As for one’s ego, we often believe ourselves to be the eternal exception to every rule. Everyone is rather prone to this, even psychologists! A sample of psychologists were recruited who specialized in addictive behaviors. The average rate of relapse for people in conseling for alcohol abuse is about 20% (I’m not certain of this figure, so do not quote me on that!!!). The psychologists were asked to say how many of their patients would relapse, and these psychologists were all fairly certain that one of their patients, at most, may relapse (around 5% of their patients). After following up with them a year or so later, they found that on average, there was a 20% relapse rate. Just to be clear, relapse without counseling is much higher.
So what’s the point of this relapse example? Well these psychologists all believe that THEIR patients are exceptions to the rules because they are such good counselors. Literally everyone, including those of us educated in human behavior, will tend to think that they are the exception to the rule. Granted, these were clinical psychologists, not social psychologists. Obviously social psychologists would never make such silly mistakes (especially not me!!!!). Okay, okay, I guess we do make the same mistakes too…
Psychology proves things by…
The scientific method – this helps us discern which theories are logically valid and sound. It helps guide the research process. What makes a study of any subject a science? The method of course! I know you all have all heard about the scientific method since grade school. There is nothing new about it I could possibly tell you. I just want to say loud and clear that the scientific method is simply philosophy in action. The goal of this method is to create results that can be support and more importantly, replicated.
Peer review – This ensures methodology was appropriate. This includes random sampling, controls to ensure the independent variable was the only thing affecting the dependent variable, justification for your design, and historical evidence to support the claim. If other psychologists do not think your design was strong enough to truly test your theory, they will not allow it to be published. Further, a scientist from outside your discipline also reviews your work. So if they feel your statistical methods were not good enough for their standards then they might reject the paper.
Even with these guidelines and safeguards, we have to keep in mind that theories are rarely proven in any science. When a study “supports” a theory, it simply means it attempted to disprove it, but failed. This is good news for a theory, but why is falsifiability important?
Let’s imagine someone creates two theories for how we choose what we will eat for breakfast. The first theory looks at genetic predisposition for early morning calorie intake, behavioral history of eating breakfast, and what foods I had the previous day. We can easily test this theory by having groups keep food diaries to track their eating behaviors, take some samples for genetic testing, etc. The second theory posits that one’s hunger is derived from little demons that are invisible and live in your belly. They are only happy when you eat too much, so they try to get you to have awful foods in the morning when you are too sleepy to think straight. These demons are actually unable to be seen, heard, or otherwise tracked with any technology. This theory is impossible to test, thus it lacks falsifiability. Why would lacking falsifiability be a problem?
If a situation cannot be created where a theory can be supported or disproven, then it cannot be used to predict anything. So it could be true that ethereal demons live in our bellies controlling our hunger, but it’s a poor theory. Perhaps one day our technology will be strong enough to find the little buggers, but for now let’s resist the urge to believe nonsense. With the genetic based test, we could create a situation to disprove the theory. I find what I suspect to be the gene for wanting toast. I create the hypothesis that all people with the gene will have toast for breakfast if they have bread to make it, and a toaster. I run the study and look at my results. It appears that 50% of all people ate toast, regardless of the presence of the gene. Darn the luck! It appears that the gene I had predicted had no effect! It was a testable theory, therefore a stronger theory than invisble hunger demons. That does not make it correct, as we just saw!
What we need to understand is that most real psychological theories are based on peer reviewed research. For it to become published, it must go through falsifiable testing, be supported by strong research design, and validated by statistical analysis. Every now and then something slips through the cracks that was way off base, but given the competitive nature of reserchers, poor theories get torn to bits! When we use statistics, however, strange results do pop up now and then. Replication is key to prevent this from getting out of hand.