Thursday, April 4, 2013


An outlier is someone or something that isn't part of the group, or in statistics, someone that doesn't follow the common pattern or trend.  In my psychology statistics class, our professor showed us a graph of test scores on introductory psychology exams from previous years, which showed that most students' grades increase from the first exam to the second. There were only 3 students whose grades were much lower on the second exam, whose data points did not fit into the pattern of the others. Our professor explained that if we included those 3 people when making a of line best fit from the data points, the entire line would change and would not accurately represent the majority of students. Therefore, we exclude outliers from the line graph.

I understand why people exclude outliers for statistical trend purposes, but seeing those 3 people get dropped made me cringe. The idea behind excluding outliers is that these people have circumstances that aren't related to the study. Ex: if someone was sick the day of the second exam, they probably wouldn't do as well on the test as they normally would have. The key phrase here is as they normally would have. We're assuming that these 3 people had unusual circumstances that caused them not to follow the trend, but if not for those unusual circumstances, everyone would have fallen into the same pattern. But we don't know this for sure. I personally don't fall into this trend. I usually do the best on my earlier exams because I get busier as the semester goes on. This pattern is normal for me. Maybe it's normal for those 3 outliers as well. Maybe there's a whole group of us whose grades follow a different pattern, but we're written off like our experience doesn't matter.

(Trigger Warning: suicidal thoughts) Someone I know was taking medicine for a physical problem, when she began feeling depressed and suicidal. No one could figure out what was wrong, and she thought the medicine might be making her feel that way. As it turned out, the medicine did have a very rare side effect of making people feel suicidal. The doctor explained that several people who had taken the medication were documented as experiencing the same symptoms, but since such a small group of people had these symptoms, the people were written off as outliers, as exceptions to the norm. As a result, none of these people were informed about this potential symptom.

I think the reason I felt so defensive of those 3 outliers in statistics class is because I was an outlier at the time. My college was rated one of the top schools for student happiness, and almost everyone who goes there loves it. I didn't like college, and when I talked about my own experience, no one believed that the school was the problem. I was constantly being invalidated, being told that I must have some other problem because my college couldn't possibly be that bad. Anytime I filled out a survey about my college experience, I worried that it would get discarded. When you're the outlier, no one believes that your experience is real.

I understand that most studies won't cover every related issue. Our professor wasn't trying to get into all the individual and personality differences that make students' test scores follow different patterns; she was probably trying to figure out what would benefit most students. I'm not saying that people need to alter lines of best fit, but people do need to acknowledge the experiences of outliers - not just the outliers who were sick the day of the psych study, but the outliers who, under typical circumstances, actually follow a trend that differs from the norm. We exist, we are real, and our experiences are just as valid as anybody's.

No comments:

Post a Comment