Scott, L. (1992). Attention and perseverance behaviors of preschool children enrolled in Suzuki violin lessons and other activities. Journal of Research in Music Education, 40, 225-235.


This study examines the effects of Suzuki violin lessons on preschool children’s development of attention and persevering behaviors, behaviors that increase students’ capacity to learn and rate of learning. To examine the impact of Suzuki violin lessons, the study collected and compared data on students receiving: (1) individual Suzuki violin lessons; (2) individual and group Suzuki violin lessons; (3) creative movement classes; (4) preschool only; and (5) no preschool or classes. This study also examines the relationship between teacher reinforcement and student attending behavior.

Key Findings:

  • Children in both Suzuki groups (those receiving individual or individual and group Suzuki violin instruction) scored higher on all attention tasks than did children in all other groups.
  • Children in both Suzuki groups spent significantly more time on the perseverance task than did children in all other groups.
  • Teachers of children in both Suzuki groups demonstrated significantly more teacher approval than teachers of children in all other conditions.

Significance of the Findings:

Findings suggest a potential relationship between Suzuki violin instruction and attention and perseverance behaviors in preschool children. The Suzuki method may provide a pedagogy and philosophy that positively affects attitudes concerning parental involvement and the learning potential of all children, as well as stresses the importance of reinforcement and mastery of each learning step. These findings corroborate findings of other studies adding to the evidence that Suzuki lessons have a positive effect on attending behaviors that may transfer to other instructional settings.


The study compared students in five groups or research “conditions,” students receiving: (1) individual Suzuki violin lessons; (2) individual and group Suzuki violin lessons; (3) creative movement classes; (4) preschool only; and (5) no preschool or classes. Children were chosen for their availability. Students and teachers in the first three conditions were observed in regular instructional settings, while those in conditions four and five were observed during tasks only.

Researchers used observation measures to document and analyze what they observed. Inter-rater reliability was calculated for 20% of each of the their observational measures and ranged from 89.5% to 97%. The attention task used in this study was designed to measure students’ ability to detect unpredictable signals in the form of the presence or absence of colored light in a 15-minute session and scored for the number of correct and incorrect responses. The perseverance task measured the time subjects spent replicating a block model and scores were given for likeness to the model. Attention and perseverance behaviors were analyzed through the observation of videotaped recordings of subjects performing the two tasks designed by the experimenter. Analysis of classroom and lesson videotapes provided data on student and teacher behavior. The researchers analyzed data using one-way analysis of variance (ANOVA) and procedures for multiple comparisons (Neuman-Keuls and Dunn’s procedures).

Limitations of the Research:

The most serious limitation of this quasi-experimental design is the “self-selection” of students to the five groups (or conditions) which the researchers studied. Demographic characteristics of the participants, including socioeconomic status (SES) and ethnicity/culture, as well as cognitive and motivational factors were not statistically controlled. In other words, home environment is likely a confounding factor. Additionally, the tasks designed for use in this study have unknown psychometric properties (e.g., construct validity).

Questions to Guide New Research:

Are these findings replicable with other age groups? Are these findings replicable with other dependent measures of attention and perseverance? Are these findings replicable using a randomized design?