Bias in Science

One of the funny things about political bias is that it's incredibly difficult to detect it in yourself. You spend so long identifying with a certain ideology that - eventually - it seems like a self evident fact. We can see a number of examples of this in modern society, but one example I want to draw on is this article from 2010.

Note, first of all, the smug sense of satisfaction. The author has found a theory that explains why Liberals are better than Conservatives. Tres bien! No need to worry folks, we've settled in once and for all. As much as I'd love to lay into the personality factors present in this article I want to stick to the psychological issues at work.

The problem with this article is not the internal logic (although that is a bit iffy) but just bad science in general. First of all note the definition of Liberal:

"...the genuine concern for the welfare of genetically unrelated others and the willingness to contribute larger proportions of private resources for the welfare of such others." 

 This definition is fine as it goes, so long as it is true. But is it? One of the main problems with deductive reasoning is that you have to start with a true premise and work down, but how do you know the premise is true? You have to work up (inductively) from evidence. Is there evidence that this definition is true? I'm sure many Liberals would identify strongly with it, but I know plenty of Conservatives (particularly Christians) who would ALSO identify with it. What are we to make of that? The author offers no reasoning for this definition, merely that it is "reasonable" to assume that's what defines Liberalism (presumably the opposite idea; not caring for others and unwilling to contribute private resources to help people is indicative of Conservatism, but as I said earlier can we reasonably assume that?), the author also claims that this "usually" translates to support for welfare programs. The author does not entertain the idea that a person may wish to help others but also oppose welfare programs (trust me, these people exist.

Another sticky issue that the author blithely tosses aside is the fact that Conservatives are more charitable (presumably because they are more religious). This, according to the author, is explained by the theory because truly altruistic Liberals don't mind giving indiscriminately and so support blanket higher taxes and welfare programs. Conservatives on the other hand are deeply prejudiced people who want to decide who they help (the example is, unsurprisingly, old white men) and who they don't want to help (poor black women).

Finally the author presents us with some seemingly unselfconscious circular logic. A common complaint, they say, is that Liberals control the major social institutions. This, apparently, is true. It's true, according to the author, because Liberals are smarter, and therefore more likely to control things. How do we know they are smarter? because those same institutions have done studies proving it. Apparently unable to acknowledge this bias the author ends by alluding to the apparent evolutionary novelty of Liberalism (as opposed to Conservatism) and stating that the smarter people (liberals) are therefore more adaptive to modern society.

Now, I don't want you to get the impression I am ragging on Liberals and vaunting the Conservatives. I'm not a fan of either group and I don't identify with either one. In fact, like a lot of young people, I tend to take a more eclectic view of personal philosophy and political ideology. Sometimes I take a liberal stand, sometimes a conservative one and sometimes (more often than not) I take a mixed stand.

My point is that politics is one of those areas where we have a vested interest in preserving the integrity of our group and actively harming the integrity of the out group. Liberals are convinced Conservatives are evil, and vice versa. The problem is that when this bias bleeds into science we can become incredibly stupid. One only has to take two ideologically opposed media institutions, for example, to see how twisted facts can become; for example The Guardian - a broadly left wing paper in the UK - will often carry stories of the depravity and general incompetence of Conservatives in and out of government. Similarly the Telegraph - a right-wing paper of similar repute - will carry studies, facts, figures and so on about how bad the Liberals are.

Who is right? We don't know, because unless you dig up the research and look for yourself (and let's be honest, who does that?) you are at the mercy of the authors bias. I could probably do a similar analysis as I have done above by picking apart an article on why Conservatives are better than Liberals and you'd see the exact same level of bias.

Hopefully this blog post has given you a bit of a flavour for the problems with a) hypothetico-deductive reasoning based on faulty premises and b) the dangers of political (or other) biases in science.

Just remember that there are plenty of idiots on both sides of the fence... 

"Debate is Over Syndrome": Applicable to Psychology?

I just read an interesting article here. Joel Kotkin talks about a well known (but little acknowledged) problem in broad intellectual society. The problem he describes is a simple one, the majority view is taken as "settled" fact and dissent is punished, sometimes quite vociferously, by members of the in-group.

Kotkin talks about this in the context of Climate Change, a contentious issue that I don't want to get in to here. He points out that a frequent refrain is heard; "The science is settled". Debate on climate change is stifled and ridiculed and people have lost jobs, careers, and fame because of their views. Worse, he points out that the people who are doing the punishing are usually the ones who are most certain of their own right-ness. Namely left-leaning intellectuals and well-funded right-leaning media moguls.

Typically, only one of those groups are considered to be biased; the right-leaning ones. However true of both groups is a blindness to the possibility that an error could have been made. In climate change this is seen in the scarily obtuse way it is discussed, the ridicule aimed at dissenters, and the sheer arrogance of proponents towards those who question established facts.

Climate change is a contentious issue as I said, I don't want to debate whether its real, or if it is, who or what is causing it. As far as I am concerned there is something there, but the science is far from settled.

Now on to the real reason I brought up this essay. I'm not so much interested in political bias (although it is a fascinating subject) but rather Psychologist Bias. I am a Radical Behaviourist, and as such definitely part of a (strong) minority. My views are casually belittled and in some cases ridiculed by the public and certain outspoken intellectuals. Most of this arises from misunderstandings, especially among the public; "Don't behaviourists believe we don't have thoughts? Don't they hurt animals and think people are the same as pigeons?" and so on...

Another example, more prevalent among academics, is the belittling of experimental methods. The behaviourist is not a fan of statistical inference. Not because we are afraid of statistics (although I do dislike them!) but because we don't see as many opportunities to use them as a standard experimental psychologist might. In Behaviour Analysis we use a method called single-case designs (not to be confused with case studies) which use a number of methods (mostly non-statistical) to analyse individuals or small groups and their behaviour.

I don't want to get too technical as I realise a lot of my readers are not trained scientists and it would probably get rather boring if I started expounding on the relative merits of p values and ANOVA's. What I want to point out is that the view I talked about above, the "debate is over syndrome", is quite prevalent among some psychologists. Now to be sure, I don't want to paint with a broad brush and I realise that many psychologists are quite respectful of behavioural psychologists and understand the reasons for why we do what we do, none the less there are some who believe that "the debate is over" and behavioural psychology lost. Anything I can do as a behavioural psychologist pales in comparison to the power of a statistical test and a hypothetico-deductive, mentalistic attitude.

There is still a lot we can learn from behavioural psychology, and from the methods and philosophical approach it employs, and I think it's highly premature to claim that the debate is over and cognitivism won. If you find yourself thinking this, I encourage you to meditate on these words by philosopher Ayn Rand; "Check your premises".

It is not, nor ever will be, my intention to offend. Cognitive Psychology has opened up a whole new, fascinating world of research to the psychologist and it would be foolish to ignore it's impact on the science, but at the same time it's important to remember that behavioural psychology is a viable, exciting science that still has a lot to teach us, and I, for one, hope more academics come to realise that more could be done if we set aside the dictatorial mindset of "debate is over" and returned to the halcyon days of science as an exercise in curiosity.