Your music playlist might reveal subtle clues about your intelligence

A new study published in the Journal of Intelligence suggests that a person’s everyday music listening habits contain subtle clues about their general cognitive ability. Scientists discovered that the lyrics of the songs people choose to play provide more insight into their intelligence than the musical beats or melodies do. These findings provide evidence that the digital footprints we leave behind in our daily lives could eventually help approximate cognitive skills without formal testing.

Traditional intelligence assessments rely on formal tests given in highly controlled, stressful environments. Yet, cognitive abilities are used constantly to navigate the complexities of everyday life outside of the laboratory.

With smartphones and digital apps capturing so much of what we do, researchers saw an opportunity to study cognitive ability in a natural setting. They chose to focus on music listening because it is a very common daily activity that engages multiple brain networks involving memory, emotion, and auditory processing.

Past research linking music to intelligence has mostly relied on laboratory experiments or self-reported surveys. In those settings, people might misremember what they listen to or claim to like sophisticated music to look good. By using digital tracking data, the scientists aimed to capture exactly what people were listening to in the real world.

“Most research on cognitive abilities, or intelligence in simpler terms, focuses on situations where people try to perform their best, such as tests, school performance, or job tasks. Because of that, we know a lot about how cognitive abilities relate to achievement, but much less about whether they appear in everyday, low-stakes behaviors,” said study author Larissa Sust, a postdoctoral researcher at Ludwig Maximilian University of Munich.

“At the same time, many daily activities now leave digital traces that allow us to study such real-life behavior more naturally than before. Our study was motivated by this gap: we wanted to see whether patterns in an everyday digital activity might also reflect differences in cognitive ability. As a starting point, we chose music listening, which is a common behavior that can easily be tracked on smartphones using custom research applications.”

The researchers tracked the smartphone usage of 185 participants over a period of five months. They utilized a custom research application installed on the participants’ personal phones to log every song played.

The participants also completed a short cognitive ability test on their smartphones. This test measured their capacity for fluid reasoning, vocabulary comprehension, and mathematical knowledge. These skills together make up a person’s general cognitive ability, which reflects how well someone can think rationally and adapt to new situations.

Over the course of the study, the participants listened to 58,247 unique songs. The researchers then gathered detailed information about these tracks from online music databases like Spotify. They extracted audio characteristics, such as the tempo and the acoustic qualities of the sound.

They also analyzed the lyrical content of the songs using a specialized linguistic tool. This tool categorized the words in the lyrics based on psychological themes, emotional tone, and social references. In total, the scientists gathered 215 different features related to audio, lyrics, and general listening habits for each participant.

To make sense of this massive amount of data, the researchers employed machine learning. Machine learning is a type of artificial intelligence where computer programs analyze large sets of data to identify complex patterns. They trained these computer models to see if the music listening features could predict a participant’s score on the cognitive ability test.

The researchers tested different types of computer algorithms. Only the complex, nonlinear models successfully detected meaningful patterns in the data. This suggests that the relationship between music habits and intelligence is highly intricate rather than simple and direct.

The computer models detected a small but reliable link between a person’s music listening behavior and their cognitive test scores. The most informative predictors were not the musical sounds, but the words within the songs. The lyrical preferences of the participants provided the strongest evidence of their cognitive ability.

“When we looked more closely at how our prediction models worked and which aspects of music listening were most informative, one finding surprised us,” Sust told PsyPost. “The lyrics of the songs people listened to were more useful for predicting cognitive ability than the musical features.”

“In other words, the themes and language used in the lyrics seemed to matter more than aspects like tempo or musical key. This was unexpected because previous research often suggested that melodic preferences play a larger role (e.g., when predicting personality traits), and many people assume that intelligence is mainly reflected in preferences for certain genres, such as classical or jazz music.”

Specifically, the models found that people who listened to songs with less positive emotional tones tended to have higher predicted intelligence scores. The researchers suggest that sad or melancholic music might appeal to those who use music for introspection and reflection.

Songs with lyrics focused on the present moment, perceived honesty, and home-related topics were also associated with higher cognitive ability. On the other hand, preferring lyrics with many social words or tentative language tended to predict lower intelligence scores.

Audio characteristics contributed very little to predicting cognitive ability, with one notable exception. The models found that a preference for songs with low liveness was a strong predictor of higher intelligence. Liveness refers to the probability that a track was recorded in front of a live audience.

The scientists propose that live recordings are often highly energetic and less controlled. Individuals with higher cognitive ability might prefer studio recordings because they often use music for focused, intellectual engagement rather than high-energy stimulation.

Listening habits also played a role in the predictions. Participants who spent more time overall listening to music tended to have higher intelligence scores. Additionally, preferring songs in languages other than German, which was the native language of the sample, was associated with higher cognitive ability.

“One key takeaway is that cognitive abilities (or intelligence) may be reflected not only in tests or high-stakes performance but also subtly in everyday behavior,” Sust explained. “In our study, patterns in people’s music listening contained small but detectable signals related to their cognitive ability, suggesting that the digital traces we leave behind in daily life could potentially help approximate intelligence.”

“While music listening alone provides only limited information, combining multiple types of digital behavior (e.g., what books people read, what places they visit) in the future might make such predictions more accurate and could eventually support adaptive digital tools or early detection of cognitive decline.”

While these patterns are interesting, the researchers note some potential misinterpretations and limitations. The predictive power of music listening alone was quite small, meaning an app cannot accurately judge a person’s intelligence just by looking at their playlist.

“On their own, these effects are therefore likely not strong enough to be practically useful,” Sust noted. “However, they suggest that everyday digital behavior may contain small signals of cognitive differences, which could become more meaningful if combined with many other types of behavioral data.”

The relationships observed in the study are purely correlational, meaning that listening to certain music does not cause a person to become smarter or vice versa. The researchers caution that other unmeasured factors, such as a person’s age, could be influencing both their intelligence test scores and their music preferences.

“An important caveat is that the associations we found may be influenced by other factors, known as confounding variables,” Sust said. “For example, age could play a role, because it is related both to cognitive abilities and to the kinds of music people tend to listen to. We are currently working on follow-up analyses to better understand and account for such effects.”

The study, “Deep Beats, Deep Thoughts? Predicting General Cognitive Ability from Natural Music-Listening Behavior,” was authored by Larissa Sust, Maximilian Bergmann, Markus Bühner, and Ramona Schoedel.