Better Metrics for Education

Updated on  | By | Permalink

Why today's tests are unnatural and some thoughts on how to improve them.

Have you ever taken a standardized test in your life that you thought properly measured the skill it was trying to test? Ever in your daily work found yourself thinking, “Wow! This job is just like that test I took in high school!”? Unless you make your living as a Mechanical Turk, probably not. Standardized tests seem to be almost universally regarded as artificial, but what exactly is it about them that is so far from the reality they are supposed to simulate?

Why Standardized Tests are Unnatural

Wrong temporal scale of pressure
Time management is an important skill in any field, but the majority of exams probe this skill on far too small a scale. In life, time management on the order of days, weeks, and months is far more crucial to success than time management on the order of minutes and hours. Standardized tests demand that not a minute is wasted, that essays be churned out in 20 minutes a piece, and that quantitative results be produced so fast that reflection is impossible. Who works like that?

Too close to purgatory
Unless you’re a deceased pagan in 14th century poetry, you probably spend very little time working in purgatory. Why then do tests take place in empty, silent, foreign rooms, devoid of pleasant stimulations such as music, window views, or our favorite cushy office chair? There’s a reason that people are so particular about their work environments - because it matters. Learning research shows that information learned in a particular context is often best recalled in that same context. By placing students in such unnatural conditions, ones they are unlikely to ever encounter again, we’re placing them at an unnecessary (and uneven) disadvantage.

Isolated from other people
It’s time to give up the myth of the lone genius. Successful people, regardless of their field, know how to work with others. Yes, those people also know how to work independently and produce unique and creative contributions, but they balance their time. Tests too should strike a balance between collaborative and independent work.

Isolated from information tools
It’s also time to stop ignoring the fact that just about every worker in the world has Google at their fingertips. No one needs to memorize the periodic tables of elements, the quadratic formula, or how to spell “paramecium.” Far more important is how to find and use information. Let students bring books, iPhones, and laptops to exams and start testing skills that matter.

A Marketplace for Metrics
So what kind of metrics do we need? To argue for one or a few in particular would only serve to bias the educational system in favor of certain social classes or types of learners. What we really need is a vibrant marketplace of metrics. Allowing a single metric (i.e. the SATs) to so fully dominate a measuring process as complicated as predicting the lifelong success of an eighteen year old is naive and irresponsible. I doubt a small group of middle-aged white male “experts” gathered at a roundtable in Washington, D.C. are going to solve the problem either. Instead, we need a wide field of people with different strategies competing to provide better measures of student aptitude.

Exactly what types of metrics this competition would converge on, I don’t know. I imagine students taking one or more general exams as well as a few less common ones focusing on the particular strengths they would like to emphasize. This test selection process would not only be a great exercise for young students in reflecting on their own strengths and weaknesses, it would help schools and employers better understand which characteristics a student most wants to emphasize.

That said, there are two particular skills alluded to above that I would hope to see quickly worked in to some of the competing metrics - collaboration and long-term time management. I’ve never seen a standardized test that probed either skill and yet, I can’t think of any two more important to my daily success.

This rant was inspired by a lunchtime discussion with Phil Zager and David Livingston. Thank you to both for providing such great conversation.