Taylor & Francis Taylor & Francis Group http://taylorandfrancis.com

9 SPORT ACTIVITY IN CHILDHOOD

Early specialization and diversification

Paul R. Ford and A. Mark Williams

Sport activity in childhood takes place in formal and informal settings. Formal settings involve adults planning, organizing and leading activity for children, such as in coaching or physical education sessions and in competitions. Informal settings involve children leading the sport activities themselves without adult intervention, such as in playgrounds, parks, streets, gardens, car parks, homes, and beaches (Ford, 2016). In some sports in certain countries, children who are considered proficient or with potential are identified to engage in formal, competitive and systematic talent development programmes. In these programmes, children often specialize solely in the sport which has become known as early specialization (for a review, see Malina, 2010). Early specialization is defined as starting in the primary sport during childhood (5 to 12 years of age), starting in a talent development programme in childhood, and during that period engaging in one sport only, or at least primarily (Baker, Cobley, & Fraser-Thomas, 2009). Moreover, early specialization involves engaging in a relatively high volume and intensity of training in that sport, as well as in competition in that sport through tournaments, matches and/or leagues (Baker et al., 2009).

Many children engage in sport outside of these formal programmes, such as in informal activity in a single sport, perhaps with some engagement in other sports and/or some formal recreational activity sport, which is known as early engagement. Informal playful sports activities are self-directed and are engaged in with the intention of fun and enjoyment (Côté & Hay, 2002). This activity is hypothesized to lead to benefits such as increasing intrinsic motivation (cf. Hendry, Crocker, & Hodges, 2014; Imtiaz, Hancock, & Côté, 2016) and acquiring skills, such as decision-making (e.g., Roca, Williams, & Ford, 2012). Other children engage in sport mainly in formal, recreational settings. At the opposite end of the continuum from early specialization, is a pathway termed early diversification. Early diversification is defined as engage- ment in a number of different sports during childhood in mainly informal, but also formal settings, with late or delayed specialization into formal activity in a single primary sport in adolescence (Côté, Baker, & Abernethy, 2007). In adolescence, these pathways appear to converge to contain increasing specialization for talented athletes in a single sport through a high volume and intensity of training and competition. The predicted characteristics of the early specialization, engagement, and diversification pathways are shown in Table 9.1. The predicted and confirmed outcomes associated with and following early specialization, engage- ment, and diversification can be found in Table 9.2.

117

Researchers appear to have generally decided that, to paraphrase the renowned English author George Orwell in his book Animal Farm (1946), ‘early diversification good, early specializa- tion bad’. The latest incarnation of this idea was forwarded in an International Olympic Committee Consensus Statement on youth athletic development (Bergeron et al., 2016). It states that ‘youth should avoid early specialisation, as diverse athletic exposure and sport sampling enhance development and athletic capacity, reduce injury risk and increase the oppor- tunity for a child to discover the sport(s) that he/she will enjoy and possibly excel at’ (p. 845). Several other key review papers published in scientific journals have taken a similar position (Baker, 2003; Baker et al., 2009; Côté, Lior, & Hackfort, 2009; Côté & Hancock, 2016; Jayanthi, Pinkham, Dugas, Patrick, & LaBella, 2012; Mostafavifar, Best, & Myer, 2013; Wojtys, 2013). In contrast to this position, coaches and practitioners appear to have decided in Orwell’s terms that while ‘early diversification is good, early specialization is better’. In some countries a large

Paul R. Ford and A. Mark Williams

118

Table 9.1 The predicted characteristics of early specialization, engagement, and diversification

Early specialization Early engagement Early diversification

Childhood Start age in primary sport Childhood Childhood Not specified Start in ‘talent programme’ Yes No No Play in primary sport Low High Low to medium Practice in primary sport High Low to medium Low or no Competition in primary sport High Low or no Low or no Other sports Low Low to medium High

Adolescence Play in primary sport Decreases to zero Decreases to zero Decreases to zero Practice in primary sport High Increases to high Increases to high Competition in primary sport High Increases to high Increases to high Other sports Decreases to zero Decreases to low Decreases to low

Table 9.2 The predicted and confirmed outcomes associated with and following early specialization, engagement, and diversification

Early specialization Early engagement Early diversification

Positive outcomes in primary sport Performance improvement Yes Yes Possibly through

transfer Expert performance Yes Yes Yes

Other positive outcomes Continued participation Some Yes Yes Intrinsic motivation Lower Increased Increased Enhanced social skills Lower Possibly medium Increased

Negative outcomes Overuse injury incidence Higher Possibly medium Lower Burnout and dropout Higher Possibly medium Lower

industry has been built around youth development programmes that employ thousands of experienced adults to work with children (and adolescents) who specialize in a single sport in well-resourced and expensive facilities.

In this chapter, we present a critical review of research, theory and practice on sport activ- ity in childhood. We support the views of others that childhood sport activities should contain positive features and lead to several positive outcomes (for reviews, see Côté et al., 2009; Côté & Hancock, 2016; Fraser-Thomas, Côté, & Deakin, 2005; Vierimaa, Erickson, Côté, & Gilbert, 2012). However, in this chapter, we mainly limit assessment of the evidence to the develop- ment of skill or talent, as is the focus of this book. From this perspective, we seek to identify youth sport activities that cause the development of expert performance at the highest levels of professional sport without the occurrence of any negative consequences, such as overuse injuries or burnout. First, we review evidence for and against early specialization and diversifi- cation, which we focus on because these are the two main developmental activity pathways identified by researchers. To our knowledge, there is no single research study that contains children in controlled groups who either specialize or diversify, which measures between-group differences in variables and outcomes of interest at key time points from their current engage- ment in childhood into adulthood. Such longitudinal research is difficult to undertake for many reasons. Scientists have tended to use case-control and cross-sectional studies from which the quality of evidence is lower because of decreased internal validity when compared to cohort studies or randomized control trials (Petticrew & Roberts, 2006). Nevertheless, in the second part of the chapter, we forward evidence-based recommendations seeking to resolve research and practical issues in this area.

Assessing evidence for early specialization

The main rationale used to support early specialization is the positive relationship repeatedly found between amount of time spent practising a task or domain and level of competence. In this rationale, an earlier start age in a sport coupled with specialization enables a larger accu- mulation of time spent practising and, thus, higher levels of competence and achievement at a later date when compared to later start ages or not specializing. In addition, the term practice is usually used in this argument as a general term to describe engagement in formal sport-specific activities during childhood, including training or coaching sessions. It is beyond the scope of this chapter to review all of the research showing that time spent in practice or training leads to skill acquisition and improved competence, as well as improved physical fitness, but these are generally considered to be scientific laws within certain limits (e.g., Schmidt & Lee, 2011; Kenney, Wilmore, & Costill, 2015).

One of the most influential scientific theories on practice is often cited (e.g., Bullock et al., 2009) as support for early specialization. Deliberate practice theory was introduced by Ericsson, Krampe and Tesch-Römer in 1993. A central part of the theory is the monotonic benefits assump- tion. It holds ‘that the amount of time an individual engages in deliberate practice activities is monotonically related to that individual’s acquired performance level’ (Ericsson et al., 1993, p. 368). As such, they state that ‘individuals who start early and practice at higher levels will have a higher level of performance throughout development than those who start later’ (p. 392). There is no mention of the term ‘early specialization’ in the 1993 paper or in Ericsson’s subse- quent writings on the theory (1996; 2003; 2006; 2007; Ericsson & Towne, 2010). The theory covers all domains of achievement, such as musicians, surgeons or law enforcement, so the earlier start age it refers to does not specify childhood. They do state that ‘expert performance is not reached with less than 10 years of deliberate practice’ (Ericsson et al., 1993, p. 372). In

Early specialization and diversification

119

sport, of course, where the attainment of expert performance is most often required in late adolescence or sometimes earlier, such as in gymnastics (Law, Côté, & Ericsson, 2007), the 10- year rule infers a start age in childhood.

One of the main methods used to examine the question of early specialization or diversifi- cation was introduced by Ericsson et al. (1993). The retrospective recall method necessitates that expert athletes recall via questionnaires or interviews the number of hours they have spent in practice and other sport activities since they began in their sport. Multiple researchers have shown that hours accumulated by expert adult athletes in practice activities since starting in their sport are greater compared to control groups of lesser-skilled athletes (for recent reviews, see Baker & Young, 2014; Ford, Coughlan, Hodges, & Williams, 2015). In a few of these stud- ies, expert adult or late adolescent athletes reported that they experienced the main characteristics of early specialization in childhood (Ford et al., 2012; Law et al., 2007). Law et al. (2007) used this method to demonstrate that Olympic and international standard adult female rhythmic gymnasts started training in the sport at 6–8 years of age, competing at regional level at age 7–8 years, and started spending all their leisure time training in the sport from 11–12 years of age. During childhood, the Olympic gymnasts accumulated around 2,000 hours of sport-specific training and engaged in one other sport activity. It is likely gymnasts specialized early because expert performance in this sport is required in mid-adolescence, as opposed to in late adolescence or early adulthood in most other sports. Similarly, Ford et al. (2012) showed that late adolescent elite soccer players in Mexico and Sweden started training in the sport around 5 years of age, participated in competitions at age 8 years, and started at an elite academy at 12–13 years of age. During childhood, the elite players accumulated over 1,600 hours of sport-specific training and engaged in one to three other sport activities. However, there was large variation in training hours between players across this period (e.g., min = 224 hrs; max = 5,680 hrs) and both groups accumulated around 1,200 to 1,400 hours in informal, playful soccer activities during childhood, albeit with large variation in hours between players (e.g., SD = 800 hrs). Moreover, late adolescent elite players in England, Brazil, and Portugal demonstrated fewer of the characteristics of childhood specialization in the sport.

Retrospective recall studies have two methodological shortcomings that limit their findings in relation to the question at hand. First, these studies only describe the current youth devel- opment system in place in that sport and country and those who stayed in the system. It does not follow that the system is optimal and it may be that many other potential athletes dropped out of the system during their youth or were not selected. Second, retrospective recall studies only show the activities that preceded the attainment of expert status in adulthood or late adoles- cence. They do not show that the early activities caused the attainment of expert status. In the next section, we review retrospective recall studies that show how the childhood sports activi- ties that preceded the attainment of expert status in adulthood or late adolescence do not meet the criteria of early specialization. In addition, we review one of the strongest pieces of evidence against early specialization, namely the negative consequences, costs and outcomes of engaging in this pathway.

Assessing evidence against early specialization

Early specialization has been associated with a number of negative outcomes, costs or conse- quences for youths, such as increased incidence of overuse injuries, burnout, dropout, overtraining syndrome and decreased social development. Olympic rhythmic gymnasts reported more overuse injuries and lower health across their development when compared to lesser-skilled international gymnasts (Law et al., 2007). Both groups reported more injuries in

Paul R. Ford and A. Mark Williams

120

adolescence compared to childhood. In addition, Olympic gymnasts significantly increased training hours across adolescence, suggesting a positive relationship between training load and injury incidence in this period. Higher workloads and some other aspects of early specializa- tion (e.g., being elite level) have been associated with greater incidence of upper limb overuse injuries in youth gymnasts (DiFiori, Puffer, Mandelbaum, & Mar, 1996) and baseball pitchers (Lyman et al., 2001, 2002; Olsen et al., 2006; for a review, see DiFiori et al., 2014). Similarly, higher training loads and other aspects of early specialization (e.g., frequent intense competi- tion activity, sole-focus) has been associated with greater incidence of burnout, dropout, and overtraining in youth tennis (Gould, Tuffey, Udry, & Lowhr, 1996), gymnastics (Dubuc, Schinke, Eys, Battochio, & Zaichkowsky, 2010), swimming (Fraser-Thomas, Côté, & Deakin, 2008), and golf (Cohn, 1990). These examples of the negative consequences of early special- ization led several researchers to recommend early diversification as an alternative pathway (Baker, 2003; Baker et al., 2009; Bergeron et al., 2016; Côté et al., 2009; Côté & Hancock, 2016; Jayanthi et al., 2012; Mostafavifar et al., 2013; Wojtys, 2013).

Further evidence against early specialization can surprisingly be found in deliberate practice theory (Ericsson et al., 1993). The 1993 paper is often misquoted, hence the need for us to use direct quotes from it in this section. In the 1993 paper they outline ‘three phases of develop- ment toward adult expertise’ (p. 369). Of particular interest to this chapter is the first phase of development detailed in the paper. Ericsson et al. state they ‘rely on Bloom’s (1985) character- ization of this (first) period of preparation’ (p. 369). They further state that ‘from many interviews with international-level performers in several domains, Bloom (1985) found that these individuals start out as children by engaging in playful activities in the domain. After some period of playful and enjoyable experience they reveal ‘talent’ or promise. At this point, parents typically suggest the start of instruction by a teacher and limited amounts of deliberate practice’ (Ericsson et al. 1993, p. 369). They summarize this as follows: ‘the first phase of partic- ipation in a domain begins with an individual’s introduction to activities in the domain and ends with the start of instruction and deliberate practice’ (p. 369). The inclusion of this early, pre-deliberate practice stage is often overlooked in reviews of Ericsson et al. (1993). Elsewhere, Ericsson (2003; 2007) uses a version of the ‘power law of practice’ (Newell & Rosenbloom, 1981) to demonstrate how performance improvement is rapid in the earlier stages of experi- ence in a domain, whereas the rate begins to slow or plateau later in the process. Ericsson clearly states that individuals should begin to engage in deliberate practice once further improvements to performance in the domain start to slow or plateau, suggesting a pre-period of informal activities. We hypothesize that a positive relationship exists in certain domains between the complexity of the domain and the duration of this early period of playful activi- ties. For example, more complex fields, such as certain sport domains, will require longer durations of this first stage of playful activities before performance improvement begins to slow or plateau when compared to less complex domains. In the next section, we review several retrospective recall studies with professional athletes that show they did not engage in the early specialization pathway in childhood, but that they did engage in informal, playful activities during this period.

Assessing evidence for early diversification

Several researchers have advocated early diversification as the pathway that does not lead to the negative consequences of early specialization, but does precede the attainment of professional status in adulthood (Baker, 2003; Baker et al., 2009; Côté et al., 2009; Côté & Hancock, 2016; Fraser-Thomas et al., 2005; Vierimaa et al., 2012; Jayanthi et al., 2012; Mostafavifar et al., 2013;

Early specialization and diversification

121

Wojtys, 2013). The main evidence for early diversification as the activities that precede the attainment of professional status in adulthood is twofold. First, the negative consequences of early specialization support the idea that children should not engage in that pathway. Moreover, as shown in Table 9.2, early diversification is predicted to protect against overuse injures and dropout or burnout and to have other positive outcomes, such as improving social skills. Second, retrospective recall studies examining the developmental activities of athletes are predicted to show early diversification precedes the attainment of later expert status. In the following section we provide a critical review of these retrospective recall studies.

In order to ascertain the sport activities engaged in during childhood, it is necessary to have data on start age in the primary sport and other sports; start age in a talent development programme in the primary sport; the amount of play, practice and competition activity in the primary sport during childhood; and the amount of play, practice and competition activity in other sports during childhood. Table 9.3 (below) uses these criteria and summarizes some stud- ies that have been cited elsewhere as support for early diversification. The criteria for inclusion in Table 9.3 are that the participants in the study are adults playing at the highest level of their professional sport. As such, many papers examining adolescent athletes including those who participate in talent development programmes were excluded because a number of those participants will not achieve the highest professional level and, therefore, their activities create a confounding variable in these data sets that limits their findings (Bridge & Toms, 2013; Côté, 1999; Ford, Ward, Hodges, & Williams, 2009; Ford et al., 2012; Hayman, Borkoles, Taylor, Hemmings, & Polman, 2014; Henriksen, Stambulova, & Roessler, 2010; Leite & Sampaio, 2012; San & Li, 2014; Wall & Côté, 2007; Ward, Hodges, Starkes, & Williams, 2007). In addition, a few studies contain adult participants who did not play at the highest level of their professional sport (Hill, 1993; Ford & Williams, 2008, 2012; Roca et al., 2012; Surya, Bruner, MacDonald, & Côté, 2012) and that were excluded from Table 9.3 because their data sets are confounded by this fact. All of these studies have added to knowledge in this area, but for the purposes of this chapter the childhood activities of athletes at the highest level of their professional sport are considered.

We found that no studies provided full data sets on the four aspects of start age in the primary sport; start age in a talent development programme in the primary sport; the amount of play, practice and competition activity in the primary sport during childhood; and the amount of play, practice and competition activity in other sports during childhood. Some stud- ies with professional-level athletes that we wished to include in Table 9.3 had insufficient sets of this data (Coutinho, Mesquita, Fonseca, & De Martin-Silva, 2014; Haugaasen, Toering, & Jordet, 2014; Hodges, Kerr, Starkes, Weir, & Nananidou, 2004; Weissensteiner, Abernethy, Farrow, & Muller, 2008; Young & Salmela, 2010). Moreover, the level of detail required to complete Table 9.3 and answer the key question in this section appeared – on construction of the table – to be available only from studies on athletes from the highest level of single sports. Studies that present combined data from professional athletes across multiple sports were excluded from Table 9.3 because they did not contain sufficient amounts of the data we required to be included (Baker, Côté, & Abernethy, 2003a, 2003b; Barreiros & Fonseca, 2012; Hopwood, MacMahon, Farrow, & Baker, 2015; Leite, Baker, & Sampaio, 2009; Memmert, Baker, & Bertsch, 2010; Moesch, Elbe, Hauge, & Wikman, 2011; Moesch, Hauge, Wikman, & Elbe, 2013). These multi-sport studies are addressed in a separate section later in the chapter. Similarly, qualitative studies with professional athletes lacked the details we required for Table 9.3, so were excluded from it and are reviewed in their own section below (Carlson, 1988; Côté, 1999; Durand-Bush & Salmela, 2002; Johnson, Tenenbaum, Edmonds, & Castillo, 2008; Monsaas, 1985; Phillips, Davids, Renshaw, & Portus, 2010; Storm, Kristoffer, & Krog, 2012).

Paul R. Ford and A. Mark Williams

122

123

Ta bl

e 9.

3 T

he fi

nd in

gs o

f k ey

s tu

di es

e xa

m in

in g

th e

de ve

lo pm

en ta

l a ct

iv iti

es a

nd p

at hw

ay s

of p

ro fe

ss io

na l a

th le

te s

St ud

y B

er ry

e t a

l. (2

00 8)

B ak

er e

t a l.

(2 00

5) La

w e

t a l.

(2 00

7) So

be rla

k &

C ôt

é (2

00 3)

H or

ni g

et a

l. (2

01 6)

C ou

nt ry

A us

tr al

ia C

an ad

a C

an ad

a C

an ad

a Eu

ro pe

( G

er m

an y)

A th

let es

A us

tr al

ia n

R ul

es Tr

ia th

lo n

R hy

th m

ic g

ym na

sts Ic

e H

oc ke

y So

cc er

N um

be r o

f a th

let es

in w

or ld

Sm al

l Sm

al l

M ed

iu m

M ed

iu m

La rg

e

C hi

ld ho

od St

ar t a

ge in

p rim

ar y

sp or

t C

hi ld

ho od

A du

lth oo

d (tr

ia th

lo n)

C hi

ld ho

od C

hi ld

ho od

C hi

ld ho

od A

do le

sc en

ce (e

.g .,

ru nn

in g)

St ar

t a ge

s ele

cte d

to ‘t

al en

t p ro

gr am

m e’

N ot

a va

ila bl

e N

on e

N ot

a va

ila bl

e Ea

rly a

do le

sc en

ce Ea

rly a

do le

sc en

ce Pl

ay in

p rim

ar y

sp or

t Ve

ry lo

w N

on e

N o

H ig

h M

ed iu

m Pr

ac tic

e in

p rim

ar y

sp or

t Lo

w N

on e

H ig

h Lo

w Lo

w C

om pe

tit io

n in

p rim

ar y

sp or

t Lo

w N

on e

N ot

a va

ila bl

e Lo

w Lo

w O

th er

s po

rts Lo

w t

o m

ed i

Related Tags

Academic APA Assignment Business Capstone College Conclusion Course Day Discussion Double Spaced Essay English Finance General Graduate History Information Justify Literature Management Market Masters Math Minimum MLA Nursing Organizational Outline Pages Paper Presentation Questions Questionnaire Reference Response Response School Subject Slides Sources Student Support Times New Roman Title Topics Word Write Writing