Music Listening Predicted Improved Life Satisfaction in University Students During Early Stages of the COVID-19 Pandemic

Strip away commuting, the office, the gym, and the Saturday plans — strip away everything that structures a day — and what fills the silence? Media. That's the honest answer for most people. But here's the thing: if what fills that silence matters for how you feel, then the question becomes which media you reach for. A study tracking Australian university students through the early months of COVID-19 just showed that music and television both filled the void — and pulled wellbeing in completely opposite directions. Krause and colleagues at James Cook University set out to ask that exact question. When the usual coping strategies disappear overnight — the social contact, the physical routines, the structured environments — does the particular media someone turns to predict whether they feel better or worse? Music had been studied as a mood regulation tool before the pandemic, with prior work from Lonsdale and North, Boer and Fischer, and Schäfer consistently linking it to stress relief. Television, social media, and gaming were murkier — the evidence was mixed. So the team designed a study to track real students through real disruption and see what the data showed. Queensland, Australia gave them an unusually clean window. The state declared a public health emergency on January 29, 2020, shut down non-essential services and closed its borders on March 23, and tightened restrictions further to a two-person household rule by the end of that month. Universities pivoted entirely to virtual attendance for the rest of 2020. For the one hundred twenty-seven students in this study, that meant adapting to new learning environments, losing peer contact, coping with financial stress from lost employment, and spending most of their time at home. Their everyday coping infrastructure collapsed in a matter of weeks. The design Krause and colleagues built to track this was a repeated-measures study across six online questionnaires spaced two weeks apart, running from April through mid-July 2020. The first questionnaire asked about life before any COVID-related changes; each of the five that followed asked about the previous week. That yielded four hundred seventy total assessments from one hundred twenty-seven participants. The sample skewed young — the median age was twenty, though the range went up to fifty-three — and was seventy-four percent female, twenty-four percent male, and two percent non-binary, with eighty-three percent Australian nationals. At each time point, participants rated how frequently they engaged in four media activities on a five-point scale from never to at least once a day: listening to music, playing video or computer games, watching television or movies or streaming videos, and using social media. Life satisfaction was measured with a single slider item — zero to one hundred, completely dissatisfied to completely satisfied. Across all assessments, music averaged four point two out of five, social media averaged four point three four, television and streaming averaged three point eight one, and gaming averaged just one point nine eight. Life satisfaction averaged sixty-six point two six out of a hundred. To pull apart what was driving what, the team used multilevel modeling — a statistical approach that separates two fundamentally different kinds of associations. Between-person effects ask whether people who generally use more of a medium tend to have higher or lower life satisfaction than other people. Within-person effects ask whether the same individual feels better or worse in weeks when they use a medium more than is usual for them. That distinction matters enormously. A person who always listens to a lot of music and also happens to be generally upbeat will show up in a between-person comparison as a music listener with high wellbeing — but that doesn't tell you whether music caused anything. The within-person analysis is sharper: it controls for each individual's baseline and asks only whether fluctuations in their media use track fluctuations in their mood. The intraclass correlation coefficients — ICCs, the statistic that shows how much of total variability lives between people versus within people over time — confirmed that both levels mattered. Life satisfaction had an ICC of zero point five one, meaning about half its variability came from stable differences between people and half from week-to-week fluctuation within individuals. Television and streaming had the lowest ICC at zero point four seven, showing the most temporal variability. Social media was the most stable at zero point six nine. Music sat at zero point six four — relatively consistent across the study, but still with meaningful within-person movement. Across the six time points, life satisfaction was statistically lower than pre-COVID levels through most of the study, only returning to baseline-comparable levels near the end. Gaming spiked immediately after lockdown began, then faded back. Social media use rose through the middle of the study and then dropped. Television and streaming were elevated at time points two and six. Music was comparatively stable, dipping slightly at time point three. Then came the multilevel results. Between-person effects — nothing. People who averaged more media use of any kind were not systematically more or less satisfied with life than people who averaged less. That's the null you might expect: the kind of person who listens to lots of music probably differs from the kind of person who doesn't in a dozen ways that also affect wellbeing, making between-person comparisons hard to interpret. Within-person effects — that's where the findings landed. On weeks when participants listened to more music than was typical for them, they reported higher life satisfaction. The coefficient was two point nine seven on the zero-to-one-hundred scale, with a ninety-five percent confidence interval from zero point two three to five point seven one. And on weeks when participants watched more television, movies, or streaming video than was typical for them, life satisfaction went down — a coefficient of negative two point five three, with a confidence interval from negative four point eight three to negative zero point two four. Video games showed no significant within-person effect. Social media showed no significant within-person effect. Just those two, pointing in opposite directions. Hold that for a second. Music and television are both activities you do while sitting still, alone, at home. Both are common leisure activities. Both filled the pandemic hours. And yet they're associated with opposite movements in how satisfied people felt with their lives that week. That asymmetry is the finding that needs explanation. Krause and colleagues offer a mechanism rooted in prior work on music and emotional regulation. People tend to listen to music deliberately — they choose it, they control what they hear, and they often choose it specifically to shift how they feel. The literature the team draws on frames music as an active, self-directed coping tool, one that can evoke social connection even when listened to alone, with lyrics and imagined relationships providing a kind of social surrogate when real contact is cut off. More user control, the paper argues, is linked to more positive outcomes. Television and streaming, by contrast, may involve lower levels of active input. Krause and colleagues speculate that heavy television use could align more with avoidance or rumination than with active coping — watching because you don't know what else to do, rather than watching as a deliberate choice to regulate your state. That's plausible. But they are careful here, and their carefulness is worth naming explicitly: this study cannot tell you which way the arrow points. Did listening to more music cause students to feel better? Or did students who were already feeling better that week reach for music more readily? The within-person design gets closer to causation than a simple cross-sectional correlation — you're comparing each person to themselves — but it's still observational. The direction remains unresolved. Other limits the team flags include the sample being university students in one region of Australia, which constrains how far the findings generalize. Pre-COVID measures were recalled retrospectively. Only about half the sample completed five or six of the six surveys. There were no contemporaneous measures of mood, health, or other wellbeing variables that might have mediated the relationship. What survives those limits is something genuinely useful. The pandemic created an almost laboratory-like condition where everyone's routines collapsed simultaneously, making within-person variation visible in ways that are hard to engineer deliberately. What that variation revealed is that media type matters — not just media use in aggregate, but which media. Treating all screen time as equivalent, the way a lot of public conversation does, misses this entirely. The practical implication is direct: guidance about coping and leisure during periods of social isolation should distinguish between kinds of media rather than bundling them together. What remains open — and what Krause and colleagues flag as the necessary next step — is whether the apparent benefit of music comes from something inherent to music itself, or from how people use it: the control it affords, the social connection it evokes, or the emotional intentions people bring to it. That question will need real-time methods and motivational measures to crack. For now, the data say: reach for the music, not the remote. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.