The proposed recertification of teachers in Alberta announced earlier this month is controversial. Would data from assessments of student learning be useful in measuring teacher effectiveness?
In Alberta there are province-wide assessments of individual student learning in Grade 3, 6 and 9. There are two obvious problems in using these assessments to measure teacher quality.
First, student performance when assessed in Grade 3 and Grade 6 is clearly a product of exposure to many teachers and other staff. Second, the assessment outcome is clearly a product of both the school and the home environment of students. Students who arrive from more advantaged homes will do better. A teacher once jokingly said to me as he transferred from a school with many underprivileged children to a school with many privileged children that he had instantly become one of the best teachers in the city if teacher quality was measured by assessment results.
We know from work in Canada and many other countries that groups of educators vary in quality. My studies for the C.D. Howe Institute for Alberta, Ontario and British Columbia provide useful measurements of the quality of the group of educators at a school where student background is taken into account. Although a student's socioeconomic background matters, good teaching at a school matters just as much or more.
Can the same techniques measure individual teacher effectiveness? We rarely have data on the background of individual students. Education ministries often do not record which students are taught by which teachers. These facts make it impossible to use Canadian assessment results to measure the quality of individual teachers. This is not the case in other countries.
In many American jurisdictions there is an annual, although controversial, assessment of student skills. If a student has one main teacher, then to assess how good a teacher is, you can calculate an average learning gain over the year for that teacher's students. These measures show there is substantial variation in what is called teacher value-added.
The authors of a significant 2011 study found that students from New York City exposed to high value added teachers in early grades had better life outcomes, postsecondary attendance and higher earnings.
Good teaching can be measured using value-added from annual assessments and the resulting measurement is a useful predictor of life success. Annual assessments linked to individual teachers are needed to carry out this task.
In Canada, there would be huge political costs to implement annual standardized assessments and to link student data to individual teachers. Deciding to spend resources on annual assessments would be, to say the least, controversial. Even the assessments currently in place are constantly threatened in spite of their clear value in assessing the quality of education at a school level and ensuring a common curriculum is taught across a province. Linking students to teachers appears to be a non-starter in labour negotiations with teacher unions.
If the results of an annual assessment were not used to evaluate teachers, it might be hard to argue that the additional information from a standardized assessment every year would be worth the additional cost. In the absence of annual student assessment data, would teachers, parents and school administrators have confidence in a process with no objective measures of teacher competence?
The current process for teacher evaluation does not work. The Alberta task force notes that with 40,000 teachers in Alberta, not one teacher ever lost a teaching certificate due to incompetence in 10 years. Yet the Alberta School Act (Section 18) states that a teacher must "provide instruction competently to students." The Act (Section 94) also states that the Minister can make regulations "governing what constitutes unprofessional conduct or unskilled or incompetent teaching." In law, an evaluation process already exists. The government and the task force will have to clarify the evaluation process they have in mind. A minor variant of the current process is unlikely to be an improvement. The creation and use of annual student assessment data to evaluate teachers, although it would clearly contribute to a meaningful evaluation process, seems unlikely.
David Johnson is Professor of Economics at Wilfrid Laurier University and Education Policy Scholar at the C.D. Howe Institute.