Introduction
Rhythm is essential to perceiving music and engaging with music1. Most humans are capable of perceiving structure in rhythm, both in the form of rhythmic patterns and as a regular beat. However, there are large individual differences in rhythmic abilities, and these do not necessarily depend on musical training2. Examining such individual differences is not just important for our understanding of human musicality, but also since rhythmic abilities have been linked to atypical language and motor development, and the extent to which patients are susceptible to rhythmic interventions. Several tests have been developed to test rhythmic abilities in the general population, but it is unclear whether these tests measure similar or unconnected aspects of rhythm perception, be it interval-based perception, the perception of rhythmic patterns, or beat-based perception3. We aim to 1) examine the validity of existing rhythm abilities tests, and 2) examine the presence of multiple rhythmic skills.
Methods In two experiments, participants were tested on often-used tests of rhythmic abilities. Both experiments included tests of duration discrimination and anisochrony detection, as previously implemented in the Battery for the Assessment of Auditory Sensorimotor and Timing Abilities (BAASTA4). Both experiments also included three tests from the Harvard Beat Assessment Test (H-BAT5), and the Beat Alignment Test, as implemented in
the Goldsmith Musical Sophistication Index2. In Experiment 2, we additionally included a rhythm discrimination test with both metric and nonmetric patterns6. Experiment 1 was lab-based, and a total of 121 participants completed all tests. In Experiment 2, we aimed to replicate the lab-based results in an online experiment, in which a total of 306 participants completed all tests.
Results
In both experiments, we find that while most tests aimed at measuring beat-based perception correlate with each other, many of these tests also correlate with the duration discrimination tests, questioning their specificity. Also, factor analysis did not reveal a clear separation between tests measuring beat-based perception, intervalbased perception, and the perception of rhythmic patterns.
Conclusion
Our results suggest that tests aimed at rhythmic abilities capture a broad range of processes, and that different rhythmic skills are related within individuals. Future research may focus on improving the design of tests of rhythmic abilities, and on examining in how much rhythmic abilities are subserved by a general underlying mechanism, or rather, with more focused testing, may consist of separate skills.
Rhythm is essential to perceiving music and engaging with music1. Most humans are capable of perceiving structure in rhythm, both in the form of rhythmic patterns and as a regular beat. However, there are large individual differences in rhythmic abilities, and these do not necessarily depend on musical training2. Examining such individual differences is not just important for our understanding of human musicality, but also since rhythmic abilities have been linked to atypical language and motor development, and the extent to which patients are susceptible to rhythmic interventions. Several tests have been developed to test rhythmic abilities in the general population, but it is unclear whether these tests measure similar or unconnected aspects of rhythm perception, be it interval-based perception, the perception of rhythmic patterns, or beat-based perception3. We aim to 1) examine the validity of existing rhythm abilities tests, and 2) examine the presence of multiple rhythmic skills.
Methods In two experiments, participants were tested on often-used tests of rhythmic abilities. Both experiments included tests of duration discrimination and anisochrony detection, as previously implemented in the Battery for the Assessment of Auditory Sensorimotor and Timing Abilities (BAASTA4). Both experiments also included three tests from the Harvard Beat Assessment Test (H-BAT5), and the Beat Alignment Test, as implemented in
the Goldsmith Musical Sophistication Index2. In Experiment 2, we additionally included a rhythm discrimination test with both metric and nonmetric patterns6. Experiment 1 was lab-based, and a total of 121 participants completed all tests. In Experiment 2, we aimed to replicate the lab-based results in an online experiment, in which a total of 306 participants completed all tests.
Results
In both experiments, we find that while most tests aimed at measuring beat-based perception correlate with each other, many of these tests also correlate with the duration discrimination tests, questioning their specificity. Also, factor analysis did not reveal a clear separation between tests measuring beat-based perception, intervalbased perception, and the perception of rhythmic patterns.
Conclusion
Our results suggest that tests aimed at rhythmic abilities capture a broad range of processes, and that different rhythmic skills are related within individuals. Future research may focus on improving the design of tests of rhythmic abilities, and on examining in how much rhythmic abilities are subserved by a general underlying mechanism, or rather, with more focused testing, may consist of separate skills.