Barriers to voluntary participation in sport for children: a systematic review

Literature searches were carried out in June 2015 using four electronic databases; EMBASE, Medline, CINAHL and SportDiscus using the search terms barrier*, stop*, prevent*, participat*, taking part, Sports/, sport*, “physical education”, PE, child*, young person*, adolescen*. An example search can be found in Appendix. Authors of the systematic reviews identified in the initial searches were also contacted to see if they were aware of any other relevant studies. Searches were updated in November 2016 with an additional search of Google Scholar. Search terms were simplified for Google Scholar to child*, barrier*, sport, and participation [28, 29]. A stopping rule was prospectively applied to the Google Scholar search results whereby screening of titles and abstracts was stopped after three consecutive pages where no new records were taken forward to full text screening.

This review is specifically focused on the barriers to children’s voluntary participation in sport. Only peer-reviewed records, describing original research and available in English, were included. Studies were required to discuss barriers to voluntary participation in sport in children up to the age of 18 years.

Studies were excluded if they only concerned the impact of non-participation in sport or the effects sports participation can have on variables such as the female athlete triad, smoking, or alcohol consumption. Studies where sport was included as an intervention (i.e. ‘forced’ participation) were excluded. Studies were also excluded if they only reported on participants with additional needs or were focused on injury from sports participation or pre exercise testing. Guidelines, policy documents, and other non-peer-reviewed publications were excluded. Studies were excluded if they did not investigate barriers to participation in sport or were not about participation in sport. Both qualitative and quantitative records were included.

A title screen was carried out by researcher 1 (SS) to remove any duplicates from the searches. Abstract screening was conducted independently by both authors. Any studies identified by either researcher as either providing likely or unclear evidence for inclusion were retrieved for full text review. The full texts were then independently reviewed against the inclusion/exclusion criteria by both authors. A consensus meeting was held to determine the extent of agreement and to resolve any disagreement, and agree the records to be included.

A data extraction form was developed, piloted on four records, and revised before data extraction began. Both author’s independently extracted data, and discrepancies were reviewed and resolved through discussion and revisiting the record. Extracted data included author, year of publication, country or location of study, study design, number of participants, age range of children in the study, type of barrier to participation, socioeconomic information, type of sport, and whether sport took place in or out of school.

To appraise quantitative studies we used the Clinical Appraisal Skills Programme (CASP) tool for cohort studies [30]. This tool contains 12 questions. Questions 3, 4, 6 and 12 were not used however as they are only relevant to intervention studies (Table 1). For each question there are three response ratings: ‘yes’, ‘no’, or ‘can’t tell’.

For qualitative studies we used the Clinical Appraisal Skills Programme (CASP) tool for qualitative studies [31]. This tool contains ten questions (Table 2). For each question there are three response ratings: ‘yes’, ‘no’, or ‘can’t tell’.

Study appraisal was conducted independently by the two authors and any disagreements in scoring were resolved through discussion and revisiting the record.

Study appraisal against the CASP questions is given in Tables 1 and 2.

In general studies met most criterion for quality; of 222 judgements 161 were that ‘yes’ that quality criterion is met, 41 were ‘no’ that the criterion was not met, and 20 were ‘CT’ meaning a judgement could not be conclusively made for that criterion/study. There was also observable trends across study types. Across quantitative studies for example, all were judged to have recruited in an acceptable way and almost all were judged to not have identified all important confounding factors. For qualitative studies, all were judged to have appropriately selected a qualitative approach to their research, whereas few reported or made clear that the relationship between researchers and participants had been adequately considered.

Support for the judgement presented in Tables 1 and 2 is given in Additional file 1 with some illustrative examples here.

In the study by Gordon [32] it was unclear if the study addressed a clearly focused issue and was rated as ‘can’t tell’. Two of the nine quantitative studies were judged to have identified all important confounding factors and took them into account in their design and analysis. Some caution should be attributed when using the results from Kirshnit [33], Irwin [34] and Hardy [35] as they scored ‘no’ on several quality questions.

In the qualitative studies it was unclear if the research design was appropriate for Barnett 2013 and Dismore [36]. It was also not clear if the recruitment strategies used by Azzarito [37] and Dismore [36] were appropriate. Differences in results may be explained by the differences in quality of the studies.

There is a potential bias in the quantitative studies as almost all were judged to not have identified all important confounding factors and as such analysis did not account for these. Four quantitative studies [32‐35] were scored ‘no’ on the majority of the quality appraisal questions. Those studies addressed a focused issue but did not take account of confounding factors in either the design or analysis of the results and the precision of the results was questioned. Some did not list odds ratios and confidence intervals making it difficult to assess the accuracy of their findings [32‐35].

Few qualitative studies reported or discussed the nature of the relationship between researchers and participants. It is therefore concerning that Barnett [38], Eime [39], Eimear-Enright [40], Fisette [41], Holt [42], Oliver [43], Stanley [44] and Totaro-Garcia [45] do not clearly provide this information. The majority (n = 10) of the qualitative studies scored as ‘Yes’ to sufficient rigour in data analysis indicating a well thought out and constructed process.

The estimates (barriers) reported in the two quantitative studies [30, 46], and one the qualitative study [41] which scored ‘yes’ on all CASP criteria provide high level evidence, i.e. replication of the study is unlikely to change the estimates. For qualitative studies we did not consider criterion 10 in this judgement as it is not relevant to reliability. All other studies were judged to have some factor or factors that might impact on the reliability of their estimates, i.e. further studies that do address these reliability issues may report different results.

Study characteristics of the nine quantitative studies are shown in Table 3 and for the 15 qualitative studies are shown in Table 4. Eight quantitative studies focused on the generic sports [32, 33, 46‐50] context with one focusing on swimming participation alone [34] in children aged 5 to 18 years. All quantitative studies made use of questionnaires and surveys. Boiche [46] investigated potential factors for dropout in sport or continuation of sport. Perry [49] sought to identify perceptions of motivators and barriers to physical activity, including sports participation. Irwin [34] also sought to identify barriers and facilitators to participation but focused on swimming. Casper [47] sought to identify constraints to participation in physical activity including sports participation and how these differ across age, gender, socioeconomic status and ethnicity. Dollman [48] focused on socioeconomic position and sport participation and how varying socioeconomic position influenced personal, social and environmental factors for participation in physical activity including sports. Gordon [32] focused on leisure activity involvement (including sports participation). All studies reported the relationship between gender and sports participation and the effect of increasing age. Two studies made use of predefined barriers for those participating in the study. One, Kirshnit [33], used electronic pagers to assist children with filling out a survey. Random messages were sent out to the pagers and on receipt children were asked to respond to a short survey. All quantitative studies made use of questionnaires and surveys. The most frequently reported barriers across the quantitative studies were ‘time’ (n = 4), ‘cost’ (n = 3), ‘opportunity/accessibility’ (n = 3) and ‘friends’ (n = 2). Where ‘friends’ was listed as a barrier studies had reported that children did not have friends to attend sessions with, or that they had no friends at the sport session and hence no one to partner with.

Across qualitative studies, one study focused on parental views of their child’s participation in sport [51] and another study looked at both parent and child perspectives of the benefits associated with participation in sport low income families [42]. Almost half of the qualitative studies focused on female experience of participation in sport only [37, 39‐41, 43, 52, 53]. Azzarito [37] sought to explore the views of ethnic-minority girls and their bodies and in what space they would be physical active in their daily lives. Eime [39] made use of the socioecological model to investigate factors affecting participation in sport and physical activity in rural girls whilst Fisette [41] explored the self-identified barriers to engagement in PE and enjoyment in girls. Enright [40] also investigated the PE environment but sought to identify how a negotiated PE curriculum might look. Kimm [52] focused on a biracial cohort of sedentary females and sought to identify barriers to their participation in activity. Oliver [43] sought to understand self-identified barriers to physical activity for 5th grade females whilst Wetton [53] sought to understand the barriers to participation to develop interventions to increase team sports participation in females. Armentrout [51] sought to understand organisational barriers and personal reasons for children discontinuing participation in sport. Barnett [38] sought to understand the perception of the relationship between movement, physical activity and sport for adolescents and whether views differed according to the amount of participation. Basterfield [54] investigated how perceived barriers to participation in sports both inside and outside of school changed within the same cohort over a three year time frame. Dismore [36] investigates children’s attitudes towards PE and school sport and the influences on these for children and Holt [42] sought to understand perceptions of benefits associated with participation in youth sport in children and parents from a low income background. Quarmby [55] investigated how different family structures affected children’s time activities including sports. Stanley [44] explored lunchtime activity and children’s perceptions of factors influencing their participation. Totaro-Garcia [45] investigated the physical activity characteristics of adolescents attending a PE service. The majority of the qualitative studies reported on sport in general but one study focused on ice hockey [51]. Six studies took place in the general sports setting (can include sport both inside and outside of school and may include organised sport) [38, 39, 41, 52, 54, 55], three took place within the school setting [43‐45], three took place outside of school [38, 42, 51], three focused on physical education within school [42,45,47] and two of these looked exclusively at females [37, 40]. The qualitative methods used in the study’s included interviews [36, 42, 45, 53], open survey questions [51, 52, 54], focus groups [38, 39, 44],survey and semi structured interview [55], focus group and interview [41], visual ethnography [37], participatory action research [40] and feminist action research [43]. Of the 15 qualitative studies six reported ‘time’ as a barrier to participation in sports, five reported ‘cost’, six reported ‘not being good at sport’, and six reported ‘fear of being judged/embarrassed’. Three studies reported ‘conforming to a gender stereotype’ three reported a ‘lack of resources’, and two studies reported ‘conformity’, and ‘boys dominating sports’ as barriers to participation.

The barriers identified in the quantitative studies were also identified in the qualitative studies. In contrast there were a number of barriers identified in the qualitative studies that were not identified in the quantitative studies. These additional barriers were coded as ‘politics’, ‘lack of enjoyment’, ‘self-conscious’, ‘lack of interest’, ‘competitive’, ‘conformity’, ‘bullying’, ‘lack of voice and choice’, ‘lack of motor skills’ and ‘doesn’t like strangers’, and related largely to emotion, feeling and lived experience.

Two visual representations were developed from the barriers identified across all studies included in the review. Figure 2 shows how practical barriers such as ‘transport’ and ‘location’ can link to other barriers such as ‘cost’. ‘Time’ was reported as a barrier in almost half of the studies in the review. This was either time in a child’s schedule or time for a parent to commit to taking their child to a sports session.

Figure 2 summarises the practical barriers to sports participation identified in this review. Time may be associated with time in a child’s schedule or time in a parent’s schedule. Children may wish to participate in sport but may be unable to because of a clash with their parents schedule such as working hours. The cost of participation can in some sports be fairly high and act as a barrier to some children. The cost of running a sports club may also act as a barrier to participation for children as there may be a lack of resource to facilitate provision of sporting opportunities. The type of equipment available can also influence a child’s view of participation, if it is old they are less likely to want to try a sport. Location can also influence participation, is the space suitable for the sport that is being held there or is it a compromise between cost and provision of opportunity. Transport can be linked to time, cost and location. If the location for sports participation is substantial distance from the child’s home then more time is required to travel to the venue which can incur a greater cost in transport. The most commonly reported practical barriers were used to develop the visual representation shown in Fig. 2.

Figure 3 summarises the personal barriers to sports participation identified in this review. Person centred barriers can be linked to internal and external factors which influence them. The barriers shown in this summary were taken from the qualitative studies. The experiences of children within the studies showed that a ‘bad experience in PE, ‘ peer disapproval’, ‘stereotype, ‘gender stereotype’, and ‘negative appraisal were influenced by external factors such as peers in a class or a teacher and were grouped under ‘external factors’ in this visual representation. Reports of ‘sporting ability’, ‘self-conscious’, ‘fear of judgement’ and ‘conformity’ within the qualitative studies were linked to a child or participants reaction to a situation in the sporting context and are listed as ‘internal factors’ in this visual representation. Competition was described in the qualitative studies in different situations, there were children who were competitive (a barrier to participation for others) and there were children who were put off by competition, therefore it represents both an internal and external factor.

The barriers cost and time were frequently reported barriers in the qualitative studies. They also appeared to associate with location, transport and resources (Fig. 3); if sports facilities were located further away it was more difficult to find time for some children to participate in them. This may be due to the child’s schedule or that of their parents.

“Well we looked at doing rock climbing out here [at the university]. [But] it’s a little bit trickier here because they have an indoor facility here but given that I [cannot drive due to medical reasons] and my wife’s in school so she’s not really there to drive us, getting out there and back on the bus would shoot three hours. Right? Like here it’s a good 40 minutes one way, then an hour lesson and then 40 minutes back.” – Holt et al. [42].

Traveling further can cost more in terms of fuel or public transport, for example a parent in the study by Armentout et al. [51] stated, “He didn’t want to quit playing, I told him we can’t afford it and we didn’t have the time”, when talking about ice hockey practice.

A parent in the study by Holt et al. [42] reported on the additive cost of travel, training costs, and costs of competitions as a barrier:

“Yeah. And each time she [daughter] does it, there’s a cost…its $5 for every race or is it even more. And I drive her there [another cost]. I don’t think that [the organizers] realize how sometimes those costs actually prevent my kids from joining running club... You don’t want your child to be labelled necessarily as the one that can’t afford it.” – Holt et al. [42].

Higher cost was associated with particular sports in some of the studies, for example ice hockey was described as expensive [42, 51].

“cost to play sport [ice hockey] unreasonable” – Armentrout et all [41].

“Like we’re not even talking hockey [an expensive sport].” – Holt et al. [42].

“It was an expensive program for the amount of practice and ice time available” -.

Armentrout et al. [51].

Single parents reported cost as a barrier to participation in sport for their child. This was also reported by parents in a partnership but was seen as a greater barrier by single parents.

“When [name of children] grow older the price changes sometimes. And the fees get more costly. And we went through a separation, me and my husband, and then you know sometimes it gets difficult. But I don’t want [my son] to know [my financial circumstances]. I would work extra hard for him to pay for his sports.”- Holt et al. [42].

Time was frequently reported as a barrier in different contexts. Armentrout et al. [51] reported a lot of feedback on amount of time for ice hockey practice.

“Early morning practices were tough”, “too much practicing for young kids”, “unreasonable practice times” and “too much time commitment”.

Other parents reported their job commitments as affecting participation for their children.

“… our son has to adjust to our schedule unfortunately because we can’t change it and I can’t help it. I have three jobs at the same time because we have to pay our bills and I have to support my family here.” – Holt et al. [42].

Time was a barrier that related to the status of the parent; whether they were in a partnership or were a single parent had an impact on their child’s sports participation. Children with a single working parent found their parents schedule a barrier to their participation in sport.

“My mum’s job [referring to barriers to activity], she gets back at like half five so if I want to go anywhere to do anything it’s normally too late so instead like I just play X box and stay in when she’s not back.” – Quarmby et al. [55].

The practical barriers (Fig. 2) may be more easily overcome than the personal barriers (Fig. 3). Practical barriers are potentially linked to more changeable situations such as time a child has to spend on the sport; this can change at various times during their school and college lives. Location can also be adaptable depending on the type of sport the child wishes to engage in.

Some of the personal barriers identified include ‘conformity’, ‘fear of judgement’, ‘disapproval’ and ‘gender stereotypes’. ‘Conforming to gender stereotypes’ was a barrier reported in three of the papers included in the review [41, 43, 53]. ‘Conforming to parent’s expectations’ and ‘media expectations in terms of gender stereotype’ were mentioned in two studies.

“Well yeah, it is for girls, but it’s just in their [my parents’] mind; it isn’t…I don’t think any girls would go out and play with a football, because it’s not like what it is supposed to be seen as, people playing football.” – Azzarito et al. [37].

“Sport is seen as a manly thing to do… they [media] don’t see it as a girly thing”.

– Wetton et al. [53].