Ask a medical student about the first patient who crashed on their watch, and you may get the blood pressure, the room number and what the supervising physician said. Ask about the enzyme table they reviewed five times last month, and you may get a shrug. One experiment captures that gap in miniature. Adults watched a slide show that turned, partway through, into a story about a car accident. A week later, those who had taken a placebo recalled the accident section better than the neutral parts. Those who had taken propranolol, a beta-blocker that stops adrenaline-like hormones from acting on their receptors, showed no such advantage (Cahill et al., 1994, as cited in LaBar & Cabeza, 2006).
Emotion differs from the strategies earlier in this series, because you do not decide to do it while studying. A better word for it is a modulator: a condition that changes how learning proceeds. It can make a memory last without making it accurate.
Two Routes Into Memory
How could anyone tell whether emotion helps a memory form or helps it last? In the slide-show study, the drug was taken before the story, so it could have acted on attention during the story or on the memory afterwards. James McGaugh’s laboratory got around that problem by waiting until learning was over. Rats learned a task, and drugs or hormones were given only afterwards, so they could not act directly on the learning or the test. Many were infused into the amygdala, a small structure deep in the brain next to the hippocampus. They changed memory tested a day or more later, not within a few hours (McGaugh, 2004). That pattern points to consolidation, the slow process by which a fragile new memory becomes stable. On this account, stress hormones released during arousal act on the amygdala, and the amygdala adjusts consolidation in the brain’s memory regions (LaBar & Cabeza, 2006).
The human case is thinner. Studies that give a drug only after learning are mixed: in one, a 40 mg dose of propranolol given after the material was learned had no effect on emotional memory (LaBar & Cabeza, 2006). In a small brain-imaging study (13 young adults analyzed), Ritchey, Dolcos, and Cabeza (2008) showed people negative and neutral pictures, then tested memory after 20 minutes or after a week. Vivid remembering of the negative pictures, which the authors call recollection, faded more slowly than vivid remembering of the neutral ones. Overall recognition, simply picking out a picture as seen before, showed no such difference. During viewing, the amygdala worked more or less in step with nearby memory regions, and that link predicted memory for negative pictures more strongly at a week than at 20 minutes. The authors allow that this could instead reflect memory strength set during learning. Hamann (2001) agrees the human findings fit the animal model, but notes that only a handful of studies meet stricter tests, such as measuring memory before and after consolidation could occur.
Consolidation cannot explain everything. Emotional material can be remembered better even on an immediate test, before consolidation has had time to work. So Hamann (2001) argues for a second route acting at encoding, the moment information is first taken in, through attention and elaboration (linking new material to what you already know). One experiment shows the attention part. People studied words on their own or while also judging sounds. The extra task cut memory for neutral words but not for arousing ones (Kensinger & Corkin, 2004, as cited in Kensinger, 2004). Kensinger reports this result from a couple of behavioral experiments.
Arousal Matters More Than Pleasantness
Making material interesting is not the same as making it arousing. Researchers usually describe emotion along two dimensions: arousal, from calm to excited, and valence, from unpleasant to pleasant. Arousal, rather than valence, appears to be what engages the amygdala (LaBar & Cabeza, 2006). A frightening moment and a thrilling one can both qualify, though Hamann (2001) notes the evidence is stronger for negative material.
Valence on its own can still help, by another route. In the relatively few studies that test it, negative words low in arousal were also remembered better than neutral words. Kensinger (2004) points to elaboration as the likely route. This advantage depends more on attention than the arousal effect does: in the studies Kensinger reviews, dividing attention reduced or removed it. Young and older adults both remembered arousing words better than neutral ones, but only the young showed the advantage for negative, low-arousal words (Krendl et al., 2003, as cited in Kensinger, 2004). Whether pleasant, low-arousal material gains in the same way is still open.
None of these reviews tested ordinary curiosity directly. My reading is that mild interest works, if at all, through elaboration, the subject of an earlier post in this series, which depends on effort and attention. A pleasant study session is unlikely to be running the machinery that made the car-accident slides stick.
When Vivid Memories Mislead
Arousal also changes what gets kept. Attention settles on the central gist of an emotional event, and peripheral details lose out (LaBar & Cabeza, 2006). Eyewitness researchers call this weapon focus: witnesses often remember the weapon but not the culprit’s clothing (Kensinger, 2004). For a student, the parallel might be a gripping clinical story that survives while the lab values and the reasoning around it fade. That parallel is my inference; none of these studies used course material.
Confidence is the second problem. Faul and LaBar (2023), reviewing research on mood and memory, note that emotional memories can carry high confidence despite being inaccurate (Rimmele et al., 2011, as cited in Faul & LaBar, 2023). They also describe word-list studies in which an induced mood produced false memories for mood-matching words that had never been presented. People often reported these as confidently remembered experiences, and in one study the effect lasted a week (Knott & Thorley, 2014, as cited in Faul & LaBar, 2023). That work concerns mood, a longer-lasting state, more than single emotional events. The warning carries over all the same: a vivid memory feels trustworthy, and that feeling is a poor guide to whether it is correct.
Part of the advantage may even come from how people answer. In one recognition study, people said they “remembered” negative words more often than other words. They were also more willing to call a negative word “old,” meaning seen before, whether or not they had actually seen it. An analysis that separated memory from that willingness suggested the willingness mainly caused the extra “remember” responses (Dougal & Rotello, 2007, as cited in Crowley et al., 2019). Crowley and colleagues acknowledge this response-bias problem and still conclude that emotion strengthens memory.
Stress, Pressure and the Exam Room
More arousal does not mean more memory. In LaBar and Cabeza’s (2006) summary, brief rises in the stress hormone cortisol during encoding generally improve emotional memory, though some studies found the same effect for neutral material, or the opposite effect. Cortisol at the time of retrieval impairs recall. The effects also depend on dose, typically along an inverted U: benefits rise up to a point, then fall. Hamann (2001) adds that arousal can sometimes impair memory, particularly when it brings high levels of stress and cortisol.
Pressure during practice carries its own cost. Hinze and Rapp (2014, as cited in Weinstein et al., 2018) gave students practice quizzes under high or low pressure. Both groups scored the same on the quizzes, but the high-pressure group did worse on a later test. Test anxiety, Weinstein and colleagues conclude, may reduce the learning benefit of retrieval practice. They concede that removing every high-pressure test is probably impossible; instructors can still add low-stakes chances to practice retrieval.
How students handle anxiety may matter more than how much they feel. Hattie and Donoghue (2016) synthesized 28 meta-analyses of learners’ dispositions, where each meta-analysis pools the results of many studies. Self-efficacy, a student’s belief in their own competence, had by far the largest link with achievement, and reduced anxiety came well behind. What counts, the authors argue, is less the level of anxiety than the coping strategies a student develops, such as learning to cope with errors and seeking help. Building students’ confidence with hard problems, they add, can also help reduce anxiety.
Using This Tomorrow
Almost everything above comes from laboratory words, pictures and slide shows, and it measures retention of those items. None of it tests understanding of a complex system or transfer to new problems. No source here supports manufacturing emotion before a study session, and the cortisol findings suggest stress could backfire.
Check vivid memories against the source. The cases you remember best feel the most trustworthy, and that feeling says little about accuracy, so before you rely on one, look up the details in the chart, the guideline or your notes.
When a story sticks, rebuild what surrounds it on purpose. Go back to the principle and the numbers it illustrated, and practice retrieving those along with the story.
In the divided-attention study, the extra task hurt memory for neutral words while arousing words held up. So give dull material like the enzyme tables your undivided attention, with no lecture audio or open chat running in the background. Interest will not carry that material for you; retrieval practice and spacing have to.
Treat exam-day stress as partly a recall problem. In the laboratory studies LaBar and Cabeza summarize, cortisol at the moment of retrieval impaired recall, though none of these studies tested real exams. Separately, Hattie and Donoghue point to coping strategies, such as learning to handle errors and seeking help, as mattering more than the anxiety itself. Treat your self-quizzes as practice too, with no scores to post or compare.
If you teach, keep practice quizzes low-stakes. When you open a lecture with a dramatic case, put the principle at the center of the story rather than in a side detail. My guess is that students keep whatever the drama is about.
What This Looked Like for Me
Some cases I will never forget: a patient who decompensated (deteriorated suddenly) in front of me during my intern year, a missed finding I had to own in a morning conference, and a save that came down to a decision I almost got wrong. The clinical details around those moments are still sharp years later, and I never made flashcards for any of them. The enzyme tables and drug panels I ground through for Step 1, the first U.S. medical licensing exam, needed relentless review to hold, and many of them faded once the exam was over.
I have learned to distrust that vividness as much as I value it. More than once I have been certain I remembered a case a particular way, then reread the chart and found I had compressed the timeline or sharpened a moment that was messier at the time. The central event was intact. What had shifted were the details around it, which resembles the gist-and-periphery pattern in the studies.
I also can’t isolate emotion as the cause. The cases that stuck were the ones I discussed, retrieved, taught and revisited, so arousal is tangled up with retrieval practice and elaboration. Even so, two things hold. Emotionally charged cases deserve more checking than their vividness invites, and the dull parts of medicine still need deliberate work.
The Core Lesson
Emotional arousal helps memories last, through attention at encoding and, on strong evidence in rats and thinner evidence in people, through stress hormones acting on consolidation. It keeps the gist and drops the periphery, and the confidence it brings says little about accuracy. Brief arousal can help memory, but pressure during practice and stress at recall can hurt it. All of this concerns retention rather than transfer. Check what your most vivid memories tell you, and rely on retrieval practice and spacing for material that has no story attached.
The next post turns to another way memory misleads: proactive interference, where what you already know gets in the way of what you learn next.
This post is part of a series on the science of learning. Each post covers one evidence-based principle and how to apply it to your own studying. Follow us on Instagram @acemeded to keep up with future blog posts and related content
References
Crowley, R., Bendor, D., & Javadi, A.-H. (2019). A review of neurobiological factors underlying the selective enhancement of memory at encoding, consolidation, and retrieval. Progress in Neurobiology, 179, Article 101615. https://doi.org/10.1016/j.pneurobio.2019.04.004
Faul, L., & LaBar, K. S. (2023). Mood-congruent memory revisited. Psychological Review, 130(6), 1421–1456. https://doi.org/10.1037/rev0000394
Hamann, S. (2001). Cognitive and neural mechanisms of emotional memory. Trends in Cognitive Sciences, 5(9), 394–400. https://doi.org/10.1016/S1364-6613(00)01707-1
Hattie, J. A. C., & Donoghue, G. M. (2016). Learning strategies: A synthesis and conceptual model. npj Science of Learning, 1, Article 16013. https://doi.org/10.1038/npjscilearn.2016.13
Kensinger, E. A. (2004). Remembering emotional experiences: The contribution of valence and arousal. Reviews in the Neurosciences, 15(4), 241–253. https://doi.org/10.1515/REVNEURO.2004.15.4.241
LaBar, K. S., & Cabeza, R. (2006). Cognitive neuroscience of emotional memory. Nature Reviews Neuroscience, 7(1), 54–64. https://doi.org/10.1038/nrn1825
McGaugh, J. L. (2004). The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annual Review of Neuroscience, 27, 1–28. https://doi.org/10.1146/annurev.neuro.27.070203.144157
Ritchey, M., Dolcos, F., & Cabeza, R. (2008). Role of amygdala connectivity in the persistence of emotional memories over time: An event-related fMRI investigation. Cerebral Cortex, 18(11), 2494–2504. https://doi.org/10.1093/cercor/bhm262
Weinstein, Y., Madan, C. R., & Sumeracki, M. A. (2018). Teaching the science of learning. Cognitive Research: Principles and Implications, 3, Article 2. https://doi.org/10.1186/s41235-017-0087-y
