Studying the daily practices of those teachers whose students register high standardized test scores in reading and math has begun to alter preparing, selecting, and evaluating teachers. Journalists (here and here) and researchers ( Allington) have reported (and shown on videos) what such teachers do to manage their classrooms as they teach content and skills from kindergarten through geometry and senior English. Data-driven instructional practice in action.
A short hop-ski-and-jump away are new policies to evaluate teachers on the basis of student test scores (often called “value-added measures”) and use of these data-driven practices in their classrooms. That is happening in Washington, D.C., Cincinnati, Houston, and other cities.
The road from data-driven instructional practice to evaluation and (don’t forget pay-4-performance bonuses), however, has potholes aplenty. Judging teacher performance on the basis of using data-driven practices and student test scores is highly contested. Dueling studies challenge the view that both approaches are scientific much like the controversy over global climate change or creationism. It ain’t pretty when researchers using gold-standard designs and methodologies hurl spitballs at one another’s studies. Ugliness aside, for now, bipartisan political support and dollars (both federal and philanthropic) support data-driven instructional practice and judging teacher effectiveness on the basis of student test scores.
When the current political climate favors top-down policies that have a weak scientific basis, what can teachers do?
Depend on their unions? Hardly. No collective bargaining contracts in southern states. Anti-union hostility at record high levels with abolition of collective bargaining for teachers in Wisconsin. Moreover, the AFT and NEA has affiliates that both support value-added measures (Denver, Cincinnati) and oppose it (Washington,D.C., New York City).
Depend on the formal data that comes from district and state tests? Not when recent reports point out that data-driven macro-reforms using penalties have yet to lift low-performing schools and districts to higher levels of achievement. Not when some teachers see the test as measuring the wrong things. In a previous post (May 12, 2011), I pointed out that such information can be useful in constructing lessons if it is timely, easily accessible in understandable formats, and help is available to figure out how to use the numbers to improve lessons. Of equal importance–if not more–are the data teachers collect from observing students daily engaged in classroom activities, teacher-made tests, and other informal ways of assessing individuals.
The dilemma, then, that teachers face is, on the one hand, valuing retail data that they receive daily from observing and interacting with students in lessons and, on the other hand, finding useful for their schools and classrooms the wholesale data that they can access from data warehouses on state and district tests and other formal assessments, some of which they may distrust. Teachers feel the pinch between valuing both informal and formal data points because they have only so much time, attention, and expertise in figuring out beforehand what all the information means for the next lesson or unit. And when the lesson begins, teachers have to manage seamlessly hundreds of on-the-spot content (do I introduce concept now or later?), context (divide class into small groups and have them use their laptops toward end of lesson?) and individual student decisions (ask Tiffany the hard question?) as the minutes tick by.
The tensions that arise within individual teachers and collectively among teachers on using data to construct and teach lessons echo conflicts that practitioners have faced and reconciled through compromises for decades: how much weight do I give to what I know from my experience teaching and watching students and how much weight do I give to the test results and other formal data (e.g., research studies, demographics)? Such tensions and the questions that arise from conflict, of course, uncover again and again the conundrum that practitioners in helping professions struggle with–how much of the practice of teaching is art, how much is science? When I am uncertain about what I should do in a specific situation, what do I lean on most? What I have experienced? What the data say? Some combination of both?
For physicians who have a far richer data base of research studies and clinical practice guidelines to diagnose and treat a variety of illnesses with just a few clicks of keyboard keys than teachers have at their disposal, similar questions of privileging practical experience over research studies when faced with a specific patient arise. An earlier post (“Do Doctors Resist Reform?” May 6, 2011) discussed how data-driven evidence-based medical practice, while rhetorically the gold standard is far from pervasive among physicians after nearly a quarter-century, and this in a profession that prides itself on learning lessons from the science of double-blind, clinical tests (see post of January 27, 2011: “Medical and Educational Research: What To Believe?”) Again, the mix of art and science in teaching and medicine, of learning from daily practice and data-rich sources remains undetermined, tension-filled, and, yes, mysterious even in 2011.