Last updated: Fri, Jun 30, 2017
In the first section I wrote and wrote about the physiology of pain. Now you should know quite a bit about it.... So imagine that you are a pain scientist and you would like to do some research into pain. How would you define “pain” for scientific purposes? Could you write an operational definition that would allow you and other scientists to distinguish pain from similar things that aren't pain? How about a definition that would allow you to measure how much pain was associated with some phenomenon? How about a definition that would allow you to compare the amount of pain from one situation with the amount of pain from another situation? What critiques would be made of your definitions?
As an easier problem, next suppose that you are a psychologist who is concerned for people in pain, and you think you've noticed that people with exaggerated responses to their pain don't seem to be doing themselves any good. How might you turn this perception (or hunch) of yours into a testable theory? You must define what you mean by "people with exaggerated responses." You must be able to distinguish these people from others or be able to measure how much people exaggerate. You must develop and state a theory that expresses clearly what you mean by "not doing themselves any good." You must test the theory. Others must test it. If it checks out, you or someone else can turn this into a medical treatment.
These are the types of problems that pain researchers face.
If you've learned about the scientific method, you've learned about hypotheses and experiments. The process goes like this: The scientist formulates an hypothesis. He/she develops an experiment to test the hypothesis. The scientist publishes his/her results, and other scientists read about it and replicate the experiment, or perhaps devise their own experiment to challenge the hypothesis. Thus the hypothesis is proven or disproven, and the process begins again.
I want to make just two points about hypotheses and testing. One is about the relationship between definitions and experiments or hypotheses. The other is about what it is that is really tested.
It should be evident that what you will prove depends to a great extent on the model you test. The definitions of terms are part of the model. Referring back to the imaginary psychologist, he/she will find that what is true about "people with exaggerated responses" depends on exactly which people this includes. When all the experiments are done and your hypothesis has perhaps been proven, you will have proven it with respect to the set of people specified by your operational definition. If you succeed in developing a treatment, it will be used for that set of people.
Taking this a step farther, suppose you refer to this defined group as "exaggerators" among your experimental team and your scientific peers. It would be incorrect to then apply the result to someone who exaggerates his/her fishing stories, unless of course your operational definition of "people with exaggerated responses" includes fishermen.
In many cases the definition of terms is very important in determining both the cost of an experimental endeavor and the practical value of the result.
My second point about hypotheses and testing is about the difference between a demonstration and the rigorous test of an explanatory model. It wouldn't, I shouldn't think, be very difficult to show that in some sense "people with exaggerated responses to their pain don't seem to be doing themselves any good." You could make it simple to show this with appropriate definitions. For example, you might take your initial idea literally and test whether they “seem to be doing themselves any good” by asking yourself whether it seems to you that they are doing themselves any good. You might think of this approach as one end of a continuum. At the other end might be a model that reveals important things behind and beneath your intuition, something important about exaggerating and its relationship to experienced pain. Some of the questions that might be asked and answered could be:
This example about exaggerating patients is contrived but is inspired by real published research. The continuum between the demonstration of an intuition and the proof of a deeper explanatory model is worthwhile to bear in mind when thinking about research results. On one end you find low cost and low explanatory value, while on the other end you find more cost and altogether a more difficult task, but also much greater explanatory power.