The Facebook friend finder

Abstract

Listen

Translate article

Riveted to my seat, I listen intently to the latest advancements in diagnostic techniques for prostate cancer. The lecturer then starts to explain the qualities of the tests—sensitivity, specificity, positive predictive value, negative predictive value . . . and I start to doze off. As a medical student, these terms often made me switch off my attention. I would just give up on understanding concepts, and merely learn definitions by rote and equations before exams, which of course were promptly forgotten afterwards. In addition to being so abstract, terms such as false positives and negatives made it hard to get a grip on things. Sure, they made sense after I had thought it through, but the effort needed to understand them was galling. Even Wikipedia fails in this respect and just provides dry definitions. Using examples that students can relate to would make things clearer and allow them to remember not just formulas but concepts that can be applied throughout their medical practice. One example is Facebook. Remember the first time you set up your Facebook account? After the initial profiling, Facebook comes up with a list of suggested “friends” whom you might know. Although almost accurate to the point of creepiness, it isn’t perfect, excluding that hot acquaintance you wanted to know better and suggesting some people you wouldn’t be seen dead with. This concept of the friend finder identifying your real friends and excluding others is similar to a screening test, which detects disease (your friends in this case) and excludes those without disease. ### The ideal friend finder—sensitivity and specificity Ideally, a perfect friend finder would be one that fulfils two criteria—it includes all your friends and excludes everyone else. Criterion one is referred to as sensitivity. When a high proportion of your friends are identified, the friend finder is considered to have a …