Special Coverage on the 2020 Invited Talk

Individual Differences in Learning and Forgetting (114)

Kathleen McDermott (Washington University in St. Louis)

Summary by Laura Mickes, Digital Content Editor

This recap is part of a special series of session summaries from the Psychonomic Society's 61st Annual Meeting. To read the rest of the series, click here.

Kathleen McDermott is one of the Psychonomic Society's 2019 Mid-Career Award recipients. The topic of her presentation is if an individual's rate of learning predicts their rate of forgetting. McDermott summarized research that shows the rate of learning is related to retention. She also reasoned that because there are retention differences doesn't mean there will be differences in forgetting. Her main question for the research described in her talk is if someone learns quickly, will that person forget quickly? 

To answer this question, McDermott and Chris Zerr had participants memorize word pairs until they had the entire set of them learned. The learning curves, plotted in the figure below, show a considerable variation in learning ranging from learning 42 out of 45 pairs on the first trial (dang! well done!) to learning 0 pairs on the first trial. 

As shown in the figure below, faster learners remember (dark red line) more than slower learners (dark blue line), 

These results are in line with other results whereby the rate of learning predicts retention. Does the rate of learning predict forgetting?

Independent variables, like the age of the individual or with items studied once vs. thrice, don't typically impact the rate of forgetting. Therefore, one wouldn't necessarily predict differences in forgetting curves. However, McDermott noted a few reasons to do so.  

But they do! As you can see in the figure below, faster learners forget slower than slower learners. 

They also fit each participant's data with different models, and the exponential-power function fit best. Still, the exponential and power functions fit well, and even the logarithmic and linear function fits weren't too shabby. 

The findings challenge past conclusions and open an entirely new set of inquiries.