Abstract:

Since the last century, scholars have increasingly focused on examining how Mozart’s music affects people’s cognitive performance, leading to rapid growth in the empirical literature on the Mozart effect. However, the effect size reported in empirical studies has been inconsistent. To address this, we conducted a meta-analysis based on a systematic and comprehensive review of studies on the impact of classical music, seeking to determine its influence on cognitive performance and the underlying mechanisms at work. We also investigated whether the characteristics of research participants (e.g., age group, gender, cultural context) and elements of experimental design (e.g., type of experimental design, types of control music, the order of music, cognitive task and cerebral hemisphere) moderate the magnitude of the Mozart effect.

We identified studies by searching Web of Science, PubMed, ProQuest, WanFang, and China National Knowledge Infrastructure from 1993 to 2022 using the following terms: (“Mozart effect” OR “Mozart music” OR “music effect” OR “classical music”) AND (cognit* OR intellig* OR spati*). Our selection criteria were as follows: (1) the study reported original empirical findings; (2) at least two out of three possible treatments (listening to Mozart's Sonata KV 448, other classical music, or silence/other sounds) were administered to the groups; (3) the study involved the generalized Mozart effect and cognitive performance; (4) participants were the general public, excluding clinical or animal samples; (5) the study was written in either Chinese or English (the languages spoken by the authors).

Ninety-one studies (with a total of 172 independent effect sizes and 7,159 participants) were included in the meta-analysis. Given that effect size could be influenced by participant characteristics (e.g., age, gender, cultural context), we applied a random-effects model. After coding the data, the “metafor” package (version 3.4.0) in R software was used to evaluate the total effect size of classical music and to analyze the publication bias test and moderating effects.

The results showed that classical music improved cognitive task performance with a small effect (g = 0.36, 95% CI [0.24, 0.49]). The impact of publication bias was minimal, and the major findings remained valid. Additionally, the moderation analyses revealed that the strength of the relationship was moderated by age group, cultural context, type of experimental design, and dominant hemisphere of the brain. Specifically, the effect size of Chinese subjects was significantly larger than that of foreign subjects (g: 0.64 > 0.27, p = 0.018), and the effect size of preoperational stage children (3~6 years) was the largest (g= 1.10). Compared with the within-subject design, the between-subject effect was significantly greater (g: 0.48 > 0.22, p = 0.037). The right hemisphere also performed much better than the left (g: 0.44 > 0.08, p = 0.019). Moreover, gender interacted with age group, cultural context and cerebral hemisphere. The direct priming hypothesis received more robust support from this meta-analysis (g: 1.29 > 0.34, p = 0.045).

To summarize, this study makes several important theoretical advances. First, this study systematically assessed the effects of listening to classical music on cognitive performance basing on a broad definition of Mozart effect, covering a wider range of musical genres and cognitive task types. It bridged the limitations of existing meta-analyses, clarified the debate on the reliability and scientific validity of the Mozart effect, and laid the groundwork for in-depth discussions. More importantly, this paper was the first to compare the effect sizes based on the "Direct Priming Hypothesis" and the "Arousal-mood Hypothesis", indicating the former to be more adept at explaining the Mozart effect. This provided a clearer theoretical guide for future researches. Finally, by examining the moderation effects of several factors, this paper explained why previous literature on the Mozart effect has reported inconsistent findings and provided more targeted design guidance for future studies. Beyond its theoretical advancements, the current paper’s results also have practical implications, such as the implications of age group differences and their interactions for children's cognitive development. The results can also aid in utilizing music education more effectively to boost cognitive performance. Future researches are encouraged to examine the long-term facilitative effect of classical music on cognitive performance, to explore the role of music preference in cognitive facilitation, and to explore more underlying moderators for the intervention effect size, such as subjects' personality traits, familiarity with music, and difficulty of the cognitive task.

Key words: Mozart effect, classical music, meta-analysis, cognitive performance