It’s Complicated: How Much Can We Know?
However, in science, an exact description of what is true is not always possible. In mathematics, 2+2=4 is always an accurate statement. But, as one increases the complexity of the science, going into physics and then chemistry, the “truth” becomes less obvious. For example, it is true that mixing hydrochloric acid and sodium hydroxide will produce salty water, but an exact description of what happens (possibly an explosion) requires one to also consider the concentrations of the reagents, the temperature, the air pressure, and more. No longer will a simple statement suffice.
In biology, the factors that must be considered are exponentially increased in comparison to chemistry. Essentially, biological systems are a complex mixture of chemical and electrical reactions controlled by application of many levels of information, not to mention the environment, so that predicting the outcome of changing one parameter can be almost impossible. The complexity, and thus the impossibility of drawing absolutely accurate conclusions and predicting the effect of a change in one parameter, further increases as one progresses into psychology, sociology, ecology and the like.
To illustrate this principle, we can consider the workof Dr. Carolyn Nersesian of the University of Sydney. This ecologist used a technique from chemistry (titration) to understand the feeding behavior of eight brushtail possums. Basically, she slowly increased the concentration of a poison in the food in a sheltered area (tree) while offering the animals untainted food in a less sheltered area that had been pre-treated with fox urine and feces The goal was to see what concentration of poison would cause the animals to risk exposure to predators by moving from the sheltered to the unsheltered area.
This is typical of scientific research–one parameter is changed while all others are kept the same. The scientist then observes what happens and assumes any change is due to the parameter that he or she altered. But, there are numerous potential confounding factors. In the example above, was the temperature in the unsheltered area different from that in the trees? Did the animals communicate, making the “decision” of one apply to all? Would the results have been different if urine from another predator had been used–or if there was no urine? The questions go on.
So how does this apply to scientific integrity? Simply, we must be aware that science cannot provide absolute answers at levels much above that of mathematics. And by the time one progresses to psychology and sociology, extreme caution must be exercised in interpretation of any data. After all, what would it take for a person to leave the safety of their home for the pleasure of Starbucks? Judging by the lines, not much. But, if it is cold and rainy…then titrate all you like, I am not going out!