48,478 students’ data from 222 studies and 573 effects – Testing Effect (IV)
48,478 students’ data from 222 studies and 573 effects – Testing Effect (IV)

48,478 students’ data from 222 studies and 573 effects – Testing Effect (IV)

If you have made it to Part IV you probably do not need an intro. Question 12 is a belter! A fundamental education modelling question all of it’s own and easily a whole blog post and research studied in of itself. Critical to any teaching and learning lead, a question that underpins timetabling and teaching practice.

Q12. Should tests be administered in or out of the classroom?

This is perhaps the most under explored question and possible one of the more important to teachers. Yang et al (2021) find that testing administered in the classroom (g = 0.514) tend to be more beneficial than those administered out of the classroom (g = 0.401) is the broad perspective.

Reflection: Take a step back. Yes testing can be applied and equally effective to 18 curriculum areas however the experience and distribution/contact for both educators and learners across the curriculum are vastly different. Core subjects, such as a English – almost every day is not comparable to Art, once a week. What Success Relearning offers these subject is very different – in and out of the classroom. It is just a hunch, but I get the gut feeling that Successive Relearning may become even more important where:

  • the number of lesson per week / cycle reduces
  • meaning that out of class testing (the weaker of the two modes) becomes more important to avoid the “knowledge” or learning fading and dissolving
  • meaning learner agency and self regulation becomes more important

Indeed, Yang et al (2021) were surprised that this aspect of testing and teaching or question, has never been explored before. Why would it? If Successive Relearning was not a part of your teaching and learning outlook?

Q13. Should tests be administered pre or postclass?

Both pre- (g = 0.186) and postclass (g = 0.536) quizzes significantly enhance learning, but post class quizzes are more effective. Backward over forward testing. That said, forward testing or “priming” or “pretreival” is fast becoming one of my favourite strategies and take very little effort.

The Eureka moments that arise when learners experiences connecting a forward / priming / pretreival question in class – joyful. “Is that linked…”

Q14. Does administration mode matter?

Paper-and-pen, clicker response systems, smartphones, online websites, and personal computers. Little
agreement has been reached about whether administration mode. Nothing to report here. Yet I get the sense that “how” teachers employ retrieval practice and Successive Relearning is important. Particularly the cost-benefit analysis of overt and covert retrieval – ie the time to physically recorded and answer, and the challenges of assessing accuracy, vs cognitively thinking through responses and time gained.

Q15. How does the effectiveness of test enhanced learning vary with treatment duration?

Testing benefits systematically increase from single class treatment (g = 0.385), treatment lasting less than a semester (g = 0.521), treatment lasting a whole semester (g = 0.547), to that lasting longer than a semester (g = 0.624). There is a significantly positive relationship between the classroom testing effect and the total number of class meetings. Hence, the relationship between test-enhanced learning and treatment duration is approximately linearly increasing.

3 more questions to go…

Opening postQuestions 2-6Question 7-11Questions 12-15Questions 16-19

Leave a Reply