IN PRAISE OF KEEPING FINAL EXAMS
`Demonstrations of learning' can't replace testing, says Paul W. Bennett.
Suspending Ontario final exams once again in 2020-21, has, so far, elicited barely a whimper, even in the universities. Or as University of Ottawa professor Andy Hargreaves quipped, “Nobody's missed standardized tests or exams.”
Caught up in the exhausting battle against COVID-19, few seemed to notice that the virus was claiming another victim in Canada's reputed “learning province”: sound, reliable, standards-based and replicable summative student assessment.
The door was opened in October when Ontario's “back-to-basics” government gave school boards the right to replace exam days with in-class instructional time. Shortly thereafter, larger metropolitan school districts, including the Ottawa public and Catholic boards, jumped in quickly to suspend exams in favour of what were loosely termed “culminating tasks” or “demonstrations of learning.”
Traditional examinations, the long-established benchmark for assessing student achievement, simply disappeared, for the second assessment cycle in a row going back to the onset of the pandemic. Suspending exams has hidden and longer-term consequences, but what's more concerning is that such decisions are rarely evidence-informed or predicated on the existence of viable, proven and sustainable alternatives.
Critics of exams contend that exams cause debilitating stress and adversely affect student well-being. Such claims are disputed by Canadian teen mental health experts, including Stan Kutcher and Yifeng Wei, as well as cognitive scientists such as Erin Maloney who cite evidence-based research demonstrating that tests and exams are examples of the “normal stress” deemed essential to healthy human development.
Sound student evaluation is based on a mix of assessment strategies, ranging from daily interaction and feedback to standardized tests and examinations. Testing remains a critical piece in balanced student assessment program, countering more subjective forms of assessment. United Kingdom student assessment expert Daisy Christodoulou puts it this way: “Tests are inhuman — and that is what is good about them.”
Teacher-made and -evaluated assessments appear, on the surface, to be more gentle and fairer than exams, but such assumptions can be misleading, given the weight of research supporting “level playing field” evaluations.
The reality is that teacher assessments tend to be more impressionistic, not always reliable, and can produce outcomes less fair to students.
Eliminating provincial tests and examinations puts too much emphasis on teacher assessment, a form of student evaluation with identified biases. A rather extensive 2015 student assessment literature review, conducted by Prof. Rob Coe at Durham University, identifies the typical biases.
Compared to standardized tests, teacher assessment tends to exhibit biases against exceptional students, specifically those with special needs, challenging behaviour, language difficulties, or personality types different from their teacher. Teacher-marked evaluations also tend to reinforce stereotypes, such as boys are better at math or racialized students underperform in school.
Constructing summative evaluation models to replace final exams is not easy and it has defeated waves of American assessment reformers. Initial successes in New England Essential Schools Coalition pilot schools with highly trained teachers proved next-to-impossible to scale-up in public systems.
Championed in the early 1990s by Outcome-Based Education (OBE) guru Dr. William Spady, system-wide student competencies assessment culminating in “demonstrations of learning” essentially crashed and burned. Grappling with the Byzantine OBE system and its time-consuming measurement of hundreds of competencies finished it off with classroom teachers. Two examples of projects that failed were the Kentucky Commonwealth Accountability and Testing System, 2007-2008, and its predecessor, KRIS, 1992-98.
Student assessment reform, recently piloted in New Hampshire as Performance Assessments for Competency Education (PACE), ran into significant problems in trying to integrate Classroom-Based Evidence with state academic standards under the Every Student Succeeds Act and in establishing consistency and comparability across schools and districts.
Professionally constructed competency-based final assessments, requiring moderation of grades and the re-scoring of classroom-based work, significantly add to teacher marking loads. That was a major contributing factor in the federal decision in December 2019 to cut funding for competency-based assessment experiments, effectively ending support for that whole initiative.
Substituting culminating student activities thrown together in place of exams is, most likely, a recipe for inconsistency, irregular marking loads, and dumbed-down standards. Teachers are doing their best under trying pandemic conditions, but let's not pretend that the crisis measures are better than traditional and more rigorous systems that have stood the test of time.