Andy Kessler’s analysis of the way research is conducted was very interesting to me. Some of the points that he made were ones that I had never thought of before, but are really important to keep in mind when looking at research and correlational studies. The statistic that stuck with me the most was “more than 70% of researchers have tried and failed to reproduce another scientist’s experiments (…) and more than half have failed to reproduce their own experiments.” These ratios astounded me; I would not have guessed that so few researchers would be unable to replicate research when studies are designed with replication in mind. This made me doubt how accurate research really is if it hasn’t been reproduced multiple times. It’s easy to see one study and think, “since it got those results, they must be true,” but if that study was performed again, there’s a very high chance that those results could be completely different. There are so many confounding variables that it is nearly impossible to account for all of them, which could lead to conflicting results of replicated studies.

I strongly agree with Kessler’s remark that “correlation doesn’t equal causation,” which is a phrase I’ve heard over and over again in science and statistics classes. It’s important to understand the impact that confounding variables can have on results and recognized that even research designed to be objective can still fall victim to inherent biases. I also believe that it is incredibly important for research to be reproduced multiple times before it’s accepted as truth.

When looking at research for our paradigm shift essays, we should all be careful when including studies that have not been replicated. For my essay specifically, I have found several analyses of comics, but if other researches repeated these studies, they would find the same results (if they didn’t use different methods) since all of the data being analyzed is consistent and wasn’t produced in an experimental context. But, I will still try to verify any research that I find by looking for multiple sources.