Or, Do PhDs roll Twenties?
Wow. This L.A. Times story — a version of this much better reported year-old piece by Scott Jaschik at InsideHigherEd — describes a colossally bad idea on which ETS (Educational Testing Service) is trying to sell graduate admissions deans.
Put very briefly: ETS has come up with an additional form for graduate school applicants (and paying customers of the GRE) to give to their recommendation-letter writers. The form asks recommenders to rate the applicants on “a scale of 1-5” on their abilities in “knowledge and creativity, communication skills, team work, resilience, planning and organization, and ethics and integrity.” (1) These ratings are then put through some kind of algorithm to produce a “PPI” score (Personal Potential Index), which purports to measure the applicant’s “non-cognitive” qualities. These, in turn, will supposedly enable admissions folks to determine whether or not the applicant is likely to complete graduate school, like, ever. (Only 57% of admitted students actually do, you see).
I won’t get into all the ways that this is a terrible, horrible, no good, very bad idea. But to quickly run down a few objections: don’t most schools already require similar rating forms in their grad applications? how can “knowledge,” “creativity,” or “planning” be non-cognitive? how would a prof. know about a student’s “resilience,” anyway? does ETS have data that backs up the correlation between PPI and grad school success?
(Other arguments against this are left as an exercise for the reader).
The more interesting question, I think, is what the ideal score is. ETS is probably going to be all boring about it, and say the best score is the aggregate of 5’s on every measure. But before we follow them in that staid path, please consider the following:
Exhibit A, from the InsideHigherEd article:
Applicants will be rated on a scale of 1-5 on questions about their abilities in these six areas: knowledge and creativity, communication skills, team work, resilience, planning and organization, and ethics and integrity.
Exhibit B, Prof. Wikipedia on “Attributes”:
An attribute is a piece of data (a “statistic”) that describes to what extent a fictional character in a role-playing game possesses a specific natural, in-born characteristic common to all characters in the game. That piece of data is usually an abstract number or, in some cases, a set of dice. Some games use different terms to refer to an attribute, such as statistic, (primary) characteristic or ability.
Most RPGs use attributes to describe characters’ physical and mental characteristics, for example their strength or wisdom. They often influence the chance to succeed in skill or other tests by addition to a die roll or by determining the number of dice to be thrown. As a consequence, usually a higher number is better, and ranges can be as small as 1–5 (for numbers of dice) …
Seems to me the successful PhD student — defined as one who completes his or her program on time (to be clear, we’re talking about a successful doctoral student, not an ideal human being) — would be high in knowledge, creativity, organization, and resilience (duh), middling in communication and team work ( best to work alone; conference papers needn’t be good; and we’ve all seen the writing in monographs), and low in ethics and integrity (you got scooped!).
The ideal grad student is a rogue, in other words. Or maybe a wizard.
I think grad deans could save themselves, and applicants, a bunch of money and time, by just going here.
Image cite: Slayer23, “20 sided dice,” Flickr, CC License
1). Scott Jaschik, “Non-Cognitive Qualities Join the GRE,” InsideHigherEd.com, May 22, 2008
2). Larry Gordon, “Graduate test goes where GRE doesn’t: personality,” Los Angeles Times, June 22, 2009.