Sunday, April 3, 2011

David Berliner, Wile E. Coyote and Campbell's Law

However beautiful the strategy, you should occasionally look at the results.

-Winston Churchill




The clip above is an excerpt from David Berliner's keynote here.

Where there's smoke there's fire; and where there's high stakes standardized testing there's cheating.

We can bemoan this inconvenience all we want. We can play the blame game until we are blue in the face, but it won't change a damn thing.

The Emperor, in Hans Christian Andersen's Story of the Emperor's New Clothes, was naked whether he liked it or not; and high stakes standardized testing corrupts absolutely whether policy makers and education pundits like it or not.

Like the Emperor's chamberlains who walked along holding up the train that was not there at all, high stakes standardized testing has its own entourage who blindly hold up it's fraudulent vail.

Andrew Rotherham from Time.com has vail in hand:
Critics of today's push for greater accountability are quick to argue that cheating is the inevitable by-product of any high-stakes system. That's ridiculous. While fraud is a fact of life, there are numerous professions with far-reaching consequences for performance in which cheating is not rampant. Besides, that argument insults teachers by implying that they can't achieve challenging goals without cheating.
What's insulting and inaccurate is that Rotherham makes the assumption that chasing high scores on bad tests is a worthy use of our finite resources; and even if there was a consensus towards doing so, it is asinine to believe that we can simply mandate a social science law such as Campbell's Law out of existence.

Education deformers can no more skirt the real world ramifications of Campbell's Law than Wile E. Coyote could avoid the punishing effects of gravity.

If we are too move forward, we all better understand Campbell's Law at least as much as we understand gravity.

Campbell's Law states:

The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it was intended to monitor.

Here are all the posts I have written on Campbell's Law:

Campbell's Law and Standardized Tests

Regression to the mean

High Stakes Testing's Kryptonite

John Merrow: Value added or value lost

2 comments:

  1. In Nichol's and Berliner's book, Collateral Damage they expand on the idea of Campbell's Law and link it to national testing via NCLB. An important read.The issue is not testing. The issue are the stakes we and other's attach to the test. Look forward to reading your other posts.

    ReplyDelete
  2. I think Rotherham also distorts the issue: "Critics of today's push for greater accountability are quick to argue that cheating is the inevitable by-product of any high-stakes system." -- In fact, don't the critics say cheating is inevitable when the high-stakes are attached to a single, corruptible tool or measurement? If the stakes are high but you have multiple and appropriate measures, the cheating may not be inevitable.

    Rotherham adds, "there are numerous professions with far-reaching consequences for performance in which cheating is not rampant." That's a difficult claim to evaluate, especially as written, but let's assume it's true for the sake of argument. He has shifted the discussion from the abuse of very narrow, easily-manipulated measures to the broader concept of "far-reaching consequences" - a much different view.

    ReplyDelete