Spelke, E. (2008). Effects of music instruction on developing cognitive systems at the foundations of mathematics and science. In Asbury, C. & Rich, B. (Eds.), Learning, Arts, and the Brain, (pp. 17–49). New York/Washington, D.C.: The Dana Foundation.


In this study, the researcher conducted three experiments that examined the relationship between the cognitive systems involved in music and those involved in math and science abilities in children. The first experiment compared the math abilities of students receiving music instruction to those in a soccer program. No clear connections were found after accounting for other demographic variables. The second experiment compared the math abilities of students with intensive music training to those with little music training. On measures of small exact numbers, large approximate number, and number words there were no associations, but the researcher did find a positive correlation in measures related to spatial cognition. The third experiment compared the music abilities of students participating in different arts forms: music, dance, theater, creative writing, and visual arts. Students with music training outperformed students without music training on tasks related to spatial cognition and reasoning. Students with visual arts training performed well on measures of geometry in visual forms.

Key Findings:

  • Results from two of the three experiments indicated that musical training is associated with higher skills in representation and reasoning in geometry as well as estimation.
  • In other art forms (focused on in the third experiment), music and dance majors outperformed other majors in estimation, number line, map activities, and geometrical invariants. Visual arts majors outperformed creative writing and theater majors on geometrical invariants. Additionally, statistical (regression) analysis showed a relationship between the amount of visual arts training and accuracy on the geometrical invariants tasks.
  • Musicians outperformed non-musicians on the geometrical invariants activity. Musical training also helped predict positive outcomes on the map activity.
  • The first experiment yielded no relationship between musical training and mathematic ability when demographic variables were factored into the analysis.

Significance of the Findings:

This study shows clear correlation between music, visual art, dance, theater, and creative writing training and specific mathematical abilities suggesting a relationship between the cognitive skills needed for each subject.


For all three experiments, the researcher administered six different measures of mathematic and spatial abilities to participants. The Multiple Object Tracking (MOT) task (Pylyshyn & Storm, 1988) measured participants’ ability to represent small, exact numbers of objects. Other tests looked at participants’ ability to compare large numbers, work with geometrical invariants, estimate, place items on a map, and place items on a number line.

The first experiment compared students in a music program to those in a soccer program to see the effects of mild to moderate musical training. There were 85 students ages five to 17 in this study. Music and sports training were calculated by weeks of participation in the program. Hierarchical regression analysis for each test was used to see how program participation might impact student abilities.

The second experiment included 61 children ages eight to 13, who varied in levels of music training as measured by weeks of music participation. There were 32 children with high levels of training and the remainder had low levels of training. The researcher conducted an ANCOVA for each test taken.

The third experiment included 80 students from a private arts high school who participated in one of five different arts programs: music, dance, theater, creative writing, and visual arts. Data were collected on their training experience and intensity of focus through self-report. Regression analysis and ANCOVA were used to analyze the data and make comparisons across training in different art forms.

Limitations of the Research:

For the first and third experiments, there were students who had experience in two or more conditions of interest. For example, in the first experiment there were many students who were enrolled in soccer and also taking music lessons, which could confound the findings. Additionally, in the second experiment, there was significant overlap between students with a music background and those who were dance majors.

Questions to Guide New Research:

Can music training improve children’s fundamental mathematical abilities? How does intensity of music instruction compare to length of music training in relation to mathematical abilities?