Education Reform

In pondering the usefulness of standardized testing, I came to something of an idea. What if instead of having standardized tests, students were simply evaluated on their grasp of the skills necessary for their current grade, and this evaluation was used to grade their previous teachers. Given the proper algorithm, the data set should be large enough to address whatever data related issues may arise such as differences in what constitutes an “acceptable” level of knowledge in students, or forces that influence performance which lie outside the prevue of the education system.

This would seem to solve the problem present with using standardized testing of teachers simply teaching to the test while leaving their students without the proper fundamentals. It could also be an impartial method of calculating pay for performance, especially if a teacher is able to improve the record of a student that is habitually underperforming.

Now granted, this isn’t really something which could be rolled out over a short period of time. It will take a reasonable number of students, and time to properly calibrate the algorithms before any findings could be made with any reasonable degree of confidence. This seems more like a project which would require financing for a long timeline from an organization such as the Gates Foundation. It would also require the buy in of the teachers, as evaluating incoming students against their baseline of what is acceptable would be a time consuming process.

At its most basic level, this system could be used to remove or retrain teachers who are failing their students and to monitor performance across a large education system. Beyond that, with a sufficient volume of data, I believe that the system could align students who learn a particular way with teachers that would be able to teach in a way that they can learn more easily from. While this may have been relatively useless even a decade ago, enhanced communications technology means that a greater pool of potential teachers is available now because of the ability to communicate without being in the same physical location. It may also even be possible to identify broader trends which might only become apparent once a large volume of data has been available for analysis, or even to track historical trends. By collecting this large amount of data, and aggregating it centrally, different educational strategies and initiatives could also be assessed against a much larger data set.

What this would really require is the buy in of a large school board to ensure that the sample size is large enough to be useful. It would also be important to be able to deliver some sort of preliminary and ongoing reports to demonstrate the utility of the mechanism.

The hardest issue I think would be setting up the evaluation framework. It would have to be the same across the area of data collection so that the data is consistent and useful. Perhaps it would just amount to a multiple choice survey that each teacher would fill out a month or so into the school year. The question then becomes which questions should be asked, and how are the answers evaluated.