Journal of Expertise Predicting Musical Aptitude and Achievement: 2019. Vol. 2(3) © 2019. The authors Practice, Teaching, and Intelligence license this article under the terms of the Creative Commons Miriam A. Mosing1,2, David Z. Hambrick3, and Fredrik Ullén1 Attribution 3.0 License. ISSN 2573-2773 1Department of Neuroscience, Karolinska Institutet, Sweden 2Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Sweden 3Department of Psychology, Michigan State University, USA Correspondence: Miriam A. Mosing, [email protected] Abstract Studies of expertise have traditionally had a strong focus on the role of one single factor, i.e. long-term deliberate practice, for expert performance. However, recent empirical and theoretical work strongly suggests that expertise is a function of many variables that may have practice-independent effects on performance, but also moderate the efficacy of practice itself. Here we study such interaction effects in a large cohort (N > 4,500) of Swedish twins, using music as a model domain, and measured expert performance (musical auditory discrimination) as well as self-reported real-life achievement as indices of expertise. Specifically, we test two recently proposed hypotheses, i.e. (1) that the efficacy of practice increases if the individual also takes part in teacher-led lessons, and (2) that practice efficacy increases with higher intelligence. The results did not support the first hypothesis. Both practice and frequency of music lessons had positive associations with the two measures of expertise but, contrary to predictions, the interaction between them was negative, i.e. the effect of each practiced hour decreased with more lessons. In contrast, the second hypothesis was supported by the data, i.e. we found a positive interaction between practice and intelligence, suggesting that higher cognitive ability is related to more efficient practice behaviors. Together the results further support that domain-specific expertise is a complex outcome, which depends on an interplay of a variety of factors. Keywords expertise, training, music, IQ, ability Introduction Deliberate practice theory, as proposed by expert performance is essentially a function of a Ericsson and colleagues (Ericsson, Krampe, & single factor: the amount of deliberate practice Tesch-Römer, 1993; Ericsson & Pool, 2016; an individual has accumulated (Ericsson, 2007; Ericsson & Smith, 1991; Ericsson & Ward, Ericsson et al., 1993; Ericsson & Pool, 2016). 2007; Feltovich, Prietula, & Ericsson, 2006), Recent empirical work in the field has historically been one of the most influential challenges this view. Several meta-analyses theoretical frameworks in expertise research. A show that, regardless of domain, deliberate core element of this theory is that—with a few practice explains only a low or modest exceptions, such as body size, height, and the proportion of the variance in expert performance detrimental effects of disease and injury— (Hambrick et al., 2014; Macnamara, Hambrick, https://www.journalofexpertise.org 184 Journal of Expertise / September 2019 / vol. 2, no. 3 Mosing, Hambrick, & Ullén (2019) Predictors of Musical Expertise & Oswald, 2014; Platz, Kopiez, Lehmann, & (Drayna, Manichaikul, de Lange, Snieder, & Wolf, 2014), the highest reported estimate to Spector, 2001; Mosing, Madison, et al., 2014; date being 36% for music in Platz et al. (2014). Mosing, Pedersen, Madison, & Ullén, 2014; In other words, most of the variance in expert Mosing, Verweij, Madison, & Ullén, 2016), performance appears to be independent of correlations between practice and performance deliberate practice. This non-practice-related (Mosing, Madison, et al., 2014), and variance is likely to reflect individual specialization within a domain of expertise differences in psychological and physical traits (Mosing & Ullén, 2018). relevant to performance. In musicians, for Taken together, these findings strongly instance, working memory capacity predicts suggest that multifactorial models are needed to sight-reading over and above the effects of account for existing data and to generate new, deliberate practice (Meinz & Hambrick, 2010), fruitful, and testable hypotheses. Building on while other studies have found force control previous work (see e.g. Ackerman, 1996; during piano strokes in pianists to be unrelated Cattell, 1987; Gagné, 2013), we have recently to musical training, but correlated with weight proposed one such model, the Multifactorial discrimination ability (Hosoda & Furuya, 2016). Gene-environment Interaction Model (the For chess, a recent meta-analysis found skill MGIM; Ullén et al., 2016). level to be related to general fluid intelligence Multifactorial expertise models such as the and short-term/working memory (Burgoyne et MGIM are, of course, perfectly consistent with al., 2016). the common observation that long-term practice Furthermore, the efficacy of practice itself can result in impressive improvements of may be moderated by other factors. For performance. However, another controversial example, Miksza has found that impulsivity is claim of deliberate practice theory is that related to less well-structured practice behaviors expertise acquisition crucially depends on a in musicians (Miksza, 2006, 2011), while other certain type of practice – that is, deliberate researchers have found positive associations practice – while engaging in other types of between music practice and openness to domain-specific activities, such as those for experience (Corrigall & Schellenberg, 2015; extrinsic reward or pleasure, is considerably less Corrigall, Schellenberg, & Misura, 2013; effective (Boot & Ericsson, 2013; Ericsson et Swaminathan & Schellenberg, 2018). We have al., 1993). The relatively few studies that have previously suggested (Ullén, Hambrick, & directly compared the effects of deliberate Mosing, 2016) that higher intelligence, which is practice and other domain-related activities on related to sustained attention and metacognition performance do not provide consistent support (Schweizer & Moosbrugger, 2004; Schweizer, for this claim that deliberate practice is the Moosbrugger, & Goldhammer, 2005; Stankov, primary predictor of performance. For instance, 2000), may facilitate effective, goal-oriented Howard (2012) found that number of played practice strategies. Finally, recent work using games predicted chess rating much more genetically informative samples have confirmed strongly than both time spent on practice and that individual differences in expert coaching. Similarly, Sonnentag and Kleine performance and its acquisition are no exception (2000) found that sales performance in to the general rule that human variation arises insurance agents was better predicted by number from an interplay between genetic and non- of handled cases than by measures of deliberate genetic factors (Hambrick, Macnamara, practice. Campitelli, Ullén, & Mosing, 2016; Polderman Addressing this issue is complicated by the et al., 2015; Ullén et al., 2016). Twin studies fact that the distinction between deliberate have thus demonstrated substantial genetic practice and other forms of practice often is influences on practice (Hambrick & Tucker- unclear (Howard, 2009). Furthermore, in the Drob, 2015; Mosing, Madison, Pedersen, Kuja- expertise literature, authors have not always Halkola, & Ullén, 2014), expert performance been consistent in their definitions of deliberate https://www.journalofexpertise.org 185 Journal of Expertise / September 2019 / vol. 2, no.3 Mosing, Hambrick, & Ullén (2019) Predictors of Musical Expertise practice. Traditionally, deliberate practice has music lessons in the prediction of musical been defined as focused and effortful practice achievement, in line with Ericsson’s recent activities that are designed either by the claims that teacher-led deliberate practice is a performers themselves or by external agents stronger predictor of performance than is such as teachers or coaches, and pursued with purposeful practice alone (Ericsson, 2016; the explicit goal to improve performance Ericsson & Pool, 2016). Second, we test (Ericsson et al., 1993; Keith & Ericsson, 2007). whether there is an interaction between practice More recently, however, it has been suggested and intelligence in the prediction of musical that deliberate practice by definition must expertise, as would be suggested by the involve a teacher or coach who is involved in hypothesis by Ullén and coworkers (2016) that designing practice activities, introducing the higher intelligence is related to more effective new term purposeful practice for practice practice strategies. activities that in previous studies were considered as deliberate practice but did not Methods involve the participation of a teacher (Ericsson, Sample 2016; Ericsson & Pool, 2016). Data for the present study were collected in Disregarding these terminological 2012 and 2013 as part of a web survey confusions, the above discussion highlights an administered to a cohort of young adult twins interesting general possibility, which is that the registered with the Swedish Twin Registry aged efficacy of practice activities is modulated by 27 and 54 (Lichtenstein et al., 2002; other factors, including both traits of the Lichtenstein et al., 2006). The web survey was individual, such as intelligence, and designed to collect extensive information on environmental influences, such as regular music-related variables and the study was interactions with a teacher. In line with this, approved