Mindware for Hiring

Happy February! Tis the season for writing recommendation letters for students. This made me revisit thinking about the process of reading applications, interviewing candidates and making potential hires. While my department completed its interview and hiring process last November (I posted some thoughts about it here), many other academic institutions are in the midst of the process right now. Today’s blog post revisits this topic in light of my recent completion of Richard Nisbett’s Mindware. For a discussion of one of his earlier chapters, see here.

What did I learn, or what principles were reinforced, when reading Nisbett’s book? For one, there is the halo effect, where certain impressions of a person (possibly formed very quickly and in a very specific context) can strongly color one’s view of a person’s actual capabilities. The impressions can be positive or negative. In some cases the impression is in line with actual ability, but in other cases it is not. Cognitive research suggests we are often not aware of how easily we can be influenced by such impressions, and how they can lead us to “see things as we want to see them”. In academia, institutional prestige seems to hold unwarranted currency. Being from a prestigious alma mater isn’t necessarily a good indicator of a capable candidate and should not be over-weighted. Note to self: The next time I’m interviewing a candidate I should carefully introspect to see if I’m being heavily influenced by a halo effect.

In a chapter titled “Spilt Milk and Free Lunch”, Nisbett introduces some economic principles: sunk costs and opportunity costs. According to Nisbett, the expression “there ain’t no such thing as free lunch” comes from the Depression-era, where bars would advertise free lunch but you had to pay for the beer. Here’s what he has to say about considering opportunity costs when hiring. “It can be sometimes quite difficult to see that the value of the unchosen alternative is actually greater than that of the chosen alternative. Every hire you make for your company constitutes an opportunity cost. If there’s no one capable who can be hired, it’s tempting to feel that noting has been lost. But if there are good reasons to believe that in the near future someone more qualified could be hired, then the present hire involves an opportunity cost to the company that might indicate that hiring should be put off.” This is particularly challenging in academia, when you have to fight for your tenure line searches, and you feel great pressure to fill the position when you have the opportunity. Note to self: Unless the candidate is stellar, consider the opportunity cost for hiring a good, but not a great candidate.

In a section titled “The Interview Illusion”, Nisbett cautions us to pay attention to variability and the law of large numbers. He relates examples from graduate school admissions where professors (yes, in a psychology department no less!) “tend to put substantial weight on the 20-30 minute interviews with each candidate. The problem here is that judgments about a person based on small samples of behavior are being allowed to weight significantly against the balance of a much larger amount of evidence…” This includes GRE scores, GPAs, essays, recommendation letters, and prior successful research experience. Apparently 30-minute interview predictions have a much lower correlation (0.1) with actual performance, while other measures do quite a bit better. Note to self: Be careful not to overweigh the short one-on-one interview over other evidence in the application file. Note to larger group: Be sure that the candidate is observed in multiple contexts by multiple people as much as possible.

Nisbett argues that “coding is the key to thinking statistically”. In cognitive experiments that compare our predictions/extrapolations based on measurable quantities versus estimating behavior/traits, we do decently on the former and rather poorly on the latter. Examples given include grades on spelling tests and points scored in basketball games, both straightforward to code; versus rating friendliness or honesty in encounters, both difficult to code. Context often dictates how you might perceive less easily code-able characteristics. There’s no good solution. Nisbett advises that “the most effective way to avoid making unjustifiably strong inferences about someone’s personality is to remind yourself that a person’s behavior can only be expected to be consistent from one occasion to another if the context is the same. And even then, many observations are necessary for you to have much confidence in your prediction.” Note to self: Distinguish between codeable versus non-codeable traits. It isn’t the case that the former is good and the latter is bad. Just don’t overweight the latter, because humans have a tendency towards doing so.

Several chapters discuss the effect of different cultural backgrounds on attitudes, behaviour, and tacit assumptions. This can be tricky to tease out because we’re often unaware of our own reference points, and how different they might be from others. Note to self: Be aware. Know thyself. Don’t superimpose thy own mindset on others. Listen more.

This leads me back to the recommendation letter. How useful is it? I don’t recall reading any “bad” recommendation letters from applicants although I have seen some lukewarm run-of-the-mill ones. (I also learned that the style of letter-writing even in the same language group can be different. For example, American recommenders were generally more effusive than their British counterparts.) Accomplishment data points in the curriculum vita are helpful, but it is also help to put them in context. That’s what a good recommendation letter can do, provided the recommender has had ample opportunity to observe the candidate, preferably under multiple contexts and occasions. But after reading many, many examples of glowing letters for “top” candidates, it can still be a challenge to pick out the stellar candidate. Some even give me pause – is this person really that awesome, littered with superlatives? Maybe. Candidates would (or should) only solicit recommenders who can provide strong positive letters. We give our students the same advice. I guess the way to think about them is as one among many pieces of evidence used to evaluate a potential hire. Note to self: Be a conscientious writer, and try to provide good solid evidence that an employer can use to hire (or accept into graduate school) the person I am recommending.