During today's debate, when Mike Huckabee said: "I had also the most, I think, impressive education record," Mitt Romney struck back by pointing out Massachusetts' test scores were superior to Arkansas scores.
But the question remains; how valid were the tests? A Townhall staffer emails me his disturbing experience grading tests for the company that graded Massachusetts' standardized tests:
During the summer of 2005, I took a temporary job working as a test scorer, grading the English exams for fourth and eighth grade students from Mitt Romney’s Massachusetts. Strange that, in this most recent debate, Romney mentioned how his state ranked first in both of these tests. My experiences pertain to those tests taken during the 2004-2005 school year, but I doubt things have changed.[# More #]
Essentially, students were asked to write a response to a problem-solving question (for example, given these 10 items, how would you build a shelter?). They were scored on a scale of 1-4 with a strange breakdown of the numbers. A 1 made no attempt at answering the question or even bothering to use the English language – these were the ones that looked as if they had been written by a kindergarten student. 2s would often make some reference to the materials with little sentence structure and a wrong answer. To earn a 3 (an above average score), the students had to either get the question essentially right without any resemblance of English or have reasonably logical thought process without the correct materials. Those which earned a 4 were where students got the question right and had utilized a fourth grade standard of the language. Strangely, we (the scorers) were actually told mid-way through the project that we had been scoring too harshly, where almost every student should be getting a 3 or 4 unless it was absolutely terrible.
The eighth graders were similar, although the question was a response to a poem as opposed to problem solving. But again, the grades were rarely deserved and very much inflated. The standards were far from accurate. Not to mention the level of education required to score this project (two college level English classes – no degree of any sort required) with very little training. We scored between 500 and 1000 scores in a 5-hour shift, where I can almost guarantee the mind began to wander after only a few exams. The margin of error was ridiculous, as was the scoring scale in of itself.