Nelson Cowan identies 9 models of WM. What's the problem? If we talk about things in general terms, then we have to expect different interpretation and understanding. We should be very clear when using technical language and position it in the specific model we are using and the limitations.

I'm my recent #CogSciSci talk, I went a step further and suggested that you start with recognition tasks and gradually change to questions with less cues.

This is a good question. You are essentially referring to something called transfer appropriate processing. So evidence from retrieval practice would suggest that task matching increases performance. However, we cannot predict exam questions so I would favour mixing formats.

Oooo, I like that. Spot on in some cases. It's a good question to ask yourself when planning activities.

Thanks so much. This is an important point - just because you do something and call it retrieval practice doesn't mean you are gaining the benefit. If we overly focus on the activity, then we may lose sight of the purpose.

Thanks so much. Hope you found it useful.

Indeed, especially when the science isn't that clear cut. I really enjoyed making the Star Wars slide, my best work I feel 🤣

So where does this leave us? Well, correlation isn't causation but it does demonstrate that just doing strategies is not enough. Doing them well and having positive attitudes and beliefs about revision is arguably more important.

Interestingly, where students had been taught about effective revision strategies, they were more likely to use things like distributed practice. However, their average grades were lower.

In a recent paper 👇 although evidence based strategies were associated with achievement, the correlation was low.
cognitiveresearchjournal.springeropen.com/articles/10....

This is important as we want students to perform well on exams. However, much of the revision advice is based on strategies that support long term learning, such as distributed practice and practice testing. A question though is are these even effective for performance?

Soderstrom and Bjork make a clear distinction between learning and performance, where they demonstrate that increases in performance doesn't necessarily lead to learning and lasting learning doesn't necessarily lead to immediate performance.

This paper is regular referenced as the revival of this area of research.

journals.sagepub.com/doi/10.1111/...

If you are talking about the testing effect and spacing effect, then yes, this is the boundaries of the research.

I never said it wasn't but if we are using these terms then they need to be accurate. I'm just explaining the evidence base.