High School Reunion
by Katie Encabo
I attended my ten-year high school reunion a couple weeks ago. While I was initially reluctant to purchase a ticket when I first heard about it, it turned out to be an entertaining way to spend an evening.
One reason I was a little hesitant has to do with my working theory that Facebook has and will replace the need for high school reunions. Digital social networks allow you to: 1) keep up with your current friends, 2) connect with people from your past with whom you want to get back in touch, 3) learn what your old acquaintances are up to, and 4) avoid the awkwardness of being forced to talk to old crushes or rivals. Isn’t all that what high school reunions are all about? As far as I could tell, Facebook’s got us covered.
But, surprise! Going out on a Friday night with a group of old friends was actually much more enjoyable than an evening spent browsing Facebook.
I currently live in the same city where I attended high school, so I’m still in touch with a couple of classmates who also live in the area. I also occasionally run into former classmates at the grocery store, restaurants, or sporting events. I sometimes dread running into these people unexpectedly (does she recognize me? Will we have to make small talk? Do I remember his name?). But seeing them all in one place, on purpose, was much less awkward. We had all voluntarily made a choice to intentionally run into people from high school by attending the event.
I found myself reflecting on that fact while at the reunion: all of the former classmates I was talking to had chosen to be in attendance.
If I had gone to the reunion as a market researcher wanting to get a good overview of my graduating class, the fact that the group of alumni in attendance were there voluntarily would definitely need to be taken into consideration and could actually hinder my research of the group. Sure, a good sample of my classmates were there–I think one of the organizers told me that nearly half of our graduating class showed up on that Friday night–but we would need to account for self-selection bias.
For example, based on the people I talked to, I would say everyone has interesting-sounding jobs. But perhaps people who chose not to attend didn’t come because they were self-conscious about not having an interesting job to talk about! So, any claim about percentage of graduates with interesting jobs could be colored by self-selection bias, or a distortion of statistics when the respondents who chose to participate in a group or activity are systematically different to the overall group we’re examining.
I also noted that I only talked to one former classmate who has a baby–no one else has kids yet. But, I did hear about five or six people not in attendance who are now parents. By hosting the event at a bar on a Friday night, people with babies were possibly in lower attendance due to the burden of securing childcare, or just the life-changing nature of having a child. The trait I was interested in measuring at the reunion (what percentage of my classmates have kids now) is impacted by the fact that they do have kids and could potentially not be there because of it.
I might have said, oh everyone I talked to looked great! My classmates are aging really well! But is it possible that those of us who do feel or look like we’ve aged beyond the ten years–already with a head of grey hair and lots of new wrinkles–chose to stay home?! Of course making speculations about those who didn’t attend is dangerous and theoretical. But my point is that you can get into just as much trouble by making assessments of a group right in front of you. If it’s not a random sample, but instead a group of self-selecting volunteers, you must be careful not to draw conclusions about an entire group based on only those who choose to attend.
The best way to measure the characteristics of a group is to be as random as possible with your group representatives. With this logic, running into people in the grocery store (assuming I ran into enough of them), while more awkward than the high school reunion, might actually give me a better sense of the class as a whole. Both class reunions and social media only include those classmates who are choosing to volunteer information about their lives.
Luckily, I was able to take off my market researcher hat for the evening and enjoy myself as a Lincoln High alum, not someone too caught up in the statistical significance of what my graduating class had amounted to. But in research and in our natural inclination to try and make simple assessments of a group, remember that we can often forget to account for self-selection bias.