Is Earlier Better? Investigating Sensitive Periods for Musical Training in School-Aged Children

Sensitive periods for musical training have been proposed, such that starting lessons in earlier childhood predicts better rhythm and melody skills in adult musicians. The goal of this thesis was to evaluate the effects of age of start (AoS) in musically trained children. We assessed whether children with an early AoS showed advantages over those who began later, with equivalent training. An inherent psychometric challenge with children is controlling for maturation. We started with tasks developed in our lab for adults, adapted them for school-aged children, and administered them to 213 children with and without music training. We calculated age-based scores, and estimated reliability and validity (Study 1). We then used age-based scores to assess contributions of AoS, training, and cognitive abilities to performance (Study 2). In Study 1, the children’s Rhythm Synchronization Task (c-RST) and Melody Discrimination Task (c-MDT) were found to have adequate convergent validity with adult analogues. Further, musically-trained children outperformed those without training, replicating findings of a ‘musician advantage’ for auditory tasks. The effect of age on task performance was largest for the c-RST, which poses the highest demands on auditory-motor integration. In Study 2, we investigated the influence of AoS on task performance at three cutoffs (AoS of 5, 6, and 7). We controlled for music training and other variables that predict musical engagement and task performance. We found a statistically significant effect of AoS and global cognitive ability, but only for the easiest task condition, Simple Melody discrimination. No AoS effects were found for the more difficult Transposed Melody discrimination, or for the c-RST, and these two were independently predicted by auditory-verbal working memory. Taken together, our results support a multidimensional model of musical task performance in childhood that includes the interaction of developmental, training-related and cognitive factors. The knowledge gained in this thesis may facilitate the application of musical training to other cognitive domains including language. Music and language share neural substrates and develop according to similar principles. The underlying mechanisms of language and music, and potential applications of our tasks to research on transfer effects, are discussed