Making a Poor Case for Teacher Pay Reform

In a stunning display Professor Jacob Vigdor, a Duke University economist, contradicts his own “teacher pay” findings and recommendations in a recent study — and the News and Observer does a poor job in reporting on it. In drawing on a large teacher and student data base in North Carolina, Mr. Vigor calls for “scrap(ping) the sacrosanct salary schedule” because experienced teachers do not raise standardized test scores at a commensurate rate relative to their lock-step pay raises and that teaching experience and teacher preparation do not matter for improving academic achievement.

However, in a previous study, which he co-authored, Vigdor and colleagues concluded that “students ‘exposed’ to a teacher ‘with very weak credentials….would be expected to achieve close to .30 standard deviations lower that if they had a teacher with the strong set of credentials (p. 29).” The largest negative effects were found to be if teachers were inexperienced and if they entered teaching with an alternative certificate. The N&O, one of North Carolina’s leading daily newspapers, misses the boat by not carefully examining the data and not challenging assumptions that a ivory-tower economist makes about teachers and teaching experience and what standardized tests can and cannot measure.

Perhaps,  T. Keung Hui, the N&O staff writer, should turn to National Board Certified Teachers –who have proven to be more effective practitioners— to learn more about what matters most for student learning. A group of NBCTs -- working with the Center for Teaching Quality -- recently released a major report on how our nation’s current standardized tests do NOT come close to measuring 21st century skills our public school students must develop. They clearly show how well-qualified, experienced teachers — who do not have the same job as other workers who sell widgets for a commission (as Vigdor suggests) — make a meaningful difference for the students and communities they serve. And they point out how our schools need to more carefully develop metrics for teaching effectiveness.

Teachers, especially our most accomplished ones, have been calling for new ways to pay teachers for advancing student learning, developing new and relevant skills, teaching in high needs schools and assignments, and leading reforms. However, just paying teachers more for test score gains won’t work not because of union resistance (as Mr. Vigdor says), but because the scores are too unstable to judge teachers solely on the basis of them. Mr. Vigdor and the N&O should ask the researchers who know the psychometrics* and the teachers who know the students. Then they could make a better case for much-needed teacher pay reforms.

*Braun, H. 2005. “Using Student Progress to Evaluate Teachers: A Primer on Value-Added Models.” Princeton: Educational Testing Service.

Comments

[William M. Fox]

Placing almost exclusive emphasis upon test-score improvement as a basis for rewarding teachers is patently unfair and, when coupled with inadequate performance-appraisal systems, drives teachers toward unethical behavior or departure to other pursuits.

A primary reason the public has not been more supportive of higher funding for education has been the poor relationship between better funding and higher educational quality as revealed by a number of studies.

Use of an appraisal system based upon the following guidelines should go a long way toward turning things around.

Those associated with schools, need to fairly identify true "stars" and "inadequate performers" as one of the bases for:

justifying good pay for outstanding teachers,

providing for self-guidance on the part of newcomers and present staff,

and providing an important basis for terminating those who cannot, or will not, measure up.

Research findings show that evaluators achieve much better agreement about who are Stars and Inadequate Performers than they do about who are Average, Above-Average, and Below-Average performers. Yet, placing individuals in the middle-three categories is a time-consuming, often arbitrary, and resentment-causing activity that most evaluators dislike having to do. Also, clearly, an average performer in a superior organization deserves much more recognition than an average performer in an inferior one. No wonder that many teachers and their unions oppose conventional merit-rating systems!

To avoid a popularity contest, assure greater fairness, and provide for constructive self-guidance, there should be behavioral documentation for both Star and Inadequate Performer nominations via the Critical Incident Technique.
To lay the groundwork for this, students, parents, veteran administrators, and experienced teachers should be polled at to what specific, observable behaviors they associate with outstanding and inadequate performance for each important aspect of a teacher's job.

Then, required behavioral documentation for Star and Inadequate-Performer nominations from fellow teachers, adminstrators, students, and parents should be based upon the most agreed-upon behaviors, and the agreed-to relative weights that should be assigned to these.

The results of this analysis can also constructively guide the initial training and subsequent selection of teachers, as well as, provide a much-needed, qualifying context for the currently over-stressed evaluation factor of test-score-improvement.

This approach also sets the stage for more productive review sessions between the rater and ratee. Since the ratee has a sound basis for self-rating, the session should start with the rater asking "How do you rate yourself for this past period through the presentation of relevant, supporting behaviors?" No rater can be all-knowing, so if behaviors are mentioned that she or he is not aware of, the rater can postpone giving his or her evaluation to provide time to check out the validity of the assertions, if this seems necessary.

A sound behavioral basis for rating also facilitates the use of motivational goal setting during the review session. For example, if the ratee wants to be a Star, what specific behavioral goals does she or he plan to adopt by such and such a time? If stardom is not the goal, which specific, Inadequate Performer behaviors will he or she need to avoid?
This approach permits a rater to be more of a counselor and coach, than one who appears to sit in arbitrary judgment.

For discussion of relevant research and related citations, see: "Improving Performance Appraisal Systems" by William M. Fox, NATIONAL PRODUCTIVITY REVIEW, Winter 1987-88, pages 20-27.

William Fox
gryfox@bellsouth.net
Professor Emeritus
Department of Management
University of Florida
(352) 376-9786