This is an excellent letter to the U.S. Department of Education, which patiently explains the harm caused by value-added modeling (VAM). It was submitted by a Néw York group called “Change the Stakes,” which opposes high-stakes testing. The letter was written by psychologist Dr. Rosalie Friend, a member of Change the Stakes. It is a good source for parents and educators who want to explain why testing is being overused and misused.
USDOE’s Proposed Regs for Teacher Education Programs
Change the Stakes submitted these comments in response to the U.S. Department of Education’s proposal to impose new accountability measures on teacher education programs, https://www.federalregister.gov/articles/2014/12/03/2014-28218/teacher-preparation-issues.
The U.S. Department of Education has proposed that teacher education programs be rated by the employment, placement, and performance of their graduates. Ratings of the performance of graduates would include the test scores of the students who are taught by graduates of those programs.
Change the Stakes (changethestakes.org), an organization of New York City parents and educators promoting alternatives to high-stakes testing, opposes this proposal.
Rating teacher education programs by what teachers do after they leave the programs is unrealistic. The decisions made by graduates and their employers are not determined by the teacher education programs. Teacher education programs are already assessed by professional accrediting boards that understand the nuances of teaching and learning.
The accountability procedures imposed on K-12 schools have diverted astounding amounts of money and time from teaching and learning. The accountability procedures have not led to any measurable improvement in student achievement. Extending these ill-conceived procedures to teacher education programs is counter-productive. Attaching high stakes to evaluation leads to the distortion of the processes that are being evaluated, as documented by Dr. Donald Campbell, the pre-eminent social scientist.
Teaching is a difficult profession. Industrial-type accountability procedures distract from the focus on teaching and learning. We want teachers to learn how to engage children in learning new ideas and using those ideas to reason and solve problems. At the same time, teachers must be able to assist children with developing socially and emotionally. This requires dealing with enormous differences among children’s backgrounds and personalities. Of course, teachers must also be expert in the skills and materials they teach. Teacher education programs must prepare teachers to think on their feet and respond to the ever changing conditions under which they labor, not to drill children for shallow, regimented tests.
Teachers’ working conditions are a major factor in their professional achievement. Social conditions, school culture, school leadership, class assignments, and relationships among colleagues are all important in determining both students’ and teachers’ success. Management expert, W. Edwards Deming, said, “It is the structure of the organization rather than the employees, alone, which holds the key to improving the quality of output.” All these factors are independent of teacher education programs.
Perhaps the most wrong-headed part of the proposal is the use of student test scores in assessing the teachers who graduated from the programs. Using student scores to evaluate teachers and then to use that “so-called” data to rate their teacher education programs is unsound and unacceptable for the following reasons.
Low Reliability of Standardized Test Results
Value-added modeling (VAM) cannot be accurately used for a small sample such as a single class. The aggregation of student test scores to derive a score for an individual teacher has been demonstrated to be wildly unstable, especially while assigning scores to a given teacher from year to year or even from class to class. The American Statistical Association has warned against the use of VAM for teacher evaluation. Using these unreliable figures to draw conclusions about the programs that educated teachers is folly.
Low Validity of Standardized Test Results
Tests cannot adequately account for every factor outside of a teacher’s instruction that impacts how students perform on a test because there are far too many other factors affecting students’ scores. Research shows that whatever teachers’ impact is, it accounts for only 1-14% of student variability in standardized test scores. If the teacher’s score is based on factors other than the teacher’s influence, it is not valid.
Studies since the 1966 Coleman report continue to show that nothing affects student achievement as much as the student’s home. Parents in poor families cannot provide their children with the same social and learning supports and enrichment that affluent and middle-class parents can provide. Furthermore, well-funded schools in prosperous communities consistently get higher test scores than cash starved schools in poverty-stricken neighborhoods.
A teacher’s effectiveness is directly affected by the composition of the class assigned to that teacher even within the same school. What kind of academic background do the children have? Are their goals aligned with the school’s goals? How cooperative are they? How well behaved or self-regulated are they?
The entire process of professional training of an educator is exceptionally complex. While a school of education affects the resulting quality of the professional educator, so much more goes into their success. Any evaluation of such an institution should be developed to be inclusive of all the contributing factors, not simply the ones for which quantitative data (however invalid and unreliable) are available.
Ignoring these additional factors and the research supporting them is an injustice not only to the programs the Education Department plans to rate but also to students, teachers, parents, and communities alike.
American Statistical Association. (2014). ASA Statement on Using Value-Added Models for Educational Assessment. http://www.amstat.org/policy/pdfs/ASA_VAM_Statement.pdf
Baker, E. L., Barton, P. E., Darling-Hammond, L. D., Haertel, E., Ladd, H. F., Linn, R. L., Ravitch, D., Rothstein, R., Shavelson, R. J., & Shepard, L.A. (2010). Problems with the Use of Student Test Scores to Evaluate Teachers: Briefing Paper 278. Washington, DC: Economic Policy Institute.
Campbell, D.T. (1976). Assessing the Impact of Planned Social Change. Dartmouth College, Occasional Paper Series, #8.
Greene, D. (2013). Doing the Right Thing: A Teacher Speaks. Victoria, Canada: Friesen Press.
Haertel, E.H. (2013) Reliability and validity of inferences about teachers based on student test scores. William H. Angoff Memorial Lecture Series. Princeton, NJ: Educational Testing Service.
Johnson, S.M., Kraft, M.A., & Papay, J.P. (2012). How Context Matters in High-Need Schools: The Effects of Teachers’ Working Conditions on Their Professional Satisfaction and Their Students’ Achievement, Teachers College Record, 114:1-39.
Viadero, D. (2006). Race Report’s Influence Felt 40 Years Later: Legacy of Coleman study was new view of equity. EdWeek [Online] Available http://www.edweek.org/ew/articles/2006/06/21/41coleman.h25.html
These comments were written by Dr. Rosalie Friend, Educational Psychologist and a member of Change the Stakes.