Principals' Test Not Predictive of Success on the Job
Exam results show racial disparities
New research has found essentially no positive correlation between how would-be principals perform on a widely used licensure exam and their success as school leaders.
The study, which looked at principals’ performance on the School Leaders Licensure Assessment (SLLA) and on-the-job evaluations, student achievement, and teacher surveys, over a 10-year period in Tennessee, also found that non-white candidates were about three times less likely than white candidates to pass the exam.
The researchers found that candidates with higher passing scores were more likely to be hired as principals. And because Tennessee has the lowest cut score among the states that use the SLLA, disparities in passing rates for white and non-white candidates could be greater in states that set higher cut scores, said Jason A. Grissom, the lead researcher and an associate professor of public policy and education at Vanderbilt University in Nashville, Tenn.
The study’s findings could have major implications for districts and states beyond Tennessee, where K-12 officials are trying to increase the numbers of racially and ethnically diverse educators. Students of color now make up a slight majority of public school enrollment, but the education workforce remains largely white.
“There are some suggestions here that there may be some value in the exam, but it’s not showing up in the kinds of job-performance measures that we really care about,” Grissom said. “If we are seeing this barrier to diversity but the payoff on the other side is not very clear, then I suggest that we really need to look at this very closely, about what is the continued justification for using this exam.”
Widely Used Test
The SLLA is used in 19 states and the territories of Guam and the U. S. Virgin Islands. The four-hour computer-based exam is one of the multiple measures that states use in awarding principal licenses and that districts use in making hiring decisions, according to Grissom’s research. In Tennessee, for example, principals must also have at least three years of experience as an educator, complete a state-approved preparation program, and receive a recommendation from such a program.
Officials at the Tennessee education department and the Princeton, N.J.-based Educational Testing Service, which designed the SLLA, stress that the test is not meant to be a predictor of on-the-job success.
“Licensing exams like the SLLA are designed to protect the public from practitioners who lack basic competence, not to predict job success,” Thomas Ewing, an ETS spokesman, said in a statement.
The exam was “developed through a rigorous process to assess the skills, knowledge, and abilities that licensing authorities have determined school leaders need when they enter the profession,” he said.
A committee from diverse backgrounds developed the exam, and the test questions were reviewed “for possible unfairness to any group,” Ewing said. The exam also goes through statistical analyses to ensure it is fair to test-takers, regardless of race, ethnicity, or gender, he said.
Paul Fleming, Tennessee’s assistant education commissioner of teachers and leaders, said the report’s findings were similar to other research on the use of the SLLA.
The department has taken several steps over the years to reduce the disparities in passing rates for whites and non-whites and to ensure that the licensing process does not intentionally exclude non-whites and minorities, he said. One of those measures, he said, was setting a lower cut score.
The difference in passing rates between white and non-white test-takers—12 percentage points—was surprising, Grissom said. The researchers could not explain why based on the data they had, and they attempted to find some answers.
They theorized that principal-candidates may sort into programs of varying quality, and that based on the test results, it was possible non-white candidates graduated from lower-quality programs. But their analysis showed that racial and ethnic disparities in passing rates existed even among candidates who graduated from the same program.
The researchers did find positive relationships between assistant principals’ performance on the SLLA and evaluations from their supervisors once they were on the job.
But even in those cases, the positive relationships were present in some, but not all, the years, and the reasons why they were more evident for assistant principals but not principals were unclear, Grissom said.
The SLLA is based on the Interstate School Leaders Licensure Consortium Standards, which set benchmarks for what principals should know and be able to do to effectively lead schools. They were redesigned in 2015 as the Professional Standards for Educational Leaders, with deeper emphasis on instructional leadership and care for students.
Ewing said ETS is reviewing the SLLA to ensure it aligns to the new standards.
Grissom said there were a few possibilities why researchers did not find a positive correlation between the exam scores and principals’ job performance. One could be that the skills and knowledge that the test tries to assess don’t capture the competencies that school leaders need to be effective.
“School leadership is not a set of knowledge and skills,” he said. “School leadership is a set of behaviors, and implementation, and relationships ... and so forth, and that’s really complex.”
Not all states use an assessment, and some, such as Massachusetts, have developed their own. One option would be to remove the licensure assessment, Grissom said.
“An alternate state strategy would be to invest very heavily in administrator preparation, in ensuring high-quality preparation across all approved programs in the state,” he said.
Vol. 36, Issue 27, Page 8Published in Print: April 5, 2017, as Principals' Test Not Predictive of Success on Job, Study Finds