Ideally, teacher candidates' clinical placements should represent an environment similar to what they'll experience in their first teaching job. The more similar, the less the learning curve. What's not as clear is what matters most about the environment: teaching the same subject? The same grade? Teaching in a similar school? In the same district?

More helpful research on this once poorly studied topic from a CALDER working paper helps to provide some insight. Its findings suggest positive benefits when prep programs place their candidates who are likely to go on to teach in a high needs classroom in just those kinds of classrooms for practice.

Researchers John Krieg, Dan Goldhaber, and Roddy Theobald pair Washington state student teaching placement data with the state's administrative data for over 5,500 teachers. They find that first-year teachers who experienced a significant demographic mismatch between their student teaching environment and their first teaching assignment were less likely to produce strong learning gains in their students than new teachers who had a better match.

Nowhere was this discrepancy more apparent than among candidates who student-taught in affluent classrooms but then took a teaching job in a high poverty school. The larger the difference between the percentage of students qualifying for free and reduced lunch from one classroom to the next, the lower the math and ELA scores produced by a new teacher's instruction.