Tag Archives: winerip

Problems with a Test vs. Problems with Testing

In the NYT this weekend, Michael Winerip details the fiasco in Florida stemming from bad scores on the state standardized writing test. This past summer, the state toughened standards but did not change the required passing score, resulting in plummeting proficiency levels statewide: only 27% of fourth graders scored proficient, compared to 81% just one year ago. Concerned about reactions nationally and from parents, the state scrambled to improve their results, ultimately lowering the score required to pass from a 4 to a 3. Suddenly, proficiency rates were right in line with past performance. All was right again (not). Excuse me for a moment while I bash my head against my desk.

Winerip is right to mock and malign the incompetence here. A couple points:

1) Radically changing the rigor of a test designed to measure longitudinal growth renders the results useless. It is fine to modify a test, and I appreciate the intent to raise academic expectations for students; making these changes at the expense of useful data, on the other hand, is misguided.

2) Fudging the numbers after the fact ultimately makes the problem worse. It is deceptive and unprofessional; I see it no differently from the state filling in bubbles for all of their students. The goal of testing is to highlight deficiencies to help guide instruction while holding schools accountable. Florida was apparently more concerned about its reputation, cheating the students and taxpayers while damaging the legitimacy of all state testing. If the state doesn’t have even have faith in their tests, why should we?

Though this case demonstrates the dark side of testing, we need to be careful about our conclusions. It is clear that there is something wrong with this test and test administration; Florida clearly needs reform in those areas. But it doesn’t mean is that holding schools accountable through testing is a bad idea. As I’ve said before, we cannot let the perfect be the enemy of the good. Standardized testing as an institution is so young that we should not be surprised that the tests aren’t perfect; it will take time before they are where they need to be. Just as we give any other invention with great potential time to mature–computers, cars, planes–testing deserves the same patience. So while Florida fully deserves our ire and frustration, let’s remember what this incident is: a speed bump and valuable lesson on what not to do as we establish accountability as a tenet of our public education system.

Performance Assessments and Certification

Today’s NYT features a piece by Michael Winerip on Ed students at UMass protesting a pilot program for an outsourced performance assessment to be used for teacher certification. Developed by Stanford University and Pearson, the Teacher Performance Assessment (TPA) would require teachers to submit a portfolio of their teaching experience, including lesson plans, student work, personal reflections, and a video of their teaching. The UMass students are concerned about the corporatization and standardization of what is traditionally a personalized process of evaluation; as Winerip quotes Barbara Madeloni, “This is something complex and we don’t like seeing it taken out of human hands. We are putting a stick in the gears.”

The TPA, which New York, Illinois, Minnesota, Ohio, Tennessee and Washington have already committed to adopt, fits with a growing trend of standardizing education nationally. If we want common core standards nationally, isn’t it logical to have a national certification process as well? The TPA also seems to prioritize teacher practice and development over the mere theoretical and content knowledge required by tests like the PRAXIS, which remains a pillar of the certification process in most states. In that sense, performance-based certification assessments seem to be another example of progressive reform.

But the practical implications are more complicated. States that implement the TPA and similar systems for certification will actually create a hurdle for “fast-track” and alternative certification programs.

The TPA features the Teaching Event, focusing on student-teaching experiences over a 3–5 day learning segment with a class of students. Throughout the learning segment candidates organize and submit evidence of their teaching (e.g., video clips of instruction, lesson plans, student work samples, teacher assignments, daily reflections) to create their own personal Teaching Event portfolio.

All teachers, then, would need to be in a position where they can have a 3-5 day “learning segment” with the same group of students. That is not difficult if a teacher has a traditional student teaching placement. But it is not something that can be easily set up independently, without the help of a college or university program. Where does that leave the increasingly popular alternative/fast-track certification programs such as Teach for America? Though the TPA is presented as a progressive, forward-thinking reform, I fear that it will have the effect of further insulating the teaching profession and blocking smart, motivated people from entering the classroom.

Don’t get me wrong. I certainly advocate using multiple metrics for evaluating teachers, and I really like the idea of using multiple sources of data–particularly student work–as part of that evaluation. But any step that further closes off the teaching profession is the wrong one. I would rather see the TPA used as a way to evaluate current teachers instead of as a further hindrance to certification.