Data Science & Psychology Data Science applied to Values, Morals, Politics, & things that matter.

Does social psychology try too hard to be perceived as a “science”?

I recently read this article in the American Psychological Society's magazine, the Observer, and it reminded me of this article by Paul Rozin, detailing how social psychology's desire to be perceived as more scientific has led it to restrict the range of methods deemed acceptable (an over reliance on confirmatory rather than exploratory methods).  As someone who came to psychological science later in life, in order to understand the world, rather than obtaining a psychology degree in order to earn a living, I have to admit some discomfort with the way psychologists attempt to use the word science.  Psychology definitely uses the scientific method and most psychologists I know are very well trained in that respect.  Yet, through no fault of psychologists themselves, I believe there are very real differences between the subject matter that psychologists study (human beings) and the subject matter of other disciplines, such that the word "science" is sometimes a misfit.  Three obvious misfits are:

• Human beings have some degree of free will which confounds reproducible results. Some will dispute this, and maybe humans are more constrained than we might think, but that is a far cry from believing that human behavior is fully determined.  In contrast, a computer is actually incapable of producing a random number that is not fully predictable, though it can measure an ostensibly random event outside of itself and report that as a random number. If psychology or neuroscience ever produces a machine that can predict which number a person will pick from 1-100 with 100% accuracy in the same way that you can predict a machine's decisions (if you know the inputs), I will retract this paragraph, but I doubt that will ever happen.  As long as you allow for some degree of free will of the subject matter, the scientific paradigm of reproducibility breaks down.  Imagine if the gas in our car could decide not to combust or the electricity in our refrigerator could decide not to flow in the direction intended.  How reliable would our cars and appliances then be?  Physicists should be quite thankful that atoms are not so willful.
• We care about individual human beings, not the "average" human being. When we do studies on electricity and figure out a method to conduct electricity with 10% less loss of energy, without any side effects, we have undoubtedly made a worthy discovery.  It doesn't matter if the same intervention causes some energy that would otherwise be captured to be wasted, as long as the net effect is positive.  Energy is fungible.  In contrast, human beings are not.  If we conceive of some intervention that makes 10% of people more happy, undoubtedly it will make x% of people less happy.  All psychology studies are full of people who react in the opposite direction to the published results.  For example, people generally conform to social norms...but we all know that there are people who do the complete opposite and rebel.  This can be mitigated by measuring individual differences, but the problem is that human beings are both the consumers and the subjects of our research.  People care about what makes them themselves happier, not the "average" person.  Since psychologists cannot say with certainty what will happen to any individual person, the consumer of psychological knowledge has to filter anything said to them through the filter of their own individuality, in order to derive utility.  While our results may often be reproducible on groups (despite free will), they never encompass 100% of people and so there will always be some degree of subjectivity to a result.  For this reason, a memoir where a person can be more certain of the applicability of a phenomenon, due to the deep description of a situation, really can lead to more useful insight than a study that describes what scientifically happens to an "average" person, from the perspective of a knowledge consumer.
• Human beings are highly evolved social animals that can intuit complex things about other human beings. The observer article highlights the findings of this probability sample survey where it was found that people believe that psychology is a science, but also that everyday life provides training in psychology (see below chart).  In both the original study and the APS article, these statements were presented as somewhat contradictory pieces of evidence, but I don't think this is necessarily true.  I believe both are true.  People intuitively know a lot about how people work.  Their perceptions may be biased by experience, but then again psychologists are biased by sampling in the same way.  I firmly believe that anyone who believes that you can't learn something profound about the human condition from a good book hasn't read a good book (or is blinded by their own biases).  I cannot see how psychologists can be truly useful to the world as long as we believe that our methods are the only ones that speak to questions about human psychology and ignore outside wisdom, much of which is more advanced than what we study.

Personally, I agree with the survey respondents in the above study.  Psychology is a science and psychologists are among the best scientists in the world, if only because we have to deal with a subject that is willful, variable, cares about itself, and where the bar is so much higher for producing marginal knowledge above that of the average person.  Psychologists should be applauded for that.  Yet, I believe that trying to tell people that our image as a discipline rests upon the belief that what we do is very similar to what physicists and chemists do will always fail because the above three reasons are self-evident to most consumers of psychology.

Psychology is what it is.  Trying to sell it as more may lead psychologists be less trusted, as we seem more out for our own gain than for doing something useful.  It may also lead psychologists to adopt unhelpful procedures, designed to prove our scientific-hood, as opposed to being of use to others, and/or to ignore those methodologies seen as less "scientific" (e.g. qualitative studies or narratives), in order to be associated with methodologies seen as more scientific (e.g. neuroscience, as in the article), even when we could learn something from both.  There is truth in any view of the world, and the best truths are arrived at from a variety of angles, in the same way that the use of many measurement techniques reduce overall error.  According to the above graph, under 50% of people (in 1986) believed that psychology has an impact on their daily lives.  I absolutely believe that psychologists should continue to use the scientific method as best we can.  However, I also believe that psychologists may actually make more of a real impact if we realize the inherent limitations of our subject matter and stop worrying so much about whether we are grouped with the physicists or the sociologists.

- Ravi Iyer