High Stakes Testing

Regarding measurement of learning, I tend to be of the J. Gould camp. (Like you didn't know already.) But if there is a huge natural variance in humans, there could also be a huge variance not in what people know (obvious), but HOW people know.

Furthermore, it is possible that folks involved in measurement conflate the two, dismiss the difference, or miss it to one side or the other precisely because it is a moving target.

In order to get to information stored in another's brain - stored in an idiosyncratic fashion - a strict path cannot work in 100% of cases. Let's say I ask my class to reflect on why something is generally true. Nobody can answer, so I work backward from it and repeat the question till somebody gets it. Then I reflect on whether I left out a key bit of context or if the students were unable to make the connection I asked of them. BUT sometimes I am surprised by what they think and how they think it.

This doesn't happen often, but it may be happening more often than we think which is what makes me think my examination of student's knowledge isn't very good.

Supporting this is the social observation that people just don't make allowances for the way other people who are equally knowledgeable see things. Sure there is emotional investment, but there are inevitable differences of opinion engendered by what we call viewpoint. Am I saying viewpoint is a lot more dangerous than we suspect in testing? That makes me wonder about professors who give credit for well-argued opposing viewpoints.

Glen McGhee pointed out Ray and Mickelson's paper: "Students "read reality with more care than they read textbooks" (14, also 11-12) "Part of their world is the realm of work ... they observe their adult kin grappling with the financial difficulties ..." Applies now to college students, too."

This is a simple fact of viewpoint. Students want to work and react with what they think is appropriate strategy. Business leaders react in dismay, thinking students are out of touch with reality.