Effects of cooperative learning in youth athletics’ motivational climate, peer relationships and self-concept

ABSTRACT

Background

Youth sport is threatened by an enormous dropout of prospective athletes. For this reason, we sought a training approach that would enhance young athletes’ learning in areas that should be critical for their retention in sport. According to current literature, a mastery motivational climate, good peer relationships and a good self-concept should have a positive impact on the length of an athlete’s career. The results of previous studies, albeit in physical education (PE), have shown the effectiveness of Cooperative Learning (CL) in achieving various affective, social, physical and cognitive learning goals. However, the impact of CL in competitive sport has not yet been investigated.

Purpose

The aim of this study was to investigate in youth athletics, (a) whether CL is more effective than previously used Direct Instruction (DI) in achieving specific goals in the social and affective domains, and (b) whether CL can promote the development of movement skills at least to the same extent as DI.

Methods

An experimental pretest–posttest comparison group design was used. Twelve Slovenian athletics trainers and 157 athletes (9.56 ± 0.73 years) were included in the sample. They were randomly divided into an experimental group (EG) that completed three consecutive CL units and a control group (CG) that continued with DI according to the control programme. The effects of the models were assessed using three standardised psychological questionnaires measuring peer relationships, motivational climate and self-concept. The movement skills of the young athletes were measured using four standardised fitness tests. One-Way ANCOVAs were conducted to examine whether the model (CL vs. DI) influenced posttest scores, taking into account pretest scores. We also captured all three elements of fidelity to CL.

Findings

There were significant posttest differences between the EG and the CG in peer relationships, emotional self-concept, mastery and performance motivational climate. The differences between the groups on the four physical variables were not significant.

Conclusion

The results confirmed that CL promotes better peer relationships, higher levels of mastery motivational climate, lower levels of performance climate and better emotional self-concept in young track and field athletes compared to DI. According to the current literature, improving these social and affective variables should reduce dropout rates in promising athletes. Therefore, we appeal to athletics federations to recognise the benefits of CL. We propose to inform trainers about the positive effects of CL and to include it in the education and training of trainers.

KEYWORDS:

• Cooperative learning
• Teaching approach
• Youth sport
• Motivation
• Relationships

Introduction

In our country, as in other European countries, youth sport is characterised by a high dropout rate among promising athletes: Slovenia – 85% (Železnik and Škof 2020), Finland – 75% (Enoksen 2011), Germany and the UK – 50-60% (Vaeyens et al. 2009), Belgium – 80% (De Martelaer et al. 2004), etc. The reasons for dropping out are too demanding trainings, early specialisation, injuries, burnout, poor relationships with trainers and among athletes, lack of motivation, etc. (Železnik and Škof 2020). Moreover, our country is unique in its small size, which means that already the pool of promising athletes is very small. Therefore, it is important to look for possible approaches to training young athletes that could help limit dropout.

Enabling the entire population to participate in sport within their abilities and interests, and increasing the number of athletes who consistently and successfully achieve excellence, are the main goals of the Long-Term Athlete Development (LTAD) model (Way et al. 2016). It is one of the modern models of talent development in sport that offer an alternative to early specialisation. There is no empirical evidence so far that one-sided, intensive, goal-oriented training of children in one sport would produce elite athletes. Moreover, the correlation between the sporting success of children under 10 and top sporting performance in the membership category is negative (Güllich and Emrich 2014). As talent identification models currently have relatively low predictive capabilities, youth competitive sport should be a place for all. The LTAD model promotes developmentally appropriate physical activity because the authors believe that individuals can only achieve improved physical literacy, athletic excellence and optimal health if their needs are addressed (Way et al. 2016). Based on their biological development, the LTAD recommendations include movement and sport skills that athletes need to develop at a certain age. In addition, the LTAD model emphasises that mental, cognitive and emotional factors are also essential for the development of every athlete. Ethics, fair play and character building should be present at all stages. 9-to 11-year-old children are in the Learn to train or major skill learning stage (Way et al. 2016). As children are most likely to learn fundamental movement skills in pre-adolescence, this should be encouraged in most training sessions. Pre-adolescence is also a sensitive period of competence development (Way et al. 2016). This means that 9-to 11-year-old children already recognise differences in their abilities, and if they see themselves as inferior, they may withdraw from participation in sport. Based on the child’s emotional development, the LTAD model recommends spending only 5-10% of the available time on competition. In fact, competition limits learning opportunities in some areas (Grineski 1996). Cooperation, on the other hand, allows children to develop different skills that cannot be achieved through the other goal structures (Hortigüela Alcalá et al. 2019). To avoid strenuous training, injuries and burnout, and to reduce dropout rates in sport, the LTAD model is an option for trainers. However, it lacks information on the approach to teaching and learning the listed skills. Although it takes a holistic approach, it puts more emphasis on physical goals than on the psychosocial aspects of children’s behaviour.

In youth sport, it is very important to focus on different learning goals – not only physical, but also affective, social and cognitive (Bailey et al. 2009). Cooperative Learning (CL) is a pedagogical model, a teaching and learning approach that enables physical education (PE) students to achieve goals in all four learning domains (Casey and Goodyear 2015). The impact of CL in competitive sport has not yet been investigated. Several studies in PE have confirmed the positive impact of CL on social learning (Bjørke and Mordal Moen 2020; Guzmán and Payá 2020; Hortigüela Alcalá et al. 2019). Although the effects of CL on the affective domain have rarely been investigated, some studies have reported improvements in PE students’ motivation and self-concept (Fernandez-Rio et al. 2017; Goodyear and Casey 2015; Goodyear, Casey, and Kirk 2014; Wang 2012). However, some researchers (Larsson and Nyberg 2017) question the effectiveness of CL in the physical domain, as it may allow for less active time than Direct Instruction (DI) – the approach still most commonly used in practise, where the teacher plays the main role and is the only person imparting knowledge (Guzmán and Payá 2020).

In CL, children work in heterogeneous pairs or small groups, cooperating and helping each other to achieve group goals (Cecchini Estrada et al. 2019). Learning in small groups allows them to develop their own knowledge while teaching peers (Johnson and Johnson 2009). A teacher primarily encourages children to learn from and with each other at their own pace. This type of learning takes into account children’s needs and interests, which is similar to the LTAD model (Casey 2017; Way et al. 2016). In addition, the cooperative nature of CL prevents the focus of physical activity from being too goal-oriented. A performance motivational climate has been found to reflect negative outcomes such as lack of fun, boredom, lower cohesion and even anxiety (Braithwaite, Spray, and Warburton 2011; Ommundsen et al. 2005). In contrast, a mastery climate is associated with better mental health (Sheehan, Herring, and Campbell 2018). Studies show that motivational climate influences the length of an athlete’s career (Moore and Weiller-Abels 2020; Weiss, Moehnke, and Kipp 2021).

The idea of integrating CL into youth sport is not about eliminating competition, but creating a cooperative space that would help young athletes improve their relationships with peers and trainers and develop various social skills (Jacobs, Teh, and Spencer 2017). Indeed, mastering the latter has a positive impact on success and satisfaction in sport (Jowett and Poczwardowski 2007). These two are among the most important motives for athletes to stay in sport. Therefore, if CL is used in youth sport, it could minimise the risk of early specialisation, physical and mental overload, and even burnout. For these reasons, CL could curb dropout in promising athletes and should therefore be tested in youth sport.

The aim of our study was to compare, in youth athletics, the effectiveness of two models – CL and DI (which is commonly used in practise) – in achieving learning goals in the physical, affective and social domains. It was hypothesised that: (a) there will be statistically significant differences between CL and DI in terms of their effects on young track and field athletes’ peer relationships, motivational climate and self-concept; (b) the differences between CL and DI in improving young track and field athletes’ movement skills will not be significant.

Materials and methods

Design

A pretest–posttest research design was used to investigate the effects of an intervention programme with CL and control programme (DI) on group and motivational climate, self-concept and movement skills of children in athletics. An experiment was conducted under non-manipulated (intact) conditions with naturally occurring groups.

Context

The Republic of Slovenia is a small country in central-southeast Europe with a population of 2.052 million. The official language is Slovenian. On the day 31.12.2021, 6534 young athletes born in 2009 were registered at the Olympic Committee of Slovenia – Association of Sports Federations (OCS-ASF 2022), of which 388 were registered with the Slovenian Athletic Federation.

Due to the socio-cultural specificities in Slovenia (e.g. low population and number of athletes, relatively low absolute resources for sport development) and the use of the Slovenian language, none of the well-known and psychometrically tested instruments developed manly in Anglo-Saxon cultural contexts could be used without adaptation. Therefore, there are only a few instruments that have been translated and validated into the Slovenian language so that they can be used in the Slovenian socio-cultural sport context.

Participants and procedure

In order to create the largest possible sample, we contacted athletics clubs from all over the country. We included all clubs that met certain criteria, the fulfilment of which made it possible to introduce the intervention in the first place: at least one group of 9-to 11-year-olds; at least twelve children training regularly; possibility of using indoor sports facilities in winter; training accessories; trainer with appropriate education (university degree, pedagogy) or qualification (at least first level and at least seven years of professional experience); trainer is willing to participate in the study; training takes place two or three times a week with the same trainer.

All twelve clubs selected are located in the urban areas of several cities in Slovenia (see Appendix A). By recruiting twelve trainers for the experiment, all their athletes aged 9–11 years were invited to participate in the study. The children were randomly divided into an experimental group (EG) (six clubs, 89 children) and a control group (CG) (six clubs, 68 children). 157 children participated in both measurements and in at least 70% of the training sessions. The EG with CL consisted of 32 girls and 57 boys (mean age: 9.26 ± 0.65 years). The CG with the traditional DI consisted of 39 girls and 29 boys (mean age: 9.89 ± 0.76 years). Further demographic data can be found in Appendix A. However, the procedure is shown in Table 1.

Table 1. Timeline.

February 2021	March-October 2021	May 2021	October 2021	November 2021 – first two weeks	November 2021-March 2022	April 2022
– approval by the University Ethics Committee – introductory meeting for trainers and parents	– coach training for CL	– trainers signed an informed consent – trainers received the entire intervention programme	– parents and children signed an informed consent	– pretest: fitness tests and psychological questionnaires (children)	– 30 consecutive training sessions with CL/DI (immediately following pretest) – recording randomly selected sessions – model fidelity	– posttest: fitness tests and psychological questionnaires (children; same tests and questionnaires as at pretest)

Model fidelity

In the next subsections, we will try to determine the fidelity of the intervention. Although each pedagogical model has its own idea and set of specific features, each model is flexible and allows practitioners to design units adapted to the specific circumstances of their context (Hastie and Casey 2014; Kirk 2013). Therefore, we ventured to use CL in the context of Slovenian youth competitive sport. To adequately understand the findings of our study, we needed to accurately describe the circumstances. We reported on all three elements of model fidelity that should be considered when researching educational approaches (Casey, Goodyear, and Dyson 2015; Hastie and Casey 2014).

A rich description of the curricular elements of the unit

The first author prepared a thorough intervention (and control) programme. The intervention programme began with Introduction to CL, as trainers and children needed to get used to their new roles (Casey, Goodyear, and Dyson 2015). The first unit started with cooperation games (icebreakers), which did not yet contain all the key elements of CL. The trainers added them gradually, in line with the programme (see Appendix B). In the second and third unit, the trainers had to form permanent, heterogeneous (by gender, abilities, knowledge, psychosocial characteristics, friendships) groups of four (±1).

The organisation of the learning process was based on cooperative structures that determined how the children worked together and what their learning goals were (Appendix B). Pairs-Check-Perform (Grineski 1996; based on Kagan (1992)) was introduced first because learning in pairs is much easier than working in larger groups. Jigsaw (Grineski 1996) was also frequently used for learning basic track and field skills. Each member of a jigsaw group was assigned to learn a segment in temporarily formed expert groups. After the children returned to their jigsaw groups, each child was asked to teach their segment to the other group members. Collective Score (Orlick 1982) was used primarily to develop movement skills (Kane and Kane 2004). Each group member contributed to the team with their score. Group goal was to improve the collective score or to be no worse than given. The main goal of STAD (Slavin 1995) was also to help others improve their score to make the most progress as a group but the grading system was more complicated. Groups received medals based on the percentage of improvement over their first attempt. PACER (Barrett 2005; Kane and Kane Jr 2004) was used to improve running technique. The pairs were given special flashcards with coordination exercises and had to take on the role of performer and observer. The trainers also assigned specific roles to the children (e.g. performer, trainer, timekeeper, referee, etc.) to work in Learning Teams (Johnson and Johnson 1994).

The cooperative structures promoted peer teaching and all five CL non-negotiables. Learning materials were provided to the children so that face-to-face promotive interaction was encouraged. Positive interdependence and individual accountability were promoted as each member of a jigsaw group received only one piece of information needed to complete a group task. Therefore, each child depended on the other group members. The children were assigned different roles so that each team member was responsible for a part of the group task. Since the children had to cooperate to achieve the group goals, positive interdependence was created (Grineski 1996). PACER also emphasised positive interdependence, by requiring all members to reach a certain level of competence in coordination exercises before the group (consisting of two pairs) could play a game. Individual accountability was also promoted by publicly presenting both the group's progress and individual results.

As part of the affective goals, the interpersonal and small group skills were defined separately for each training session (Appendix B). The trainers presented each skill to the children and they wrote it together on a special poster that accompanied them throughout the experiment. Group processing was done at the end of each session. It evolved from a whole group discussion led by the trainer to an independent debate in fixed groups.

In the CG, the trainers continued to use DI, i.e. they were the only ones who set tasks, determined the course and pace of learning, assessed the achievement of objectives and monitored the group's interactions (Metzler 2011). The work was organised frontally so that the children had the same tasks at the same time. The goal structure was either individual or competitive. The control programme corresponded to the intervention programme in terms of content. Regardless of the used model, the children worked on the same physical goals. However, there were differences in psycho-social learning, while DI does not allow for all types of goal attainment, as is typical for CL.

A detailed validation of model implementation

To determine model fidelity, i.e. whether reported learning outcomes can be attributed to the pedagogical model, we recorded four randomly selected training sessions from each athletics group in the EG (Polvi and Telama 2000; Zach, Cohen, and Arnon 2020). Data were collected through systematic event coding of the 17 categories of the Cooperative Learning Validation Tool (CLVT) (see Appendix C). It was developed by Dyson (2010) and tested and modified by Casey, Goodyear, and Dyson (2015). Observations were conducted by the first author. In the quantitative analysis of the CLVT data, the average percentages of each coded category were calculated. As part of the process of using the CLVT, qualitative data were collected in the form of Validation Tool Field Notes (VTFN) (Casey, Goodyear, and Dyson 2015; Goodyear 2017). In addition to the VTFNs, the Post Lesson Teacher Analysis Tool (PLTA) was used to report on children’s learning and the trainers’ actions from the trainers’ perspective (Bodsworth and Goodyear 2017). They were asked to write structured reflections after each training session.

The CLVT results showed that we achieved a satisfactory degree of CL model fidelity (Appendix C). All critical elements of CL were used in 75% of the sessions, but group processing was done in all sessions. Other key concepts of CL beyond the five non-negotiables (categories 2–6 in Appendix C) were also observed in about three-quarters of the recorded training sessions. We found that the percentage of observed CL key elements would be even higher if the structures and non-negotiables were not gradually added (Appendix B).

Student learning was assessed in each session and improvements were made in 92% of the sessions, reflecting high student engagement (Appendix C). The number of learning assessments and improvements observed was highest in the social or emotional domain. At the end of each training session children were asked to reflect in writing on the psychosocial aspects of their own behaviour, the behaviour of their peers and the behaviour of the trainer (VTFN). The CLVT revealed that social/emotional and cognitive goals were observed in 75% of the recorded sessions (Appendix C). However, physical goals were observed in every training session. Consistent with the CLVT findings, trainers reported improvements mainly in the social or emotional and physical domains (PLTA).

We cannot claim that full fidelity was achieved in every session, and there are certainly examples highlighted by CLVT, VTFN and PLTA that show variations in the degree of fidelity achieved. However, this moderate to high degree of model fidelity allows us to assume that children’s response to the units was the result of CL (Bjørke and Mordal Moen 2020; Casey, Goodyear, and Dyson 2015).

A detailed description of the programme context that includes the previous experiences of the trainer and children with the model

All trainers (except me – the first author) had used only the traditional approach (DI) before the study. I had discovered CL about a year before the study. It caught my attention so I first did a literature review and then started using it in practise as I work as an athletic trainer for children. The impact of CL on children’s learning is also the topic of my PhD thesis (in progress). I decided to train one of the groups in the EG to get a deep insight into the implementation of the intervention. I have also conducted a coach training for CL for the trainers in the EG. We met five times from March to October and conducted two lectures and three workshops, which lasted a total of 20 h (Table 1). In the first lecture, the trainers received explanations about goal structures, learning objectives, the history of CL, pedagogical models in general, basic features of CL and findings from previous studies on CL. In the second lecture, the non-negotiables and structures of CL were theoretically presented to the trainers. They were also given a first insight into the intervention programme. During the workshops, the trainers learned about general cooperative activities as well as specific cooperative structures with an athletics content. All workshops were recorded so the participants received video material. To check whether learning had taken place, they were given the task to choose a cooperative structure (from the programme) and a track and field discipline and present it practically in the third workshop. Other trainers had the role of performers (during the activities) and CL experts (after the presentations, during the group processing). During the experiment we were in constant contact with the trainers. We met regularly remotely and communicated by phone and email to solve various dilemmas, deepen the trainers’ knowledge of CL and adapt the plan according to the circumstances.

Details of the participating children can be found in the subchapter Participants and Procedure and in Appendix A. They had no previous experience with CL.

Data collection

As described above, quantitative and qualitative data were collected with CLVT, VTFN and PLTA to report on fidelity to CL. All other data collected was quantitative. Flamingo, flying sprint, T-test and curl-up were used to measure children’s movement skills or to report on the achievement of physical goals. Social and emotional improvements, on the other hand, were determined using three standardised psychological questionnaires to measure peer relationships, self-concept and motivational climate.

Movement skills

Standardised fitness tests were used to measure four movement skills, which were defined as the most important according to the children’s biological development (Way et al. 2016). The flamingo balance test (Sember, Grošelj, and Pajek 2020) requires standing on one leg on a special beam while the other leg is bent at the knee and the foot of this leg is held close to the buttocks. The 15-metre flying sprint measures speed (Mackala et al. 2019). The t-test (Pauloe et al. 2000) is a simple running test of agility that includes forward, sideways and backward movements and is suitable for a variety of sports. The curl-up abdominal test (Allen et al. 2014) requires you to perform as many curl-ups as possible with a cadence of one every three seconds.

Peer relationships

It was assessed using the My Class questionnaire (Fisher and Fraser 1981). Validation including factor analysis and internal consistency analysis for Slovenian conditions (Zabukovec 1998; N = 134, 9–10 years old) confirmed the psychometric properties of My Class. The questionnaire consists of 25 dichotomous questions (items) representing two factors – personal development (e.g. ‘We argue a lot in our athletics group.’) and peer relationships (e.g. ‘Every member of our athletics group is my friend.’). The Slovenian version of My Class was modified for the purposes of this study. The changes only concerned terminology (e.g. ‘class’ was replaced by ‘athletics group’). Reliability (α(personal development) = 0.70, α(peer relationships) = 0.74) was similar to the original and Slovenian (Zabukovec 1998) versions of My Class.

Motivational climate

The Motivational climate was measured with the Learning and Performance Orientations in PE Classes Questionnaire (LAPOPECQ; Papaioannou 1994). A pilot study of the Slovenian version (Cecić Erpič et al. 2004) confirmed that all items had satisfactory psychometric properties and confirmatory factor analysis showed that the items could be assigned to the same factors as in the original questionnaire. The Slovenian version consists of 25 items answered on a five-point Likert-type scale (1 = strongly disagree, 5 = strongly agree). The items represent five factors: (1) the teacher/trainer behaviour (e.g. ‘The trainer is very satisfied when each child learns something new.’); (2) the children’s learning orientation (e.g. ‘We – the children – are most satisfied when we learn a new motor skill.’); (3) the children’s competitive orientation (e.g. ‘We – the children – are most satisfied when we are better than others.’); (4) outcome orientation without effort (e.g. ‘It is very important to win without trying hard.’); (5) children’s worries about mistakes (e.g. ‘We – the children – feel bad when we make mistakes.’). The first two factors represent the variable ‘mastery motivational climate’. The second variable ‘performance motivational climate’ is composed of factors 3-5. Some terminological changes had to be made in order to use the questionnaire in the context of competitive sport (e.g. ‘teacher’ was replaced by ‘trainer’). The Cronbach’s alphas obtained were: F1 = 0.69, F2 = 0.74, F3 = 0.82, F4 = 0.74, F5 = 0.69. Reliability was similar to the original and Slovenian versions of the LAPOPECQ.

Self-concept

It was assessed using three subscales of the Self-Concept Questionnaire (SCQ). The original version is Autoconcepto Forma A (AFA; Musitu, García, and Gutiérrez 1994). It was validated for the Slovenian context by Kranjc in 1997 (Musitu et al. 1998). However, in 2018, Gosar rechecked the psychometric characteristics (N = 266, 9–11 years old) and confirmed the validity and reliability of the Slovenian version of the SCQ (Musitu et al. 2021). The three subscales measured were: social (e.g. ‘I like helping other children in the group.’), physical (e.g. ‘I am satisfied with my appearance.’) and emotional (e.g. ‘I get nervous when the trainer calls me.’). They consist of 27 items. The item response is a 3-point Likert-type scale that measures frequency. The Cronbach’s alphas obtained correspond to those of Musitu et al. (1998): social – 0.75, physical – 0.80, emotional – 0.73.

Data analysis

All data were analysed using IBM SPSS Statistics for Windows, version 26. Descriptive statistics were first compiled and pretest differences between groups were tested (see Table 2). Exploratory analyses were then conducted to determine whether the data met the assumptions for analysis of covariance (ANCOVA). To examine whether posttest scores differed between the models (CL vs. DI), One-Way ANCOVAs were used. The pretest scores were added as covariates. Finally, effect sizes were calculated and reported. For all statistical analyses, the significance level was set at p ≤ 0.05.

Table 2. Intragroup changes and between groups differences at pretest.

Variable	Pretest					Posttest
	Mean ± SD		t	p	Cohen’s d	Mean ± SD
	EG	CG				EG	CG
Personal development	25.88 ± 5.14	23.48 ± 4.95	3.00	0.00^a	0.48	24.85 ± 5.59	24.53 ± 5.80
Peer relationships	22.45 ± 3.86	24.06 ± 3.77	−2.51	0.01^a	−0.40	23.22 ± 4.26	22.72 ± 4.70
Social self-concept	28.74 ± 2.49	28.81 ± 2.90	−0.09	0.93	−0.01	28.82 ± 3.18	28.69 ± 3.37
Physical self-concept	18.4 ± 2.20	18.51 ± 2.00	−0.11	0.92	−0.02	18.53 ± 2.20	18.25 ± 2.28
Emotional self-concept	20.05 ± 2.77	20.67 ± 2.89	−1.18	0.24	−0.19	21.40 ± 3.13	20.71 ± 3.11
Mastery motivational climate	4.38 ± 0.43	4.41 ± 0.44	−0.20	0.85	−0.30	4.49 ± 0.44	4.35 ± 0.51
Performance motivational climate	2.56 ± 0.64	2.45 ± 0.65	0.98	0.33	0.16	2.22 ± 0.60	2.43 ± 0.78
Flamingo	15.99 ± 5.64	13.85 ± 5.55	2.31	0.02^a	0.38	15.65 ± 5.81	14.09 ± 6.54
15 m flying	2.59 ± 0.19	2.50 ± 0.15	3.12	0.00^a	0.51	2.57 ± 0.19	2.50 ± 0.16
T-test	14.10 ± 1.27	13.59 ± 1.13	2.48	0.01^a	0.41	13.48 ± 1.17	13.07 ± 1.15
Curl-up	18.14 ± 8.99	19.19 ± 10.03	−0.77	0.44	−0.13	24.44 ± 11.38	25.39 ± 11.27

Note. SD = standard deviation; EG = experimental group; CG = control group.

^a p < 0.05.

Results

Table 2 shows the descriptive statistics and between-group differences at pretest. Significant differences and medium effect sizes in favour of the CG were found for personal development, peer relationships, flamingo, 15 m flying and t-test. Non-significant differences and small effect sizes (which can be considered trivial) were found for other dependent variables.

To compare the effects of CL and DI on motivational climate, peer relationships, self-concept and movement skills, and to control for pretest scores on the dependent variables, we used ANCOVAs for further analysis. Post-test, there were significant differences in peer relationships, emotional self-concept and motivational climate between the EG and the CG (see Table 3). CL proved to be more effective than DI in improving peer relationships and emotional self-concept in young track and field athletes. The EG improved mastery motivational climate and decreased performance climate to a greater extent than the CG. At posttest, children in the CG actually perceived the mastery climate to be slightly lower than at pretest (Table 2). However, the effect sizes were small for all outcome variables (Table 3). The main results of the study were related to the psychosocial aspects of the young athletes’ behaviour. Nevertheless, the children’s movement skills were also measured and compared between groups to confirm that the learning process was also focused on physical goals, both in the CG and in the EG. This is because a learning approach that does not take into account the development of movement skills would be useless for youth competitive sport. The differences between the groups on the four variables representing children’s movement skills were not significant (Table 4).

Table 3. ANCOVA for posttest differences between EG and CG in group climate, self-concept and motivational climate.

Variable	Estimated marginal mean		ANCOVA		Partial eta squared
	EG	CG	F	p
Personal development	24.36	25.29	1.27	0.26	0.01
Peer relationships	23.59	22.15	5.72	0.02^a	0.04
Social self-concept	28.83	28.68	0.12	0.74	0.00
Physical self-concept	18.53	18.34	0.35	0.56	0.00
Emotional self-concept	21.54	20.53	5.51	0.02^a	0.04
Mastery motivational climate	4.49	4.35	5.01	0.03^a	0.03
Performance motivational climate	2.19	2.46	8.07	0.01^a	0.05

Note. EG = experimental group; CG = control group.

^a p < 0.05.

Table 4. ANCOVA for posttest differences between EG and CG in movement skills.

Variable	Estimated marginal mean		ANCOVA		Partial eta squared
	EG	CG	F	p
Flamingo	15.01	14.77	0.15	0.70	0.00
15 m flying	2.55	2.53	1.26	0.26	0.01
T-test	13.43	13.22	2.00	0.16	0.02
Curl-up	24.85	25.51	0.14	0.71	0.00

Note. EG = experimental group; CG = control group.

Discussion

The results confirmed a greater positive impact of CL compared to DI on children’s affective (self-concept and motivation) and social learning (peer relationships) (see Table 3), which is consistent with the findings of previous research in PE (Casey and Goodyear 2015). Peer relationships and mastery motivational climate improved only in the EG (Table 2) and the differences between the two groups at posttest were significant (Table 3). Bjørke and Mordal Moen (2020) also confirmed the positive effects of CL on peer relationships. Hortigüela Alcalá et al. (2019) found that implementing CL and teaching social skills to pre-adolescent children can help them become less eccentric. Social skills such as mutual trust, respect and support, understanding and cooperation are fundamental components of a good relationship (Hortigüela Alcalá et al. 2019; Jowett and Poczwardowski 2007). Although athletics is an individual sport, good relationships between participants are very important (Lisinskiene 2018). Developing good relationships with peers and trainers allows athletes to enjoy sporting activities, develop higher motivation, be more successful and continue to participate in sport (Jacobs, Teh, and Spencer 2017). An important contribution of our study to the current literature is also the finding that CL can improve the motivational climate in youth sport. Improving the mastery climate and lowering the performance climate can improve athletes’ mental health (Sheehan, Herring, and Campbell 2018) and increase the likelihood that they will stay in sport longer (Moore and Weiller-Abels 2020; Weiss, Moehnke, and Kipp 2021). Poor mental health (e. g. severe mood swings, depression, poor sleep quality, anxiety, etc.) and dropout can also be the result of poor self-concept in athletes (Lee 2020). Therefore, we have investigated and successfully demonstrated the greater positive effects of CL compared to DI on the emotional self-concept of young athletes.

Considering the philosophy of modern athlete development models in sport (including the theoretical background of the LTAD model), CL could be an appropriate training approach for children. As it can improve peer relationships, motivational climate and self-concept of young athletes, it could also help to improve the mental health of young athletes and reduce the dropout rate in youth sports. Therefore, it is very important for trainers to think about the type of teaching approach they use and how it affects the psychosocial aspects of children’s behaviour that are crucial for their retention in sport.

In PE and youth sport, psychosocial and cognitive goals are gaining importance, but the main effects of the learning process are still physical (Bailey et al. 2009). As this is a definition of quality physical activity for children, every teacher/trainer should ask themselves whether their choice of teaching approach enables as many children as possible to achieve as many learning goals as possible (Grineski 1996). With CL, all four learning goals of PE can be achieved (Casey and Goodyear 2015). Our study contributed to the science the finding that at least two (affective and social) learning domains in youth sport can be achieved with CL and that goal achievement in these two domains is more effective with CL compared to DI. The children in the EG also improved in the physical domain (Table 2), but the differences between the EG and the CG in movement skills were not statistically significant (Table 4). Nevertheless, the second hypothesis of our study was confirmed, although some researchers believe that CL allows less active time than DI and consequently lower improvements in movement and sport skills (Larsson and Nyberg 2017).

Conclusion

Given the huge dropout of young perspective athletes, we sought a pedagogical model that could improve children’s learning in areas that are crucial for young athletes to stay in sport. CL proved to be more effective than DI in improving the motivational climate, peer relationships and emotional self-concept of track and field athletes. Therefore, we propose to the Athletic federation of Slovenia and also to other national sports federations to establish a systematic coach training for CL so that (future) trainers get a deep insight into this pedagogical model in the context of youth sport. The goal is for trainers to use CL at least in combination with their current teaching approach.

However, we are aware of some limitations of the study. Considering the small number of young track and field athletes in Slovenia, we had to choose randomisation by natural groups. Due to the specific context, we also could not use the matched groups technique for randomisation. Another limitation could be the very different training conditions of the clubs (see Appendix A), which made the implementation of the training programmes difficult. One limitation was certainly the coronavirus pandemic and the associated closures. As a result, the average attendance of children was much lower, so we had to lower the standard for exclusion from the study (normally it is 80%). There were two limitations in determining model fidelity. If we were to study CL another time, we would record all training sessions to get a more realistic picture of model implementation. Also, we should involve another CL expert who would be able to assess model fidelity using the CLVT, but such experts do not exist in Slovenia.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Appendices

Appendix A

Table A1. Demographic information of the clubs, coaches and youth participants.

Club (group)	E (EG)	F (EG)	G (EG)	I (EG)	J (EG)	N (EG)	A (CG)	B (CG)	C (CG)	K (CG)	L (CG)	M (CG)
Facilities	Good	Good	Poor	Very good	Good	Very good	Good	Good	Good	Poor	Poor	Very good
Number of sessions per week	2	2	3	3	2	2	2	2	2	3	2	3
Length of a session [min.]	90	60	60	90	90	75	60	90	90	75	90	60
Coach’s gender	F	F	F	F	F	F	M	F	M	F	F	M
Coach’s age	39	44	29	49	36	27	23	27	30	24	36	28
Coach’s education/qualification	PE teacher^a	2. level	Sports coaching graduate^a	2. level	1. level	PE teacher^a	Student, Sports coaching; 1. level	Preschool teacher; 2. level	PE Student^a	PE Student^a	1. level	2. level
Coach – professional/amateur	A	P	A	A	A	A	A	A	A	A	A	A
Coach’s experience coaching athletics	13	18	7	13	10	9	1	7	5	4	15	7
Male athletes	9	4	5	9	1	4	4	4	7	9	1	4
Female athletes	11	3	9	7	16	14	7	3	8	4	6	8
Athletes’ age [Mean±SD]	9.45±0.61	9.14±0.38	9.21±0.58	9.31±0.60	9.47±0.64	9.29±0.77	9.91±0.94	10.43±0.54	9.21±0.58	9.62±0.65	10.14±0.38	10.42±0.52
Athletes – years in sport	1.5	0.7	1.2	2.0	1.4	2.1	1.4	2.1	1.3	1.7	2.0	1.9

Note. Group: EG = experimental group; CG = control group. Facilities: Poor – little gymnasium, track and field stadium (200-400m); Good – gymnasium, track and field stadium (300/400m); Very good – track and field hall and stadium. SD = standard deviation.

^a Educated personnel (PE teachers, sports coaches) does not need a qualification to coach.

Appendix B

Table B1. Description of CL units.

Unit	Session	Structures	Non-negotiables	Learning goals
1. – Introduction to CL	1.	/	Cooperative skills	Physical: - coordination, reaction speed Affective: - active cooperation of all - proper communication
1. – Introduction to CL	2.	/	Cooperative skills	Physical: - natural human movement - running technique - precision Affective: - active cooperation of all
1. – Introduction to CL	3.	/	Cooperative skills, group processing	Physical: - natural human movement - endurance, speed, precision Affective: - proper communication
1. – Introduction to CL	4.	/	Cooperative skills, group processing	Physical: - running technique - coordination Affective: - active cooperation of all
1. – Introduction to CL	5.	/	Cooperative skills, group processing	Physical: - speed, agility, precision Affective: - everyone included - cooperation – communication, help

Table

Unit	Session	Structures	Non-negotiables	Learning goals
1. – Introduction to CL	6.	Pairs-check-perform	Cooperative skills, group processing, face-to-face promotive interaction (worksheet)	Physical: - running technique - dynamic balance Affective: - mutual help - giving feedback
1. – Introduction to CL	7.	Collective Score	Cooperative skills, group processing, positive interdependence	Physical: - repetitive strength Affective: - listening to others carefully
1. – Introduction to CL	8.	/	Cooperative skills, group processing, face-to-face promotive interaction	Physical: - running technique - speed, agility Affective: - sharing ideas - giving feedback
1. – Introduction to CL	9.	Jigsaw	Cooperative skills, group processing, face-to-face promotive interaction, individual accountability	Physical: - endurance, coordination - precision Affective: - cooperation – communication, help

Table

Unit	Session	Structures	Non-negotiables	Learning goals
1. – Introduction to CL	10.	/	all five	Physical: - natural human movement - endurance Affective: - everyone included - proper communication
2. – Jumping and throwing	11.	Jigsaw, Collective score	all five	Physical: - place vortex throw - standing long jump - explosive strength Affective: - mutual respect and help
2. – Jumping and throwing	12.	STAD	all five	Physical: - running technique - place vortex throw - precision, strength Affective: - encouraging others - giving praise
2. – Jumping and throwing	13.	Jigsaw	all five	Physical: - long jump - explosive strength (legs) Affective: - focus on task - mutual help

Table

Unit	Session	Structures	Non-negotiables	Learning goals
2. – Jumping and throwing	14.	Jigsaw, Learning Teams	all five	Physical: - long jump - strength (empty weight, technique) Affective: - listening to others carefully - encouraging active involvement
2. – Jumping and throwing	15.	Collective Score	all five	Physical: - medicine ball throw - explosive strength (upper extremities), endurance Affective: - mutual trust - asking for help
2. – Jumping and throwing	16.	/	Cooperative skills, group processing	Physical: - long jump - explosive strength, endurance Affective: - sharing ideas - listening to others carefully - encouraging others
2. – Jumping and throwing	17.	Pairs-Check-Perform, Jigsaw	all five	Physical: - high jump (scissors) - running technique (scissors) Affective: - proper communication - mutual help, giving feedback

Table

Unit	Session	Structures	Non-negotiables	Learning goals
2. – Jumping and throwing	18.	STAD	all five	Physical: - medicine ball throw - explosive strength (upper extremities), speed Affective: - giving praise (when deserved)
2. – Jumping and throwing	19.	Pairs-Check-Perform	all five	Physical: - place vortex throw - relays (rules) - speed, agility Affective: - giving feedback - focus on task
2. – Jumping and throwing	20.	PACER	all five	Physical: - running technique - coordination Affective: - insistence, supporting others - giving feedback
3. – Sprinting and running	21.	Pairs-Check-Perform, Learning Teams	all five	Physical: - standing start - running technique - endurance Affective: - proper communication - encouraging others

Table

Unit	Session	Structures	Non-negotiables	Learning goals
3. – Sprinting and running	22.	Jigsaw	all five	Physical: - block start - speed (frequency, start acceleration) Affective: - sharing ideas - criticizing ideas, not individuals
3. – Sprinting and running	23.	/	Cooperative skills, group processing, face-to-face promotive interaction (worksheet)	Physical: - block start - explosive strength, stabilization Affective: - active cooperation - solving problems together
3. – Sprinting and running	24.	Collective Score	all five	Physical: - relays - speed endurance Affective: - insistence, encouraging others
3. – Sprinting and running	25.	PACER	all five	Physical: - running technique - coordination Affective: - insistence, supporting others - active cooperation

Table

Unit	Session	Structures	Non-negotiables	Learning goals
3. – Sprinting and running	26.	Learning Teams	all five	Physical: - block start - stride frequency, explosive strength (legs), precision Affective: - focus on task - giving feedback
3. – Sprinting and running	27.	STAD, Collective Score	all five	Physical: - block start - acceleration speed, stabilization Affective: - encouraging others - giving feedback and praise
3. – Sprinting and running	28.	Jigsaw	all five	Physical: - balance, coordination, agility, strength Affective: - proper communication - sharing ideas - criticizing ideas, not individuals
3. – Sprinting and running	29.	Learning Teams	all five	Physical: - place vortex throw, block start, long jump (refreshing) - explosive strength Affective: - independent individuals - active cooperation - mutual help and support

Table

Unit	Session	Structures	Non-negotiables	Learning goals
3. – Sprinting and running	30.	/	all five	Physical: - place vortex throw, block start, long jump (refreshing) - explosive strength, speed Affective: - mutual support - giving praise

Appendix C

Table C1. Results of the systematic event coding on CLVT.

Category number	Description of category	Percentage of sessions category coded as observed
1a	Social/emotional goals	75
1b	Physical/skill goals	100
1c	Cognitive goals	75
2	Equitable heterogeneous groups	92
3	Student centered instruction	75
4	Teacher facilitator	75
5	Cooperative learning structure	75
6	Students have shared ownership	67
7	Face-to-face promotive interaction	92
8	Positive interdependence	75
9	Small group and interpersonal skills	83
10	Individual accountability	75
11a	Physical assessment	75
11b	Cognitive assessment	67
11c	Social or emotional assessment	100
12a	Physical improvement	92
12b	Cognitive improvement	75
12c	Social or emotional improvement	92
13	Self, group or peer assessment	100
14	Students encouraging one another	42
15a	Group processing – what happened?	100
15b	Group processing – so what?	100
15c	Group processing – now what?	100
		Low	Moderate	High
16	High academically focused time	17	42	42
17	High level of student attention/interest/engagement	0	25	75