Music Ensemble: a large dataset on musicianship, cognition, and personality in musicians and nonmusicians

Why Studying Musicians Tells Us About the Mind

Many of us suspect that years of music lessons shape more than just our ability to play an instrument. Do musicians really think, remember, or even feel differently from nonmusicians—or are naturally gifted people simply more likely to stick with music? The Music Ensemble project tackles this question by creating one of the largest, most carefully designed collections of data ever assembled on musicians and nonmusicians, opening a powerful new window on how intensive training and everyday life experience sculpt the mind.

A Worldwide Look at Musical Expertise

The Music Ensemble dataset brings together detailed information from 1,438 young adults, aged 18 to 30, tested at 35 research sites across 16 countries in Europe, North America, South America, and Australia. Participants were divided into two clearly defined groups. Musicians had at least ten years of formal training and were actively practicing, while nonmusicians had no more than two years of lessons and had not played or sung for at least five years. To make fair comparisons, most musicians were paired with a nonmusician of the same age, gender, and education level, yielding 678 closely matched pairs. All volunteers visited a lab in person and completed the same tightly standardized testing session so that results from different countries and laboratories could be combined with confidence.

What the Researchers Measured

Each participant completed a broad range of thinking and perception tasks. These included short-term memory tests for numbers, visual patterns, and simple melodies; a demanding “2-back” attention task that taps the ability to update information in mind; a classic puzzle test of abstract reasoning; and a vocabulary test that reflects verbal knowledge. To capture musical listening skills, participants took a compact version of a specialized music perception test that measures sensitivity to tune, rhythm, timing, and tempo. Alongside these performance tasks, they filled out well-established questionnaires about their musical background and engagement, how rewarding they find music, their personality traits, and their family and personal socioeconomic circumstances.

Building a High-Quality Open Resource

The team devoted substantial effort to make sure the data are both reliable and widely usable. All procedures were preregistered in advance, meaning that the main plans for recruitment and analysis were publicly laid out before data collection began. Instructions and computer tasks were carefully translated and checked across multiple languages. The researchers systematically screened the data for errors, documented rare glitches in individual tasks, and clearly tagged any missing or corrected values. They then combined trial-by-trial records and questionnaire responses into several user-friendly files, accompanied by a detailed data dictionary and processing scripts, all freely available on the Open Science Framework platform. Checks of internal consistency show that the key questionnaires perform well across languages, and small positive correlations among the thinking tasks align with current theories of general intelligence.

What We Already Know from the Data

Although this paper mainly introduces the dataset rather than reporting many new findings, it confirms that the group definitions worked as intended. Musicians outperformed nonmusicians on all objective measures of musical perception and reported much higher levels of musical sophistication and engagement. At the same time, the dataset reveals important complexities: some measures are strongly interrelated, some participants’ self-descriptions do not perfectly match their initial group, and scores vary slightly across test sites and languages. The authors highlight these features so that future users can choose appropriate statistical methods—for example, multilevel models that take into account that people are nested within research sites—and can thoughtfully define their own inclusion criteria when asking new questions.

Why This Matters for Future Research

The Music Ensemble dataset does not claim to settle whether music training causes changes in thinking and personality or mainly reflects who chooses to become a musician. Instead, it provides the field with a shared, high-quality foundation to address such questions more rigorously. By making a large, international sample openly available, with rich information on cognition, personality, musical background, and demographics, the project allows scientists and students alike to reanalyze the same data from many angles, compare different statistical approaches, and explore new hypotheses about expertise, learning, and the mind. In short, this work turns long-standing debates about musicians and mental abilities into a testable, collaborative endeavor.

Citation: Talamini, F., Grassi, M., Altoè, G. et al. Music Ensemble: a large dataset on musicianship, cognition, and personality in musicians and nonmusicians. Sci Data 13, 473 (2026). https://doi.org/10.1038/s41597-026-06654-0

Keywords: musical expertise, cognitive abilities, personality traits, open dataset, short-term memory