Learn About: 21st Century | Charter Schools | Homework
Home / Edifier

The EDifier

September 14, 2017

New research: High-stakes tests influence teacher assignment decisions, impacting long-term student achievement

A new study released last month raises potential concerns about the ways in which teacher assignment decisions may impact student achievement. The study, which included data from the Miami-Dade County Public School district between 2003-2012, examined whether less-effective teachers were assigned to untested grades, and how those assignments affect students’ long-term academic achievement.

Previous studies have found that principals do take into account students’ academic growth when making decisions about teacher grade level assignments. One major factor in this decision is student scores on high-stakes standardized tests. Additionally, there has been evidence that less-effective teachers are more likely to be re-assigned to a low-stakes, untested classroom for the following school year. To further clarify whether teachers are re-assigned based on test scores, researchers measured the effect that a teacher has on students’ test score growth year over year. (Low-stakes tests given across the M-DCPS district were used to measure academic growth at the K-2 level.) They then examined the relationship between student test score growth and teacher grade level assignment in the following school year.

Researchers found that highly effective teachers in grades K-2, grades in which students are not subject to state tests, were more likely to be reassigned to grades three to five —tested, high-stakes grades— in the following school year. In contrast, highly effective teachers in third through fifth grades were unlikely to be reassigned to an untested grade. However, their lower-performing peers those third, fourth and fifth-grade teachers whose students made the least progress were more likely to be assigned to an untested K-2 grade in the following year. Researchers believe that by reassigning less-effective teachers out of tested grades, principals hope to improve student test scores over the short term. But what are the long-term consequences of concentrating the least-effective teachers in the “low-stakes” grades?

Though high-stakes standardized testing at the elementary level is focused in grades three through five, foundational skills learned in grades K-2, such as basic math and early literacy, drive success at all levels. After finding that lower-performing teachers are more likely to be reassigned to an untested grade, the researchers examined the effect that the resulting concentration of less-effective K-2 teachers could have on a student’s long-term achievement. Second graders taught by a teacher who had recently been reassigned from a tested grade had significantly lower gains in both literacy and math than their peers taught by teachers who had not been reassigned. Crucially, these effects carried into the following school year: a student taught by a recently reassigned teacher in second grade would also have lower third grade scores than their peers, reflecting a gap equivalent to having been taught by a first-year teacher during the second grade.

Clustering the least-effective teachers in untested grades— particularly K-2, where foundational skills like reading are taught— may have long-term consequences for student learning. Researchers have found that despite these lower gains for students over the long term, principals tend to focus on short-term staffing needs, and concentrate the highest-performing teachers in high-stakes, tested grades. These findings should raise questions for any district: How are student test scores used in staffing decisions, and how do those decisions affect student learning long term?

January 17, 2017

How schools hire teachers

Much attention has been paid recently to teacher shortages.  However, less attention has been paid to how schools hire teachers, which has a direct impact on the shortage issue (which varies by subject and region).  Ineffective teachers are more likely to leave, as are those who don’t “fit” well with the culture or demographic of the school.  This, of course, makes sense – people don’t typically want to stay in jobs they’re horrible at, especially ones that are high-stress and low-pay.  There are lots of other reasons teachers leave, but principals don’t typically have a lot of influence over employees needing more time for family, a job closer to home, or higher salaries.  Teachers’ satisfaction with their work environment and school leadership is paramount to their retention, as well, but today we’ll just focus on who gets hired.  Previous research has shown that principals and other hiring managers tend to hire teachers based on their relational skills – motivation, honesty, enthusiasm, and caring – over their track record for student achievement.

As a personal anecdote, I was hired as a teacher by three different schools in two Texas districts. I never had to provide a sample lesson plan, teach a sample lesson, provide prior test scores, or provide prior evaluations. It’s possible that they checked my references, but I’m not sure. I was certified for the grades and subject areas I was hired to teach, which are often hard to staff, and was willing to teach in high-poverty schools, so maybe the lack of data-collection was due to teacher demand being higher than the supply of qualified teachers.

Image result for teacher hiring

A recent study shows that my experience is not unique.  The researchers performed interviews with principals and school administrators in six large public districts and two smaller charter districts to determine how data was used.  About 70 percent of the principals reported using teachers’ previous observation or test score data when determining whether or not to hire teachers who might transfer into their campus from elsewhere in the district.  This seems to point toward a trend in which principals increasingly use data; previous studies showed that only 40 percent of principals used student achievement data in hiring decisions.  As more states and districts develop teacher evaluation systems, we may see this trend continue upward.  Performance-based assessments are even less commonly used.  The Center for American Progress, a left-leaning think tank, found that only 13 percent of surveyed school districts required teachers to teach a sample lesson with students.

You may be asking why 100 percent of principals aren’t using data, and I’d have to answer that there are multiple reasons.  First, in most teacher evaluation systems in which a teacher is observed by a school administrator, over 90 percent of teachers are typically found to be effective or highly effective.  The lack of distinctions and the subjectivity of these evaluations render them virtually meaningless in the hiring process.  While principals may look negatively on someone falling below this threshold, they may also balance this data with personal recommendations or personal experience with the subjectivity of the evaluation system.  They may also have limited options for qualified candidates to hire.

Even value-added models, often used as components of teacher evaluations systems, are not fully reliable.  Teacher scores may change from one year to the next, or from one test to another.  If principals are aware of the shortcomings of the testing system, they may be hesitant to rely on it for hiring. Many subjects are not tested on an annual basis, making it difficult to use related data for hiring teachers who don’t teach math and language arts.

Finally, teacher observations are often biased against teachers in high-poverty and high-minority schools, as well as against men.  Given that high-poverty, high-minority schools are also the most difficult to staff, it makes sense that principals might take prior evaluation scores for teachers who have taught at other high-needs schools with a grain of salt.

While data on teacher effectiveness may be flawed, it doesn’t mean that we should throw it out entirely.  If a teacher consistently has low scores or has bounced between schools due to ineffective ratings, principals would be wise to heed this information.  We can use data to weed out some bad apples; we can’t necessarily depend on it to be 100 percent accurate or predict success.

RSS Feed