Which steak would you prefer: one that is 25% fat, or one that is 75% lean? If you’re like the participants in a classic study by Irwin Levin, you’d pick the latter.

This finding represents a framing effect, where equivalent information presented in different ways influences behavior. In this example, the steak labeled with a positive frame (75% lean) is viewed more favorably than the one with a negative frame (25% fat).

Framing effects have been found on people’s attitudes, goals, decision-making, and even metacognitive monitoring (metacognition is awareness of one’s thinking about, well, thinking). For example, people are more optimistic about their future memory when asked to make judgments framed in terms of what they will remember as compared to when they are framed in terms of what they will forget.

In a set of experiments published in the Psychonomic Society journal Memory & Cognition, authors Dillon Murphy (Twitter) and Barbara Knowlton (pictured above) investigated whether framing effects occur for how people prioritize remembering valuable information. To do so, they manipulated whether a memory task was framed in terms of losses or gains to see how this would influence memory for information of varying importance.

As shown in the figure below on the general procedure, participants aka “learners,” studied 6 lists containing 20 words each. During the study phase, each word was presented for 3 seconds. After each word was presented, learners estimated the likelihood of correctly recalling it on a later test on a scale from 0 (definitely will not remember) to 100 (definitely will remember). Following the study of each list, learners had 1 minute to recall as many of the words from that list as they could remember, after which they received feedback on their performance for that list.

Learners were told that each word would be paired with a random number between 1-20 that represented how many points the word was worth. For a random half of the learners, task instructions were framed in terms of gains. Specifically, they were told, “… if the word ‘apple’ appears with a 5 next to it and you remember ‘apple’ during the test, then you will receive 5 points.” Similarly, their feedback was framed in terms of how many points they scored (e.g., “You got 100 out of 210 points.”).

For the other half of the learners, the instructions were framed in terms of losses. They were told, “… if the word ‘apple’ appears with a 5 next to it and you forget ‘apple’ during the test, then you will lose 5 points.” Their feedback was framed in terms of how many points they lost (e.g., “You lost 110 out of 210 points.”).

The authors then analyzed whether recall and metacognitive judgments varied as a function of point values and framing. The authors hypothesized that when evading losses, learners may aim to avoid the cost of forgetting by trying to remember more information overall – including the words worth fewer points. However, when seeking gains, learners may aim to prioritize remembering more of the highly valuable information (i.e., the words worth more points).

The results aligned with the authors’ predictions. Consistent with prior research on selective remembering (i.e., value-directed remembering), learners recalled more high-value than low-value words. And, this effect was larger for learners whose task instructions were framed in terms of gains than when instructions were framed in terms of losses. As shown in the figure above, learners were more selective of high-value information when instructions emphasized maximizing their score versus minimizing their losses. Interestingly, learners’ metacognitive judgments tracked recall, suggesting they were generally aware of this effect.

In a second experiment, the authors investigated how learners allocated their study time as a function of task framing. Learners again studied lists of words paired with point values and received instructions and feedback in terms of either gains or losses. However, in contrast to the first experiment, learners were told they could study the word for as long as they liked (up to 10 seconds per word).

Learners spent more time studying high-value words compared to low-value words, but there was no influence of framing on study time. And, in contrast to the first experiment, no effects of framing were found on recall, leading the authors to speculate that framing effects on selectivity may be mitigated when learners have control over study time.

In a third experiment, the authors attempted to directly compare the effects of framing on memory selectively under fixed and self-paced learning conditions. No metacognitive judgments were solicited in this final experiment. The results of this experiment largely replicated the findings of the first two experiments: Regardless of the study schedule, when instructions and feedback were framed in terms of gains (versus losses), learners were more selective about what they chose to remember. These results show that framing effects generalize to memory and metacognitive processes.

In everyday life, we are exposed to far more information than we can possibly remember, and we often have to be strategic about what’s worth remembering. For the first time, Murphy and Knowlton show that “How you frame a learner’s goals, as well as the framing of feedback, can influence how they remember.”

Featured Psychonomic Society article:
Murphy, D. H., & Knowlton, B. J. (2022). Framing effects in value-directed remembering. Memory & Cognition, 1-12. https://doi.org/10.3758/s13421-022-01317-y