To steal from Tom’s post a few days ago, I too wonder “How I did” this school year. Since my evaluation was likewise “satisfactory,” I thought I’d consider the question how a state government might: through test scores.

Colorado has joined with a few other states (Florida and New York are among those with plans in motion) to tie a teacher’s continued employment directly to test scores. It appears that student test scores must comprise “at least fifty percent” of the evaluative criteria for teacher tenure and retention. If improvement is not sustained, a teacher can lose tenure and risks being fired. That would certainly align with an “unsatisfactory” review…potentially sparked by poor test scores. 

As I read the article, it stated clearly the bill calls for teachers to demonstrate student growth. I’m not familiar with the Colorado assessment system, and a half hour of wading through the web didn’t net me many answers. I’m a skeptic of that word growth, however. Something tells me we’re not talking about a preassessment in September and a postassessment in June, which is the only kind of assessment of growth I’d feel comfortable tying to teacher pay and continued employment. The old argument of comparing apples to apples is key. If we’re comparing apples to oranges, then ready the court for appeals.*

In a once-a-year test situation, how can growth be assessed? Let’s trace it out and play the how I did game by considering my students’ performance on the recent High School Proficiency Exams (HSPE) in reading and writing and previous years’ Washington Assessment of Student Learning (WASL) tests.

In 2009, my own sophomores had a pass rate of roughly 92%** on the state reading test–let’s refer to this cohort as the “apples.” Of 2010’s crop of sophomores, the “oranges,” roughly 82% passed reading. Still good, I’m proud of them, they actually did better than I would have anticipated back in September. Hold on though, when I compare apples to oranges, the data suggests that I somehow made the oranges dumber. Does this 10% drop mean I am a failure?

When I went back and compared their sophomore scores to their middle school scores, the last time they took the standardized test in Washington, something else crazy happened. First off, not all my kids were even in Washington in middle school so that certainly skews the data. Disregarding that (because what else is good about data other than that we can pick and choose which data to use?) the awesome apples who hit 92% with me? Well, in middle school, 94% of mine passed. The oranges, the 82%? In middle school only 68% of my crew passed. From this data, I apparently failed the apples but served the oranges better.

Want to really complicate things? The apples scored better than the oranges on the 10th grade reading test–but the oranges scored better than the apples on the 10th grade writing test. For both cohorts’ writing test, the scores were essentially unchanged from middle school. So the data shows that I improved performance, hindered performance, and held performance stagnant…all at once.

I’m so confused! If test score data is supposed to be how I measure my effectiveness as a teacher, I can’t tell if I’m a savior or if I suck. 

As of right now, Colorado’s move is only to tie teacher tenure to test scores–and I’m sure that advocates for the profession are asking the same questions I am. I’m fine with assessing teachers, but this is the wrong means of assessment.

More important than all this: people see this in the news, they see the teachers and the union resisting, and they assume that that resistance is a matter of fear of people losing jobs. It isn’t. I and many other teachers would love to see an effective system for (1) remediating or removing ineffective teachers and (2) cultivating the potential of all within the ranks. People think that the argument over test scores and teacher performance is about pay or retention or tenure. It isn’t. It is about an ineffective and invalid means of assessment. The tests are intended to assess student performance–and to assumptively extend the thinking that the same assessment also is valid for measuring teacher effectiveness is flawed logic which does not yield reliable results.

To attach the assessment of one’s teaching to test scores is simply the wrong approach. The teachers and the union are resisting not to protect bad teachers, but to communicate that once-a-year test scores are not a worthwhile assessment of a teacher’s effectiveness. Teachers and the union are not opposed to a fair assessment of teacher effectiveness: we’re just saying that test scores are not it. 

So how did I do? If we just look at the test data, the results are inconclusive…I’ll probably need to give back those Teacher of the Year awards. But since the data doesn’t do much for us, how about this: let’s take the September writing samples and compare them to the June. Let’s take the September reading comprehension assessments and compare them to the June. I look at that, and I know I did well. More importantly, my students did well. In that, I can see growth, I can hold the evidence of it in my hands and in my head, not just in a chart or graph.

____________________________________________________

*Bad pun. Sorry. I honestly didn’t mean to.

**I’m lucky to be in a district that has superior elementary and middle schools which therefore deliver me students who are well prepared for high school and the high school exams. My numbers look good, but much credit goes to the teachers who came before me…wait, how does that skew my data? Not so simple, eh?