The article that we read this week detailed 5 different questions that should be asked of scientific research to insure that it is valid and accurate. The first one is whether the experiments detect differences that matter. This is important because a lot of the time people get caught up in the tiny differences that occur between samples, and then use that difference to draw their conclusions. Asking this question helps to recognize if the difference really matters. Although a difference may be present, that does not mean that it matters in the context of the experiment.
The second question is whether the study shows causation or correlation. This is important to review because sometimes the inverse relationships aren’t explored, and so it leaves other variables open to influencing the experiment. This is extremely important as a question, because we have to insure that there is a true connection between what we are testing and the variable of interest. We see this in our class as well, we are told that we want to make sure our figures and data are significant, and we go out of our way to show this. The reason is that proving the connection is so important for future experiments, because causation may cause false positives.
Knowing the mechanism was another one of the questions. This one is important because just knowing that something happens is not usually enough. We need to know how it works to really draw useful information from the experiment. I think the author included this cause they know that a lot of the time people get caught up in the sensation of the conclusion, but can’t actually act on the information because it’s so incompletely researched.
Another question asked is whether the experiment reflects reality. We want the data of the the experiment to be applicable to real life. It might be interesting data, but there’s not much we can do with something in the real world if there is now application for it.
This also ties in with the question, could anything else explain the results. This is to eliminate other variables and answers to the question and solidify the results of the experiment. It helps to see if a study actually has data that is significant, or whether or not it could be tossed out due to other factors influencing the results.
All of these points are helpful in addressing controversy. I think the best one is the question of whether other things could change the results. It’s the easiest question to ask, and lets you quickly see if a paper has findings that are significant.
