Comparing psychological constructs in early specializing and sampling male hockey players

In recent years there has been a growing trend in western nations for young people participating in sport to specialize early (Feeley, Agel, & LaPrade, 2016; Jayanthi, Pinkham, Dugas, Patrick, & LaBella, 2013) and considerable research over the past two decades has focused on the potential benefits and negative outcomes associated with early specialization. Much of this work has been guided by the Developmental Model of Sport Participation (DMSP; Côté, 1999; Côté & Fraser-Thomas, 2016; Côté & Hay, 2002), which proposes two different athlete trajectories towards an elite level of performance. In the first, athletes participate in many sports as a child and progress to specialize and invest in one sport (i.e., early sampling trajectory). In the second, athletes bypass the sampling stage and begin with sport specialization during childhood (i.e., early specialization) (Côté & Fraser-Thomas, 2016). While there is no consistent definition of early specialization, a commonly adopted one suggests early specialization is year-round training in a single sport at the exclusion of other sports, occurring during the ages of 6-12 (Wiersma, 2000). In 2015, the American Orthopedic Society of Sport Medicine (AOSSM) convened a think tank of experts on early specialization and released a consensus definition. Early specialization was defined by three criteria: (a) participation in intensive training and/or competition in organized sports greater than 8 months per year, (b) participation in one sport at the exclusion of participation in other sports, (c) involving prepubertal (roughly age 12) children (LaPrade et al., 2016). The rationale for early specialization comes from the influential deliberate practice framework (Ericsson, Krampe, & Tesch-Römer, 1993). Ericsson and colleagues (1993) noted that expert musicians spent more time in practice aimed at improving performance than their non-expert counterparts. The authors suggested this specific type of training (i.e., activities that were highly effortful, designed for the purpose of improving performance and do not provide instant gratification), which they labeled ‘deliberate practice’, was the key difference between experts versus non-experts across domains. Importantly, this framework promotes that in order to become an expert, one must engage in a large quantity of deliberate practice, and the more time spent in deliberate practice, the higher the attained level of performance. Further, the authors suggested that if someone starts deliberate practice at a later age, they would be at a disadvantage compared to their peers who began earlier as they would have potentially had less opportunity to accumulate the same amount of deliberate practice hours. Whether it is beneficial in terms of expertise for an athlete to specialize early is a complex and unresolved issue in the field. However, all would agree that becoming an elite athlete requires practice. In a review of literature on sport skill acquisition, researchers concluded that high amounts of deliberate practice are integral to developing expertise across sporting domains (Baker &Young, 2014). This theory underpins the rationale for early specialization; the sooner one begins sport specific deliberate practice, the more likely they are to reach a high level of skill.

There is, however, growing concern, that early specialization can be detrimental to the physiological and psychosocial health of youth. Early specializers experience higher rates and risk of injury as well as higher levels of emotional exhaustion, dropout and burnout (Fraser-Thomas, Côté, & Deakin, 2008; Jayanthi et al., 2013; Strachan, Côté, & Deakin, 2009). Despite the strong consensus statements that have been advanced concerning early specialization (e.g. LaPrade, et al., 2016) very little is known about the differences between athletes who specialize and those who do not, with the exception of the injury and training pattern research mentioned above. In the current study, we examined differences in two key psychological constructs - personality and anxiety – which have been largely overlooked in this area.

Personality generally reflects stable psychological characteristics that differ between individuals (Larsen & Buss, 2008). The most widely employed model of personality is the five-factor model (FFM) proposed by McCrae and John (1992). The FFM suggests there are five personality traits that everyone possess; openness to experience, conscientiousness, extraversion, agreeableness and neuroticism. Openness to experience manifests in behaviours such as being intellectual, imaginative and creative. Behaviours such as being organized, practical, and prompt are markers of someone who is highly conscientious. Extraversion can be classified by behaviours such as being talkative, social, and assertive. Someone who is high on agreeableness, will display behaviours such as being sympathetic, kind, warm and sincere. Finally, neuroticism, is defined as low emotional stability, and being moody and insecure. While the examination of personality traits among athletes has a long history (e.g., Allen, Greenless, & Jones, 2011; Eglof & Gruhn, 1995; Kirkcaldy, 1982; McKelvie, Lemieux, & Stout, 2003; Piedmont, Hill, & Blanco, 1999), no research has focused specifically on personality traits in relation to children and youths’ sport trajectories.

Similarly, sport-related measures of anxiety have been largely overlooked in examinations of the consequences of early specialization. Athletes competing in individual sports have reported higher levels of competitive state anxiety (CSA) than those competing in team sports (Flowers & Brown, 2002) and athletes higher in competitive trait anxiety (CTA) respond to competition with greater levels of CSA compared to those with low CTA (Hanton, Mellalieu, & Hall, 2001). Of particular concern are the negative outcomes associated with high competitive anxiety such as decreased performance and increased fear of failure (Aoyagi, Burke, Joyner, Hardy, & Hamstra 2009; Rainey & Cunningham, 1988). It has been suggested that early specialization could lead to higher levels of competitive anxiety, given athletes invest more time and may feel greater pressure to succeed (Baker, Cobley, & Fraser-Thomas, 2009). On the contrary, because of the amount of time spent in a single sport, athletes may be more accustomed to the pressure and may, therefore, have lower levels of competitive anxiety than their non-specializing peers (e.g., those who sample several sports). Additionally, past research has shown competitive anxiety and personality are important factors influencing sport outcomes (Eglof & Gruhn, 1995; Piedmont et al., 1999; Rainey & Cunningham, 1988). As such, this study examined differences between early specializers and samplers on measures of personality and anxiety to determine whether significant differences exist between groups on these outcomes.

Descriptive statistics are provided in Table 1. Based on the criteria noted above, 43 of 77 (55.8%) participants were classified as early specializers. As expected, there was a significant difference between groups in terms of hours spent in other sports per week for both years t (78) =4.1, p<.001 and t (78) =5.1, p<.001. As the players were on the same teams or competing at similar skill levels, there were no significant differences in terms of hours per week spent in hockey. Having the groups be similar in terms of their hockey participation meant that we could be more confident that any differences found between groups in anxiety scores or personality would be a result of the breadth of sport participation (i.e., single versus multiple sport engagement).

Table 1
Descriptive Statistics of Study Variables by Early Specializers and Samplers

Note: All personality subscales have a possible score from 5 to 20; CSA and CTA anxiety subscales have a possible score of 5 to 20

Cronbach’s alphas showed good internal reliability for CTA and CSA (.89 and .82 respectively). Independent sample t-tests indicated there were no significant differences on any of the CSA subscales (somatic anxiety, t (73) = -.34, ns; cognitive anxiety, t (74) = -.01, ns; or self-confidence, t (75) = 1.51, ns). Similarly, for CTA, there were no significant differences on the subscales of somatic anxiety t (72) = -1.11, ns; worry, t (72) = -.89, ns, or concentration, t (72) = -.98, ns. Further, there were no significant differences on any of the personality sub-scales, indicating no differences between early specializers and samplers.

This exploratory study examined measures of state and trait competitive anxiety and the “Big 5” personality traits between early specializers and samplers. No significant differences were found for any of the outcomes. This was intriguing as researchers have noted several relationships between personality and different aspects of sport in the past. For instance, studies have found that athletes have higher levels of extraversion compared to the general population (Eglof & Gruhn, 1995; Kirkclady, 1982), as well as significantly lower scores of neuroticism and higher scores of conscientiousness (Kajtina, Tušak, Barić, & Burnik, 2004; McKelvie et al., 2003). Furthermore, this combination of personality traits (higher conscientiousness and lower neuroticism) has also distinguished higher performing and more dedicated athletes from lower level less dedicated athletes (Allen, Greenless, & Jones 2011; Piedmont et al., 1999). A possible explanation for our lack of significant effects may be the age of participants. Specifically, since there are mean-level age differences in the Big Five traits across the life span (Soto, John, Gosling, & Potter, 2011), extroversion, conscientiousness, and neuroticism may manifest themselves differently in relation to training, practice, effort, and investment among older athletes than younger athletes, making differences in sport trajectory based on these traits less prevalent. Further, past research has shown parents are the key influencers in their child’s sport participation (Greendorfer & Lewko, 1978). Given the young age of the participants, parents may have still been playing a primary role in selecting the child’s sport activities, making children’s personality a less influential factor in their sport trajectory.

As noted earlier, this exploratory design focused on examining psychological differences between athletes who specialize versus sample during their early sport careers. As a result of its exploratory nature, there were several limitations that should be considered in future work. First, the age range of participants in this study was exclusive (i.e., ages 11 and 12) and future research should examine older athletes who were early specializers to see if there are any differences in competitive anxiety and personality traits once personality becomes more stable. Potentially, athletes who specialized early and now compete at a higher level may have developed higher levels of neuroticism from the added pressure of investing all their time in one sport or learned to better manage their competitive anxiety with experience. Studying a wide age range would help to better understand differences between early specializers and samplers in terms of competitive anxiety and personality traits.

Second, a key challenge in this study surrounded the categorization of athletes as early specializers or samplers, which was based on parental report and understanding of the information required. Although this study used questions from the DHAQ (Hopwood et al., 2015), a well validated questionnaire, and our categorization was built upon a recognized definition of early specialization (LaPrade et al., 2016), future research may consider developing a more effective categorization tool – specifically ensuring instructions and questions are brief, simple, and clear, to enhance parents’ read ability and accurate completion. If athletes are not being categorized properly, this could lead to groups being too homogenous and would in turn result in researchers not being able to find differences between groups that might exist had they been accurately sorted.

It is possible previous research has overstated the psychosocial differences between early specializers and samplers and that there are more similarities than differences. Identifying areas of difference and similarity will be important as this field continues to evolve. As mentioned previously, the definition of early specialization and the subsequent classifying of athletes as early specializers is neither clear nor consistent across this area of research. This limitation has been noted by several researchers; for example, Ferguson and Stern (2014) stated .there is no standardized definition for early sport specialization, with authors providing their own interpretation…. The lack of a concrete definition has led to confusion over what qualifies as early sport specialization” (p. 377-378). Similarly, Buckley et al. (2017) noted “in 2017, the topic of single-sport specialization remains poorly defined, with many unanswered questions” (p.1). Some examples of the different definitions of early specialization include an “early starting age in highly structured and intensive activities with the explicit goal of improving performance in a sport” (Mendes et al., 2018, p. 2) while others have used “year-round intensive training in a single sport at the exclusion of other sports” (Post et al., 2017, p.1405) or even “the time when the athlete defined one sport as being more important than other sports” (Moseid, Myklebust, Fagerland, & Bahr, 2019, p. 461).The lack of a coherent definition means the markers for early specialization vary based on study design. While most researchers agree that early specializers pursue primarily one sport from a young age, the “cut off” age for single sport involvement varies depending on the sport and guiding sport participation model (e.g., LTAD, Canadian Sport for Life, 2019; DMSP, Côté, 1999; Côté & Fraser-Thomas, 2016). Further, while some suggest length of training is a key marker (e.g., increased frequency and increased duration of training; Wiersma, 2000), others focus on time spent in activities (e.g., year-round; Hill & Hansen, 1988). Questions remain as to what constitutes “year-round training” or “single sport participation”. In the current study, group categorization was operationalized using criteria from Baker and colleagues’ (2009) definition, but future research needs to establish a consistent definition of early specialization as well as the best method to accurately measure key markers. An important step forward will be to determine what is meant by “early” “intensive” or “year round”, for example.

This research is timely given the growing trend towards early specialization within youth sport programs, and thepotential consequences of early specialization (Fraser-Thomas, Côté, & Deakin, 2005; Law, Côté & Ericsson, 2007; Strachan, et al., 2009). While findings of this study did not find differences in competitive anxiety and personality according to sport trajectory, continued research in the area of sport trajectories is important for understanding the effect of different types of sport participation on youth development and athlete achievement.