vested interest in the outcome. That they have a vested interest tells us nothing about the argument's validity and soundness which should be evaluated independently When there is a vested interest, it simply means we should be extra cautious about illegitimate biases (and omissions). It doesn't automatically follow that the argument is invalid or unsound. To figure that out, you have to actually evaluate the argument.

Conflict of Interest
A conflict of interest is a vested interest on steroids; i.e., when vested interests are extreme. In such cases there is usually an ethical issue involved too, and in professional settings, conflicts of interest have to be disclosed.

For example, in medical research if a university study of a drug is funded by the company that produces the drug, this is a conflict of interests for the researchers. Their labs and job depend on the funding. Thus, sources of funding must be disclosed at the beginning of any research that is produced. This is actually quite a big problem in medical research because drug studies that are funded by the drug producer systematically have higher positive results than if the same drug is studied by a neutral party. For more info check out the link]]]:Medical research and funding.

But bias in medicine (and elsewhere) isn't only on the "proponent's" side. Often people who oppose something for ideological reasons are just as guilty of bias.

An important example of a conflict of interest in medicine that wasn't disclosed was Andrew Wakefield's anti-vaccine research article in the Lancet. What he did not disclose in his research was that he had been paid several millions of dollars to do research on vaccines by a company that was developing an alternative to the conventional vaccine.

There was a clear conflict of interest because he stood to gain so much if his research showed that standard MMR vaccine was unsafe. The company that had funded the research was developing an alternative that they wanted to bring to market. If they could generate fear about the existing vaccine, they'd have the whole market to themselves.

In the end, his results were never replicated, his methods shown to be unethical, his data drawn from a statistically insignificant sample size (12 children), and the article was subsequently retracted by the publisher. However, because of the fear that came about because of his "research," there was and continues to be tremendous damage to public health. To learn more click here.

Summary:

We all have biases. What matters for critical thinking is the degree to which they distort the presentation of evidence and reasons in arguments both for and against the arguers position. Biases are illegitimate when they cause distortion such that arguments cannot be fairly evaluated.

Interesting Articles on Biases

Why You're Biased Against Being Biased

For some excellent examples of how biases affect how we interpret the world, this is a beautiful article..

Examples:

As noted above, from its creation and through Bill Clinton’s presidency, Medicare lacked a prescription-drug benefit. It was not until 2003, under President George W. Bush, that Congress added the Part D benefit, through which Medicare pays for seniors’ prescription drugs. The enactment followed a controversial House roll call vote, which Republicans held open for several hours as party leadership maneuvered to secure enough votes for passage.4 One bargaining chip to attract market-oriented Republican votes was the so-called “noninterference clause”—a provision drug manufacturers had a major role in writing and getting through Congress—which banned negotiations between Medicare and pharmaceutical companies on drug prices and prevented the government from developing its own formulary or pricing structure.5 Instead of CMS negotiating on Part D plans’ behalf, prescription drug plans compete for enrollees and negotiate directly with manufacturers.

Congress has taken no action to repeal the ban, and that seems unlikely to change. The pharmaceutical industry’s lobbying efforts topped $231 million last year;6 indeed, the industry has spent more on lobbying than any other industry since 1998. It seems unlikely that a newly elected President would start his or her administration by reigniting the fight with the pharmaceutical industry that previous presidents, of both parties, have lost.

Examples of group bias on Democrats in the 2016 election: It was widely tweeted and shared around the internet that a Wikileaks set of emails were in fact fake. It turns out there was no evidence for the claim but the idea was widely accepted anyway.7

++Examples of Conflicts of Interest and Vested Interests
Donald Trump and Wells Fargo: Trump and Wells Fargo

The pharmaceutical industry is a major funder for anti-marijuana legalization campaigns.8

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License