A group of leading teacher preparation program providers – Deans for Impact – released a manifesto and set of policy recommendations this week calling for “outcomes-based accountability and data-informed improvement.” Their recommendations are not inconsistent with the five key things we know about teacher preparation:
- Teacher preparation programs are disconnected from local labor market needs. There is no incentive for teacher education programs to slow the overproduction of elementary school teachers and speed the underproduction of secondary school math teachers even though it’s the latter where teacher preparation programs have shown their greatest value-add in terms of classroom teacher impact on student achievement. In Illinois, for example, there are roughly nine new elementary school teacher certificates issued for every one elementary school, first-time teacher hired.
- There is no specific teacher preparation program input measure – not SAT/ACT score, postsecondary education subject matter training (beyond secondary level math), program length, master’s degree attainment, institution selectivity, or even certification – that accounts for more than a small percentage of teacher effectiveness as measured by associated elementary and secondary school student achievement scores.
- The results of state teacher preparation program accountability systems are highly dependent on the statistical model used to evaluate value-added effectiveness. Ohio’s model and one used in Missouri have led to a finding of no statistically significant difference among preparation programs. On the other hand, models used in New York City, Louisiana, Washington State, North Carolina, and Tennessee all repeatedly have found significant effects. The variability of outcomes among models (and associated disagreements among academics) cautions that high stakes accountability consequences should only attach to very good and very poor-performing programs on multiple measures.
- The best hope for teacher preparation program improvement lay with those able to replicate most closely the first-year teaching experience with quality mentors and limited class size responsibility (i.e. a program rooted in quality clinical experience). Why? Because the single greatest predictor of future teacher effectiveness by a factor of seven is initial year effectiveness, as reflected by value-added measurement.
- Even though low-performing teachers improve markedly after year one in the classroom, the average low-performing teacher, measured by year one results, never catches up in effectiveness – in years two, three, or four – to the level of effectiveness of a median first-year teacher (see chart below).
Teacher preparation programs that consistently produce low-performing teachers year one in the classroom do a disservice initially and over time to elementary and seondary school students. These programs should face two options: (i) improve; or (ii) lose their eligibility to contribute toward the granting of state teacher certification (accredited programs would still be able to confer a postsecondary degree).
The good news is more and more education groups are coalescing behind the idea that data on teacher candidate placement and performance in K-12 classrooms can spur improvement in training programs.
In Deans for Impact’s policy recommendations, for example, they say:
Policy needs to provide actionable data, as well as support and tools for program improvement, to help those at the front line of our education system succeed.
To that end, we should:
- Improve data access through policies that provide educator-preparation programs with information on the performance of their graduates.
- Develop a new outcomes-focused state-level certification process that recognizes programs that voluntarily agree to prepare educators who are demonstrably effective.
We at Education Reform Now couldn’t agree more because great teachers are made, not just born.