Very few people like taking tests. But testing can be a vital part of effective learning and long-term retention of knowledge. The trick is to apply tests in the right way. that tests “ask students to look into their wells of knowledge, locate information, and express that knowledge on the page”.
As Lahey explains in an excellent piece in the Atlantic, testing comes in two basic types: summative and formative. Summative tests “measure students’ sum total knowledge or ability at a fixed point in time”. Formative tests are “designed to discover what students do and do not know in order to shape teaching during and after the test”.
In 2006 research, Roediger and Karpicke explain that the “retrieval processes engaged in during a test are responsible for enhancing learning”, and specifically “elaboration of encoding, more effortful or deeper encoding, and creation of different routes of access” explain the testing effect. Mere exposure to the learning material is not sufficient.
The authors examined many rigorous studies about testing and found strong evidence of a “testing effect” where frequent, early testing was the best way to trigger this enhanced encoding of learnt material. As one might expect, without testing the recall of learnt material drops off steadily over time. However, one study showed that early testing mostly eliminated the expected drop-off in knowledge recall nearly two months after the first test!
This is where testing goes wrong. Summative tests (which only test at the end of a learning period) are done too late to enhance recall. Their use is (at best) just a statistical measure of student outcomes. But for active, ongoing learning it is clear that testing early and often is the way to go. Lahey sums it up well:
Formative testing at its best is low-stakes and high-frequency. When students are used to the practice of being tested (or “quizzed,” if that term carries less baggage) it loses its emotional teeth and its utility as an educational tool begins to emerge.
Also published on Medium.