Value-added evaluations

I've been following along on the discussions of how to evaluate teachers, and how using kids test scores doesn't show appropriately how well a teacher teaches. After all, disadvantaged students just have so much against them to be able to get good grades, goes the argument, so teachers can't be held responsible for students not doing the work.

But an article in yesterday's LA Times Superintendent spreads the gospel of 'value-added' teacher evaluations proves to me that standardized test scores can be a legitimate way to evaluate teachers. In this case, it isn't the actual scores that are used, but a ratio of individual student's scores over time, which will show students' progress in relationship to themselves. Studies have shown that teachers really do affect students' progress, and that their effect can be measured. If a whole class gets much better Value-added ratios, then the teacher must have had something to do with it. If some improve and some don't, it is probably the students' own effort and not the teacher that makes the difference.

Maybe not surprisingly, the students from underachieving school actually show the most progress (since a tiny improvement is huge related to very little) while good students from privileged schools have a hard time showing much improvement at all, since it is related to a much larger base number.

Personally, I would like to be able to see how my teaching affects my students' improvement. I could use it as a formative assessment that tells me to try harder.

I could see how various students perform to understand in which ways I am successful, and where I need to try different methods.

The concept is vaguely like a grading system I developed years ago for a class in Denmark which could vaguely be compared to a continuation high school here. They only received final test scores for their record, so I was free to give them formative assessments any way I wanted. I invented a system with arrows: up if students did better than usual, down if they had slacked off. Initially, of course, the better students didn't like getting a down arrow while someone who did far less got an up. But gradually they figured it out, and used it to motivate themselves to do even better. At the same time the students at the lower end were not getting F's, but were getting encouraging up arrows. The dyslexic student I started the system for had been given up by previous teachers. But she went from what would have been F's to a final test grade of about C!

Knowing where I'm doing well and where I need improvement would be great. And if it turns out I'm not a good teacher anyway, I don't want to inflict poor teaching on students who deserve better.

However, another article on the same page Judging teachers: Much of what you thought you knew is wrong reviews a few misconceptions about teachers' effectiveness, for example:

Teacher experience matters. Although teachers are generally paid more for years of experience, research suggests that instructors show dramatic improvement in their first few years and then level off. Teachers with 20 years of experience are often no more effective than peers with five years.

Teacher education matters. Schools routinely pay teachers higher salaries for obtaining master's degrees. But several studies have found that educators with advanced degrees do no better than those without (with the possible exception of high school math teachers) - my emphasis.

So maybe if I teach 5-10 years I will be able to encourage, motivate and stimulate my students, and then leave before I get too set in my ways and lose my enthusiasm! Another misconception from the list needs to be taken with a grain of salt - and good sense:

Class size is key. Research suggests that modest changes in class size, such as decreasing it by four or five students, has been shown to have little to no effect on student learning.

Adding a single student or two to 25 might not make a difference, but adding gradually until you get 35 or 40 students will most likely show a great difference over time (which would also be reflected in Value-added scores.) Don't use that as an excuse to push more students into classrooms designed for far fewer students. Teachers need to be able to move about the room, get close to students (proximity is a major element in classroom management!) and there has to be room to move desks into different configurations, for collaborative learning groups, areas for special purposes, etc. A classroom is not a lecture hall!