04 December 2014

Misleading Science Data

Share
I read recently of a pseudoscientific experiment conducted on a college campus concerning how toilet paper dispensers unravel. The "researcher" did a blind test to see if there was a preference for whether the paper was dispensed over the roll or under it. Like many scientific experiments, this experiment included too many variables and lacked a control group for comparison. As is prone to happen in pseudoscientific endeavors, the person making the claims jumped to conclusions without considering other factors. Although interesting, it is hardly conclusive nor useful as a determinent of anything whatsoever besides the fact that toilet paper vanishes from the bathrooms.

Proper scientific studies involving people must inform the people that a scientific study is underway. There were zero signs indicating that our behavior was being studied, but thankfully there were also no cameras mounted in the bathrooms to violate our privacy. Had I known that they were studying something, this would have changed my behavior. If I knew I were studied, I would make sure that my preference were known, that I included all relevant items, and that I made sure that anyone who didn't know was aware that a change might result based on the data collected. Instead, they concluded that we always do exactly what we prefer. As a biochemist, I have taught biology labs, and we discuss altruistic behavior, which illustrates that sometimes people do what is best, even if it's not best for them, however rarely that may occur. Since that happens, it renders the conclusion irrelevant, because at least some people didn't choose always what they personally prefer.

Studies with too many variables often unjustly ascribe outcomes to the variable of interest when it's coincidental instead. The researcher made no mention of the sample size either in terms of users or bathrooms considered and failed to specify the limits besides "my campus". Well, even then, that's hardly telling, because that's not my campus, and the student body will change next year as will the one where I work as well. The researcher did not consider the disparate behavior of custodians, some of whom may have an already biased proclivity to install toilet paper rolls in a certain fashion. Additionally, it assumed that the rolls were used to wipe and not as tissue or stolen for personal home use or used to toilet paper a car, building, office, or person, which is a different set of outcomes altogether. The researcher failed to note whether the dispensors were all installed on the same side of stalls, which might change which roll one chooses due to flexibility rather than preference. Although one could conclude that this served as a benchmark for future studies, because they didn't reduce the variables sufficiently, we cannot conclusively link the outcome to the variable of interest.

Many pathways of both genetics and behavior intersect, but they do not always do so on purpose or for the same reasons. Even people who prefer the same things as you may prefer them for completely different reasons, sometimes reasons you find personally repugnant. Misbegotten conclusions project similar values, actions, and morays onto a diverse group of individuals, particularly when, in this instance, they are not told that their actions are being measured. I do not always choose toilet paper the same way. When I use the restroom, I sometimes notice that rolls offer both under and over paper dispensation. I prefer one, but I usually will take paper from the smaller roll, not because I prefer it, but because I know that the sooner the roll is spent the sooner that custodians will replace it with a fresh one. In this example, my behavior has nothing to do with preference of dispensation but with preference about having more toilet paper available on my next visit. What matters most to me is not how I obtain the paper but that it is there to obtain.

While it may be possible to do this scientifically in a blind study, this study was neither scientific nor conclusive. There was a time when there was no toilet paper at all. Furthermore, not all paper is the same. I will take certain brands over others due to factors more important to me than the manner in which I obtain them. This study presumed that the most important thing to everyone was how the paper came off the roll. How is irrelevant if I can't get it or if I can get it better elsewhere, and let's not forget all the times I enter a stall and change because some nincompoop unirated on the seat. I don't choose those rolls, not because of how they are rolled, but for sanitary reasons instead. Coincidence is not causality. People are claiming that it's false to claim that global warming is a hoax because it's colder this winter than normal when they use the same logic as evidence for it when it's absurdly hot in the summer. So many things they claim to be true simply vanish from the news. Lest we forget, swine flu, SARS and Ebola were supposed to wipe out life as we know it, and none of those are important anymore despite the vast army of scientists working on them. Proper science comes from proper methods. Since men are flawed, I don't expect our methods to ever conclusively answer anything, as if an imperfect thing could perfectly describe anything more complex or perfect than itself.

No comments: