Strong Schools Can Withstand Teacher Turnover?

Posted on August 28, 2013


A recent NY Times article about young, inexperienced teachers and teacher turnover focused on the YES Prep charter schools. In that article, Wendy Kopp, the founder of Teach for America, stated, “Strong schools can withstand the turnover of their teachers. The strongest schools develop their teachers tremendously so they become great in the classroom even in their first and second years.”

So, one has to ask how Ms. Kopp defines strong schools. YES Prep did win the first Broad Prize for charter schools. But how strong are YES Prep schools? Do they have strong student achievement growth? Do they graduate a high percentage of students enrolled? How well do students perform after high school?

Let’s look at some data from Texas that I have readily available.

Do YES Prep Schools have high achievement and student growth? I think anyone examining both the percentage of students proficient and the value-added growth scores indicate a pretty strong school. I won’t present the details, but the percentage of students passing the state tests is certainly above average and the valued-added measures provided by the state (which are problematic in some respects, particularly because they are correlated with the achievement of students in the year before entering a school)  are also above average. I guess they get “two claps and a sizzle” for these accomplishments.

But what underlies those achievements?

For one, as compared to schools in the same zip code , YES Prep enrolls students with greater levels of achievement (as measured by z scores of scale scores across five cohorts of incoming students as shown below. Specifically, students entering YES Prep in the 6th grade had math and reading scores at least 0.225 standard deviations greater than the scores for students entering other schools in the same zip code as YES prep schools.

Students entering YES Prep are also less likely to be designated as English Language Learners or have special needs (as measured by taking an alternate or modified version of the state tests in the year prior to entering the school).

Yes Prep Incoming students

Given that peer effects can have substantial impacts on student achievement, the above data suggests that one ingredient of the YES Prep “secret sauce” is simply enrolling more advantaged students than schools serving the same communities.

Not only do YES Prep schools enroll more advantaged students than schools in the same community, but they also lose a greater proportion of lower-performing students and retain a greater percentage of higher performing students than other schools in the same zip code. This is shown below and is based on five cohorts of students from the 6th grade to the 8th grade. The percentages for the 6th grade row indicate the distribution of 6th grade students by their 6th grade test score. The 8th grade row indicates the distribution of the students remaining in the school based on their 6th grade score. The “DIFF” row documents the percentage point change from 6th to 8th grade. So, for example, we see that 7.2% of the students in 6th grade had a TAKS math z-score less than or equal to -1.0. Two years later, only 2.2% of the students remaining at YES Prep had a 6th grade TAKS math z-score less than or equal to -1.0. At the other end of the continuum, 32.9% of the 6th grade students had a TAKS math z-score greater than or equal to +1.0 and the percentage increased to 40.3% after the effects of attrition over the two years.

Relative to comparison schools (schools in the same geographic location (schools in the same zip code plus schools in all contiguous zip codes), YES Prep schools lost a greater percentage of lower performing students and retained a greater percentage of higher performing students.

YES prep MS attrition by score

More recently, I examined the retention of students from the 8th grade to three years later (students would normally be in the 11th grade, but students were also considered to have stayed at the same school if they were enrolled in any other grade).  Not only did less than 60% of the students remain, a substantially lower percentage of the lower performing students remained at YES prep. Clearly, one reason for improvements was the disappearance of a substantial proportion of the lower-performing students. Perhaps this attrition is due, in part, to the high turnover of teachers.

YES Prep 8th grade to 11th grade attrition by math scorePerhaps YES Prep would retain a greater percentage of lower-performing students if they actually retained a greater percentage of teachers. Does Wendy Kopp believe a strong school is one that loses a substantial percentage of students and a greater percentage of lower-performing students than higher-performing students? Is this what we want all schools to look like? And what about those lower-performing kids? maybe those simply are not the “strivers” (as Mike Petrelli calls them) and reformers simply don;t care much about such students.

Finally, how do YES Prep students fare in college? Well, Texas tracks that data and makes it publicly available. The data are limited since it can only determine a student’s GPA if that student entered into a Texas public 2- or 4-year institution. However, the data provides some good insights into how students perform in their first year of post-secondary schooling.

As shown below, more than 40% of the students entering 4-year institutions earned less than a 2.0 and 60% earned less than a 2.5 GPA> Not particularly stellar for a college prep high school. Even for students entering 2-year institutions, almost one-third earned less than a 2.0.


YES Prep College GPA

In the end, I’d like to ask Wendy Kopp if these are the characteristics of a “strong” school? Do we really want schools that cream-skim (intentionally or not) schools from the local community, then retain a greater proportion of the higher performing students than lower performing students, and send kids to college with really high state test scores, but clearly not the knowledge, skills, and traits necessary to succeed in college? I guess if these are the schools we want, then having a large percentage of very inexperienced teachers with high rates of turnover is acceptable. It clearly is to Ms. Kopp.

Posted in: Uncategorized