Effects of different types of structured physical activity courses on gross motor development in preschool children: a 12-week comparative intervention study

Abstract

Background

Structured physical activity courses can effectively enhance preschool children’s motor skill development, and the type of activity may influence intervention outcomes. However, few studies have compared different courses, leaving educators without clear guidance for selecting effective interventions. This study examined the effects of three 12-week structured physical activity courses on the gross motor development of children aged 3–6, providing empirical evidence for intervention strategies.

Methods

A three-group quasi-experimental pretest–posttest design was adopted. Forty-five children aged 3–6 were recruited and assigned to one of three groups according to the course type: roller skating (RS), basketball (BT), and physical training (PT), with 15 children in each group. All participants received a 12-week intervention consisting of the assigned structured physical activity course, delivered twice weekly. Gross motor development was assessed before and after the intervention using the Test of Gross Motor Development-3 (TGMD-3). A 2 (time: pre vs. post intervention) × 3 (group: RS, BT, PT) repeated-measures ANOVA was used to examine within-group and between-group differences in TGMD-3 scores.

Results

After 12 weeks, the total gross motor score showed a significant time × group interaction. All three groups improved significantly (p < 0.01), and post-intervention scores in the BT and PT groups were higher than in the RS group (p < 0.01). The total locomotor score showed a significant main effect of time (p < 0.01), with improvements observed across all groups. Significant improvements were observed across locomotor skills, although gallop and skip improved only in the BT and PT groups. The total object control score also showed a significant time × group interaction, with improvements in all groups (p < 0.01) and higher post-intervention scores in the BT and PT groups compared with the RS group (p < 0.01). Among individual object control skills, two-hand strike of a stationary ball, two hand catch, and underhand throw showed significant interactions, with BT and PT outperforming RS after the intervention (p < 0.05). For the remaining skills, significant time main effects were detected across groups (p < 0.01). All groups improved in one hand stationary dribble, forehand strike of a self-bounced ball, and kicking a stationary ball (p < 0.01), while only the BT and PT groups improved in overhand throw (p < 0.01). A group main effect was also observed for the one-hand stationary dribble test (p < 0.01).

Conclusions

A 12-week engagement in roller skating, basketball, or physical training courses was associated with enhancements in locomotor skills, object control skills, and overall gross motor skill levels among 3–6-year-old children. Further comparison indicated that, although no significant group differences were observed in the overall gain in total locomotor scores, the basketball and physical training groups exhibited positive changes across all individual locomotor tasks. In contrast, while the roller skating group was associated with positive changes in most locomotor skills, less pronounced gains were observed in gallop and skip. Regarding object control skills, children in the basketball and physical training groups showed greater gains than those in the roller skating group in both overall object control performance and specific tasks such as two-hand strike of a stationary ball, two-hand catch, and underhand throw.

Introduction

Gross motor development in infancy and early childhood not only affects current physical activity capacity but also predicts motor skills and cognitive performance in later childhood and adulthood [1, 2]. Children’s gross motor level is related to cognitive abilities – children with lower gross motor skills tend to have poorer visuomotor integration and working memory [3, 4]. Fundamental movement skills (FMS) developed between ages 3 and 8 are crucial for future motor learning; FMS are often called the “building blocks” of lifelong physical activity for children and adolescents [5]. Solid fundamental skills benefit children throughout their lives [6]. Research indicates that children under age 8 have great potential to learn a variety of basic motor skills (e.g. walking, running, jumping, throwing) [7]. Therefore, a high level of gross motor ability in early childhood greatly promotes children’s spontaneous participation in physical activities and positively influences their cognition, psychological health, and social adaptability. This strong early motor foundation also supports the learning of specialized sports skills later on and helps instill lifelong exercise habits.

Exercise interventions can effectively promote young children’s motor skill development [8]. Preschoolers’ physical activities typically include structured courses, free play, and spontaneous games. While free play is important for healthy development, in China rapid urbanization has greatly limited outdoor activity space for preschoolers. Children often lack sufficient opportunities for physical activity in kindergartens or community areas. Many urban kindergartens have small campuses and large class sizes with low teacher-child ratios, constraining children’s exercise time. Moreover, many residential neighborhoods lack playgrounds specifically for young children. As public awareness of the benefits of physical activity and basic motor skills grows, more parents are enrolling children in various structured physical activity courses to compensate for the lack of exercise opportunities. Structured physical activity courses are goal-directed, organized activities led by teachers, designed according to principles of children’s physical education and developmental needs. Multiple studies have shown that structured, planned physical activity interventions can significantly improve fundamental motor skill development in preschool children [8–10]. In the context of limited play space, and with empirical support, structured courses are considered an effective way to promote motor development in Chinese children.

Currently, various structured physical activity courses for preschoolers have emerged, giving children the opportunity to participate in different types of courses. Because of differences in goals and content, different course types may have different effects on enhancing gross motor development. However, most existing research has focused on the impact of single courses on preschoolers’ gross motor development – for example, on basketball [11], soccer [12], or swimming [13]– and there are few comparative studies assessing differences among courses. This lack of comparison makes it difficult to understand the unique mechanisms and effects by which different courses promote motor development, and leaves educators without empirical evidence when selecting specific sports for intervention. It is thus challenging to create timely, developmentally targeted, and efficient exercise intervention plans for children.

Roller skating, basketball, and physical training courses are three widely offered structured physical activity programs for preschoolers in China. These courses are popular with children and each has distinctive features. The roller-skating and basketball courses both contain clear cultural elements and sport-specific movements, and are typical sports adapted for preschool education (roller skating representing individual sports, basketball representing team sports). The physical training course is a comprehensive fitness program that integrates multiple types of movements. It is important to emphasize that specialized sports training methods should not be applied in early childhood physical education. The basketball and roller skating courses examined in this study were not competitive, elite training programs, nor did they adopt adult-style instructional methods. Instead, they employed playful, game-based, and context-rich educational approaches, incorporated fundamental physical fitness exercises, and selected sport-specific movements that were basic and developmentally appropriate for preschool-aged children. By comparing the effects of these three courses on preschoolers’ gross motor development, this study aims to understand how different types of structured courses differently affect preschool children’s gross motor skills. The results will provide a scientific basis for parents, kindergartens, and training institutions to select and design preschool physical activity courses, thereby promoting motor development in young children in a targeted manner.

Methods

Trial design

This study employed a three-group quasi-experimental pretest–posttest design to compare the effects of different structured physical activity courses on children’s gross motor development. A total of 45 children aged 3–6 years were recruited through convenience sampling and assigned to one of three groups based on the course type they participated in: roller skating (RS), basketball (BT), and physical training (PT). Each group comprised 15 children. Before the intervention, all participants were assessed with the Test of Gross Motor Development, Third Edition (TGMD-3) to establish baseline gross motor scores. Each group then underwent a 12-week intervention consisting of the assigned type of structured course, with sessions held twice per week. Immediately after the 12-week intervention, the TGMD-3 was administered again. A 2 (time: pre- vs. post-intervention) × 3 (group: RS, BT, PT) repeated-measures ANOVA was conducted on each score to test the main effects and interaction of time and group. The study protocol was approved by the Ethics Committee of Beijing Sport University (Approval No. 2024192 H) and conformed to the Declaration of Helsinki.

Notably, this study adopted a preliminary exploratory quasi-experimental design, comparing three active intervention groups without a no-intervention control group. This design allowed for descriptive comparison of motor skill development across popular preschool activities, with the primary aim of generating hypotheses for future controlled studies.

Participants

An a priori power analysis was performed using G*Power (version 3.1) for a repeated-measures ANOVA with three groups and two time points. Assuming a medium effect size (f = 0.25), α = 0.05, and statistical power (1-β) = 0.80, the analysis indicated that at least 42 participants would be required.

Strict eligibility criteria were applied to reduce potential confounding and enhance comparability across groups. Children were included if they were 3–6 years old, physically healthy, enrolled in non-sports-specialized kindergartens where physical activity primarily consisted of unstructured free play, and had not received structured sports training prior to the study. Children were excluded if they had medical conditions contraindicating physical activity (e.g., chronic illness, musculoskeletal problems), sensory or cognitive impairments (e.g., vision, hearing, or learning disabilities), or were concurrently engaged in other structured extracurricular sports programs during the study period.

Preschool children who met these criteria were recruited through open enrollment at an integrated sports training institution in Shenzhen, China, with parents voluntarily registering their children after receiving detailed information about the study’s objectives and procedures. A total of 58 children expressed interest in participation. After screening against the eligibility criteria, 3 were excluded because they were concurrently engaged in other structured sports programs at different institutions. Thus, 55 children were formally enrolled in the study (17 in RS, 20 in BT, and 18 in PT). During the 12-week intervention, 10 children withdrew due to illness, family emergencies, or other personal reasons, leaving 45 participants (15 in each group) who completed the intervention and were included in the final analysis, meeting the required sample size to achieve adequate statistical power (0.80). The participant flow is summarized in Fig. 1.

Fig. 1.

Participant flow diagram

Prior to the intervention, the study objectives and the basic procedures of the training courses were explained in detail to parents and legal guardians. Informed consent was subsequently obtained, and permission was granted by the training institution. Based on course participation, children were assigned to the basketball (BT), roller skating (RS), or physical training (PT) groups. Each group trained twice per week, with at least one day between sessions to allow for adequate recovery. Baseline demographic and anthropometric characteristics are presented in Table 1.

Table 1.

Demographic characteristics of the participants in each group (M ± SD)

Group	n	Age (years)	Height (cm)	Body weight (kg)	Sex (boys/girls, n, %)
RS	15	4.5 ± 0.8	109.7 ± 7.4	18.6 ± 3.6	7 (46.7%)/8 (53.3%)
BT	15	4.8 ± 0.5	111.2 ± 5.4	19.7 ± 3.9	8 (53.3%)/7 (46.7%)
PT	15	4.3 ± 0.7	108.9 ± 5.3	18.6 ± 1.9	7 (46.7%)/8 (53.3%)

Intervention

Children in each group received a 12-week intervention of the assigned structured course, with 60-minute sessions held twice weekly. Courses were taught by a team of teachers qualified in early childhood physical education. Each class was limited to 4–8 children and conducted by one lead teacher assisted by one helper. To ensure consistency and quality across the different courses, all instructors received standardized training before the study. This training included: (1) a detailed explanation of the intervention plan and study objectives; (2) an overview of the course structure (warm-up, main activities, and cool-down); and (3) demonstrations of the specific teaching content and methods for the roller skating, basketball, and physical training courses, so that each teacher was familiar with the activities and instructional techniques for their course. During the intervention, researchers closely monitored and recorded each group’s progress, attendance, and practice to ensure adherence to the planned schedule.

Each of the three courses (roller skating, basketball, physical training) followed a three-part structure (warm-up, main activities, and cool-down) with a total duration of 60 min per session. The warm-up always included running-based exercises (such as light running and high-knee marches) to activate the body. In the main portion, each course included differentiated exercises based on its theme. The roller skating and basketball courses both combined various fitness exercises with sport-specific skill drills. For example, the roller skating course included drills like gliding, arm swings, and lateral stepping, whereas the basketball course included fingertip taps, dribbling, and footwork drills. Notably, the roller skating course emphasized lower-body practice (e.g. leg drills), while the basketball course included more full-body activities (e.g. dribbling while moving, shooting). The physical training course’s main activities consisted primarily of fun obstacle-course exercises: each session featured three circuit routes, each with 3–4 stations. The activities included crawling, jumping, throwing, kicking, hitting, dodging, balance-beam walking, S-shaped runs, and other movements to comprehensively promote motor development. In the cool-down phase, each course included static stretching and relaxation exercises to lower heart rate (see Table 2). The three courses were all conducted in a playful, engaging, and context-based manner, aligning well with the physical and psychological developmental needs of preschool children.

Table 2.

Content design of the different physical activity courses

Coursecomponents	RollerSkatingCourse	BasketballCourse	PhysicalTrainingCourse
Warm-up	Running warm-up, high knee raises, joint mobilization, etc.	Running warm-up, quick shuffle steps, high knee raises, backward kicks, etc.	Running warm-up, hopping/jumping, various crawling movements, etc.
Main Activities	Exercise 1: Fitness drills (e.g. high knee raises, backward kicks, static squats, single-leg arm swings; mainly to train lower-body strength, balance, and coordination). Exercise 2: Basic roller-skating skill practice (e.g. gliding, arm swings, lateral stepping) in a game-like context.	Exercise 1: Basic basketball skill practice in fun contexts (e.g. obstacle courses, mini-games) focusing on fingertip passing, dribbling, shooting, and footwork. Exercise 2: Fitness drills (e.g. quick reaction runs, standing long jumps, backward kicks, shuttle runs).	The course is mainly composed of fun circular obstacle courses: each class has three circuit routes, each with 3–4 stations. Activities include crawling, jumping, throwing, kicking, striking, dodging, balance-beam walking, S-shaped runs, etc., comprehensively promoting motor development.
Cool-down	Static stretching and relaxation.
Class Duration	Total 60 min.

Before finalizing the intervention curriculum, the research team conducted observations of existing courses and interviews with teachers at various training institutions to identify typical teaching content. The intervention content used in this study was then designed by integrating mainstream teaching models, ensuring that it was representative and enhancing the generalizability of the findings.

Measurements

Gross motor development was assessed using the Test of Gross Motor Development-3 (TGMD-3) before and after the intervention. The TGMD-3 [14] is designed for children aged 3–10 years and includes two subtests: locomotor skills and object control skills. Each subtest contains specific movement items (e.g. locomotor: run, hop, etc.; object control: kick a stationary ball, overhand throw, etc.). During testing, a trained examiner guided each child to perform each prescribed movement. Trained raters then scored the performance of each movement according to the TGMD-3 criteria. The scores for each movement were summed to produce a locomotor skill subscore and an object control skill subscore; the overall total score was the sum of these two subscores. Because the average age was similar across groups and the sex distribution was balanced, analyses were conducted using raw scores to enhance sensitivity in detecting pre- to post-intervention changes. To ensure scoring consistency, each child was evaluated by the same rater in the pre- and post-tests. Before data collection, all raters received training on scoring. After training, they re-scored a subset of children to calculate inter-rater reliability. The Kendall’s W coefficients all exceeded 0.70, indicating high reliability of the assessments in this study.

Statistical analysis

Data were entered using Excel 2019, analyzed with SPSS 26.0, and visualized using Prism 10.0. Data are presented as mean ± standard deviation (M ± SD). Preliminary checks indicated no missing values or outliers. Normality was assessed with Q–Q plots and homogeneity of variances with Levene’s and Brown–Forsythe tests. Baseline equivalence among groups was tested using one-way ANOVA on pre-test scores. To evaluate intervention effects, a 2 (time: pre vs. post) × 3 (group: RS, BT, PT) repeated-measures ANOVA was conducted for each test score. When a significant interaction effect was observed, simple effect analyses were performed using LSD-adjusted pairwise comparisons to examine both group differences at each time point and within-group changes across time. When no interaction effect was found, significant main effects were further explored: time effects were assessed with LSD-adjusted pairwise comparisons of pre- and post-test scores within groups, and group effects were examined with LSD-adjusted post hoc comparisons among groups. Effect sizes for ANOVA were reported as partial eta squared (η²ₚ). According to Cohen’s guidelines, values of 0.01, 0.06, and 0.14 represent small, medium, and large effects, respectively [15]. Statistical significance was set at α = 0.05 (two-tailed); p < 0.05 was considered significant and p < 0.01 highly significant.

Results

Data summary, baseline comparisons, and ANOVA results

Basic descriptive statistics of preschool children’s gross motor development test scores before and after intervention for each group are shown in Table 3; a comparison of baseline gross motor development scores is shown in Table 4; the results of the repeated-measures ANOVA for the main effects of time, group, and their interaction are shown in Table 5.

Table 3.

Basic descriptive statistics of preschool children’s gross motor development test scores before and after intervention for each group (mean ± SD)

Test Items	RS			BT			PT
	pre-test	post-test		pre-test	post-test		pre-test	post-test
Horizontal jump	3.47 ± 1.41	5.93 ± 1.58		2.87 ± 1.25	5.40 ± 1.18		3.20 ± 1.32	6.40 ± 1.99
Slide	5.33 ± 1.59	7.00 ± 1.60		5.07 ± 1.94	7.80 ± 0.56		5.00 ± 1.41	7.60 ± 0.83
Run	5.40 ± 1.99	7.87 ± 0.35		5.20 ± 2.24	7.53 ± 0.83		5.33 ± 1.05	7.87 ± 0.35
Hop	2.20 ± 1.37	3.60 ± 2.50		3.00 ± 1.25	4.20 ± 0.94		2.40 ± 1.24	4.53 ± 1.13
Gallop	3.13 ± 2.23	4.40 ± 1.55		2.13 ± 2.13	5.00 ± 1.65		3.33 ± 2.32	6.07 ± 1.16
Skip	2.00 ± 1.89	3.47 ± 2.33		2.13 ± 2.20	4.40 ± 1.55		1.73 ± 1.98	3.73 ± 1.44
Total scores of locomotor movements	21.53 ± 5.46	32.27 ± 6.90		20.40 ± 7.19	34.33 ± 3.31		21.00 ± 4.78	35.80 ± 4.66
Forehand strike of self-bounced ball	0.53 ± 0.74	2.47 ± 2.39		0.87 ± 0.92	3.53 ± 1.46		0.60 ± 0.91	2.80 ± 1.70
Kick a stationary ball	2.53 ± 0.99	4.00 ± 0.85		2.93 ± 1.22	4.60 ± 0.74		2.80 ± 1.08	3.80 ± 0.86
Overhand Throw	1.07 ± 0.96	2.27 ± 2.02		1.33 ± 1.05	3.27 ± 1.16		0.93 ± 0.80	3.07 ± 1.49
One hand stationary dribble	1.53 ± 1.92	2.93 ± 1.98		3.80 ± 0.68	5.87 ± 0.35		2.13 ± 1.25	4.93 ± 1.22
Two-hand strike of a stationary ball	3.00 ± 1.07	3.87 ± 1.51		2.87 ± 1.30	5.47 ± 2.03		2.87 ± 2.00	6.27 ± 1.71
Two hand catch	2.40 ± 1.84	3.87 ± 1.51		2.80 ± 0.94	5.47 ± 0.52		2.27 ± 1.28	4.87 ± 0.64
Underhand throw	3.00 ± 1.31	3.20 ± 1.93		2.27 ± 1.39	5.13 ± 1.36		3.13 ± 1.13	5.53 ± 1.60
Total scores of object control movements	14.07 ± 4.61	22.6 ± 7.08		16.87 ± 3.16	33.33 ± 3.62		14.73 ± 4.37	31.27 ± 4.37
Total scores	35.60 ± 7.76	54.87 ± 12.61		37.27 ± 8.92	67.67 ± 6.24		35.73 ± 7.87	67.47 ± 7.43

Table 4.

Comparison of baseline gross motor development scores across groups note: bold indicates statistical significance

Test Items	RS (mean ± SD)		BT (mean ± SD)		PT (mean ± SD)		baseline comparison
							F	p
Horizontal jump	3.47 ± 1.41		2.87 ± 1.25		3.20 ± 1.32		0.77	0.469
Slide	5.33 ± 1.59		5.07 ± 1.94		5.00 ± 1.41		0.17	0.845
Run	5.40 ± 1.99		5.20 ± 2.24		5.33 ± 1.05		0.05	0.955
Hop	2.20 ± 1.37		3.00 ± 1.25		2.40 ± 1.24		1.56	0.222
Gallop	3.13 ± 2.23		2.13 ± 2.13		3.33 ± 2.32		1.25	0.298
Skip	2.00 ± 1.89		2.13 ± 2.20		1.73 ± 1.98		0.15	0.86
Total scores of locomotor movements	21.53 ± 5.46		20.40 ± 7.19		21.00 ± 4.78		0.14	0.871
Forehand strike of self-bounced ball	0.53 ± 0.74		0.87 ± 0.92		0.60 ± 0.91		0.63	0.537
Kick a stationary ball	2.53 ± 0.99		2.93 ± 1.22		2.80 ± 1.08		0.51	0.603
Overhand Throw	1.07 ± 0.96		1.33 ± 1.05		0.93 ± 0.80		0.70	0.501
One hand stationary dribble	1.53 ± 1.92		3.80 ± 0.68		2.13 ± 1.25		10.88	0.000
Two-hand strike of a stationary ball	3.00 ± 1.07		2.87 ± 1.30		2.87 ± 2.00		0.04	0.962
Two hand catch	2.40 ± 1.84		2.80 ± 0.94		2.27 ± 1.28		0.59	0.561
Underhand throw	3.00 ± 1.31		2.27 ± 1.39		3.13 ± 1.13		2.00	0.148
Total scores of object control movements	14.07 ± 4.61		16.87 ± 3.16		14.73 ± 4.37		1.92	0.16
Total scores	35.60 ± 7.76		37.27 ± 8.92		35.73 ± 7.87		0.19	0.827

Note: Bold indicates statistical significance

Table 5.

Repeated-measures ANOVA results for the main effects and interaction

Test Items	Time				Group				Time× Group
	F	P	η²ₚ		F	P	η²ₚ		F	P	η²ₚ
Horizontal jump	94.22	0.000	0.692		1.49	0.236			0.69	0.507
Slide	60.58	0.000	0.591		0.28	0.761			1.25	0.296
Run	72.95	0.000	0.635		0.35	0.710			0.04	0.959
Hop	34.16	0.000	0.449		1.46	0.243			1.10	0.341
Gallop	34.67	0.000	0.452		2.95	0.064			1.74	0.188
Skip	34.06	0.000	0.448		0.57	0.570			0.52	0.601
Total scores of locomotor movements	179.11	0.000	0.810		0.44	0.645			1.58	0.217
Forehand strike of self-bounced ball	54.35	0.000	0.564		1.79	0.180			0.49	0.619
Kick a stationary ball	70.26	0.000	0.626		1.84	0.171			1.44	0.247
Overhand Throw	36.62	0.000	0.466		1.96	0.153			0.96	0.392
One hand stationary dribble	66.01	0.000	0.611		22.42	0.000	0.516		2.47	0.097
Two-hand strike of a stationary ball	61.53	0.000	0.594		2.86	0.068			6.57	0.003	0.238
Two hand catch	121.03	0.000	0.742		3.74	0.032	0.151		3.64	0.035	0.148
Underhand throw	42.48	0.000	0.503		4.41	0.018	0.174		8.65	0.001	0.292
Total scores of object control movements	323.10	0.000	0.885		11.67	0.000	0.357		11.89	0.000	0.361
Total scores	377.14	0.000	0.900		4.34	0.019	0.171		8.00	0.001	0.276

Note: Bold indicates statistical significance

Baseline data comparison

According to Table 4, among all baseline indicators, only the “one hand stationary dribble” exhibited a statistically significant difference between groups (p < 0.05). For all other indicators, the differences among the three groups were not statistically significant (p ≥ 0.05), indicating a generally similar baseline level of gross motor development among the three groups.

Total gross motor development score

After 12 weeks of intervention, there was a significant time × group interaction effect on the total gross motor development scores (F_2,42 = 8.00, p = 0.001, η²ₚ=0.276; Table 5). Simple effect analyses showed that the pattern of changes differed across groups. Within-group comparisons showed that the total score of each of the RS, BT, and PT groups significantly increased after the intervention (p < 0.01; see Fig. 2a). Between-group comparisons indicated no differences among the three groups before the intervention, but after the intervention the total scores of the BT and PT groups were higher than that of the RS group. (p < 0.01; see Fig. 2b).

Fig. 2.

Analysis of the interaction effect of total scores of gross motor development in preschool children (* indicates p < 0.05, ** indicates p < 0.01) (a) Comparison of pre- and post-test total scores within each group (RS, BT, PT). (b) Between-group comparison of total scores at pre-test and post-test

Locomotor movements

After 12 weeks of intervention, there was a significant main effect of time on the total locomotor movement score (F_1,42 = 179.11, p < 0.001,η²ₚ=0.810; Table 5). Follow-up analyses of the time effect indicated consistent improvements across groups. All three groups showed significant increases (see Fig. 3).

Fig. 3.

Analysis of the main effect of time on scores of locomotor movements in preschool children (* indicates p < 0.05, ** indicates p < 0.01)

Similarly, horizontal jump(F_1,42 = 94.22, p < 0.001,η²ₚ=0.692), slide(F_1,42 = 60.58, p < 0.001,η²ₚ=0.591), run(F_1,42 = 72.95, p < 0.001,η²ₚ=0.635), hop(F_1,42 = 34.16, p < 0.001,η²ₚ=0.449), gallop(F_1,42 = 34.67, p < 0.001,η²ₚ=0.452) and skip(F_1,42 = 34.06, p < 0.001,η²ₚ=0.448) exhibited a significant main effect of time (Table 5). Follow-up analyses of the time effect revealed significant improvements across multiple locomotor skills. All groups showed significant improvements in the test scores of horizontal jump, slide, and run (p < 0.01; see Fig. 4a, 4b and 4c). The hop scores increased significantly in the RS and PT groups (p < 0.01) and also in the BT group (p < 0.05; see Fig. 4d). The gallop and skip scores increased significantly in the BT and PT groups (p < 0.01; see Fig. 4e and f).

Fig. 4.

Analysis of the interaction effect of total scores of object control movements in preschool children (* indicates p < 0.05, ** indicates p < 0.01) (a) Horizontal jump: comparison of pre- and post-test scores. (b) Slide: comparison of pre- and post-test scores. (c) Run: comparison of pre- and post-test scores. (d) Hop: comparison of pre- and post-test scores. (e) Gallop: comparison of pre- and post-test scores. (f) Skip: comparison of pre- and post-test scores

Object control movements

After 12 weeks of intervention, the total object control score showed a significant time × group interaction (F_2,42 =11.89, p < 0.001, η²ₚ= 0.361; Table 5). Simple effect analyses showed that the pattern of changes differed across groups. Within-group analysis showed that the total object control score of all three groups improved significantly (p < 0.01; see Fig. 5a). Between-group analysis showed no significant differences before the intervention, but after the intervention the BT and PT groups had higher scores than the RS group (p < 0.01; see Fig. 5b).

Fig. 5.

Analysis of the interaction effect, main effect of time, and main effect of group on object control movements in preschool children (* indicates p < 0.05, ** indicates p < 0.01) (a) Comparison of pre- and post-test total scores within each group (RS, BT, PT). (b) Between-group comparison of total scores at pre-test and post-test

Following 12 weeks of intervention, the scores for two-hand strike of a stationary ball (F_2,42 = 6.57, p = 0.003, η²ₚ=0.238), two hand catch (F_2,42 = 3.64, p = 0.035, η²ₚ=0.148), and underhand throw (F_2,42 = 8.65, p = 0.001, η²ₚ=0.292) each showed a significant time × group interaction (Table 5). Simple effect analyses demonstrated differential patterns of change across groups. In the two-hand striking a stationary ball test, within-group analysis indicated significant improvements in both the BT and PT groups (p < 0.01; see Fig. 6a). Between-group analysis showed no differences before the intervention, but after the intervention the PT (p < 0.01) and BT (p < 0.05) groups had higher scores than the RS group (see Fig. 6b). In the two hand catch test, all three groups showed significant improvements (p < 0.01; see Fig. 6c), and after the intervention the BT and PT groups showed higher scores than the RS group (p < 0.01; see Fig. 6d). In the underhand throw test, only the BT and PT groups showed significant improvement (p < 0.01; see Fig. 6e), and after the intervention those groups had higher scores than the RS group (p < 0.01; see Fig. 6f).

Fig. 6.

Analysis of the interaction effect, main effect of time, and main effect of group on object control movements in preschool children ( * indicates p <0.05, ** indicates p <0.01) (a) Two-hand strike of a stationary ball: comparison of pre- and post-test scores. (b) Two-hand strike of a stationary ball: between-group comparison. (c) Two hand catch: comparison of pre- and post-test scores. (d) Two hand catch: between-group comparison. (e) Underhand throw: comparison of pre- and post-test scores. (f) Underhand throw: between-group comparison. (g) One-hand stationary dribble: comparison of pre- and post-test scores. (h) Forehand strike of a self-bounced ball: comparison of pre- and post-test scores. (i) Kick a stationary ball: comparison of pre- and post-test scores. (j) Overhand throw: comparison of pre- and post-test scores. (k) One hand stationary dribble: between-group comparison of the averaged pre- and post-test scores

The one hand stationary dribble (F_1,42 = 66.01, p < 0.001, η²ₚ=0.611), forehand strike of self-bounced ball (F_1,42 = 54.35, p < 0.001, η²ₚ= 0.564), kick a stationary ball (F_1,42 = 70.26, p < 0.001, η²ₚ = 0.626), and overhand throw (F_1,42 = 36.62, p < 0.001, η²ₚ = 0.466) tests all showed a significant main effect of time (Table 5). Follow-up analyses of the time effect indicated notable within-group improvements across these object control skills. On the one handstationary dribble, forehand strike of self-bounced ball and kick a stationary ball tests, all three groups showed significant within-group improvements (p < 0.01; see Fig. 6g, 6h and i). In the overhand throw test, only the BT and PT groups showed significant improvement (p < 0.01; see Fig. 6j).

Finally, the one hand stationary dribble test showed a significant main effect of group (F_2,42 = 22.42, p < 0.001, η²ₚ= 0.516; Table 5). Post hoc comparisons revealed significant differences among the groups. Specifically, the BT group scored significantly higher in overall pre- and post-test scores than both the RS and PT groups, and the PT group also scored higher in overall pre- and post-test scores than the RS group (p < 0.01; see Fig. 6k).

Discussion

Effects of different courses on preschoolers’ overall gross motor skills

After a 12-week engagement in the respective structured physical activity courses, all groups showed increases in total gross motor development scores, suggesting that participation in these courses was associated with improvements in preschoolers’ overall gross motor skill levels. This finding is consistent with previous research showing that organized exercise programs led by qualified coaches can significantly improve children’s fundamental movement skills [16], and that such intervention effects are sustained in preschool children [17]. A recent meta-analysis also found that structured physical activity interventions (whether active play or skill-oriented physical education) were more effective than unstructured play in improving children’s fundamental movement skills [18]. Structured physical activity courses are typically led by trained teachers and conducted with rich equipment and game-based scenarios in a goal-oriented and organized manner. Two main factors may explain why such programs are effective in enhancing gross motor development.

First, instructor-led organized physical activities encourage children to engage more frequently in running, jumping, throwing, and other movements. This provides them with abundant opportunities to practice, receive feedback, and refine each fundamental movement skill. Such repetition and feedback align with motor learning principles and facilitate neuromuscular adaptation and the consolidation of movement patterns. Although the course names and content differed in our study, each intervention included varied fundamental movement exercises in its basic practice. The period before age eight is considered a peak and sensitive period for motor development, often referred to as a “window of opportunity“ [19]. Payne et al. [6] emphasized that acquiring a new gross motor skill requires the opportunity to perform that action in a given context, and that transfer is likely when different skills share common components. Structured physical activity courses offer children repeated exposure and opportunities to practice a variety of movements, which may explain the observed improvements in their overall gross motor development.

Second, structured programs often incorporate fun and contextualized designs—such as small-group cooperation and interactive games—that boost children’s motivation to participate and increase their activity intensity. Studies have confirmed that structured physical activity courses are more effective than free play in promoting moderate-to-vigorous physical activity in preschool children [20]. In such courses, teachers ensure that all children are actively engaged in higher-intensity activity. Prior research has shown that higher levels of physical activity at preschool age positively predict the future development of fundamental motor skills [21] and that physical activity interventions benefit motor skill development [22]. One proposed mechanism is that increased intensity is associated with greater arousal, which can enhance attention and motivation and potentially improve motor learning efficiency. Structured physical activity courses allow children to sustain moderate-to-vigorous physical activity for longer periods, which may be another factor associated with the improvements observed in their gross motor skills.

The findings in the basketball and physical training courses are consistent with earlier studies. For example, Fotrousi et al. [23] found that a mini-basketball program significantly improved the gross motor skill levels of children in the intervention group. A systematic review also noted that basketball games can promote fundamental motor skills in children [11]. Similarly, structured physical training interventions have been shown to effectively improve children’s gross motor development [24, 25]. In contrast, most existing research on roller skating has focused on its effects on physical fitness. For example, inline skating has been found to improve balance and strength in children [26]. Few studies have examined the impact of roller skating on motor skill development. This study is the first to investigate the impact of roller skating courses on the gross motor development of preschool children. The results indicate that, similar to physical training and basketball courses, roller skating courses were also associated with improvements in children’s gross motor development. These findings not only reinforce existing evidence that organized physical activity courses are associated with improved motor development in preschool children but also provide additional empirical support by introducing roller skating as a viable intervention.

Differences in the effects of different courses on locomotor skills

Our findings show that although no significant group differences were observed in the overall gain in total locomotor scores, the basketball and physical training groups exhibited positive changes across all individual locomotor tasks. In contrast, while the roller skating group was associated with positive changes in most locomotor skills, less pronounced gains were observed in gallop and skip. Both the basketball and physical training courses included diverse locomotor activities. The physical training course, by design, is intended to facilitate overall development through a variety of movement tasks. Analysis of the intervention plan showed that it included running, long jumping, and single-leg hopping, with these skills practiced repeatedly in a circuit format. Therefore, children in the physical training group had many opportunities to practice locomotor movements. The basketball course, although sport-specific, naturally includes a range of basic movements. For example, its footwork training component involved forward, backward, and lateral running, as well as probing and crossover steps. In contrast, the roller skating course focused on a more limited set of movement patterns, including parallel gliding, single- and double-leg side gliding, turns, and curved skating. Its physical fitness drills included running, backward kicks, lateral steps, and standing long jumps. Content analysis revealed a lack of forward gliding and alternating two-foot jumping exercises in the roller skating curriculum, which likely explains the limited improvements in gallop and skip. These skills are technically demanding and require targeted practice. This finding further confirms that improvements in motor skills are not automatic but depend on practice opportunities [6]. This study utilized the TGMD assessment tool to evaluate preschool children’s locomotor skills. Results indicated that the basketball and physical training groups exhibited improvements across a broader range of individual locomotor skills compared to the roller skating group. However, it is important to acknowledge that the TGMD assessment tool does not comprehensively measure all aspects of children’s locomotor skills. Future studies should incorporate additional assessment tools, such as the BOTMP [27] or KTK [28], to more comprehensively evaluate the effects of various physical activity courses on preschool children’s motor development.

Differences in the effects of different courses on object control skills

Prior studies have shown that object control skills tend to benefit more from interventions than locomotor skills [16]. Compared with locomotor skills, the acquisition of object control skills not only requires the use of specific equipment but also relies more substantially on teacher assistance or instructional guidance, which introduces greater variability in the learning process. The attainment of proficiency in these skills is therefore more dependent on external stimulation and facilitation. In our study, the development of object control skills exhibited more pronounced differences. Compared to the roller skating course, participation in basketball and physical training courses was associated with greater improvements in overall object control scores and in specific skills such as two-hand strike of a stationary ball, two hand catch, and underhand throw.

Basketball involves complex hand-ball coordination, including dribbling, passing, shooting, and catching. Analysis of the basketball course in our study showed that it included various ball-handling drills and throwing tasks, which closely resemble the object control skills assessed by the TGMD-3: one hand stationary dribble, two hand catch, overhand throw, and underhand throw. While striking is not a routine part of basketball training, it shares mechanical similarities with throwing, particularly in terms of hand–eye coordination and force transmission. This may explain the significant improvements in striking tasks observed in the basketball group. Although the basketball course did not explicitly train kicking, participants in that group showed significant gains in kicking a stationary ball. This may be due to the transfer of learned control strategies such as initiating movement, generating force, aiming, and following through—skills practiced in throwing tasks and transferable to kicking. This further demonstrates that when motor skills share common factors or components, skill transfer is more likely to occur [6]. The physical training course also included a variety of object control tasks, such as throwing, striking, and kicking, with repeated practice provided through circuit training. This ensured broad exposure and supported improvements across all object control skills.

By contrast, the roller skating course offered limited upper-limb training, focusing mainly on lower-limb tasks. This may help explain why participants in the roller skating group showed significant improvements in lower-limb kicking movements (kick a stationary ball), but only limited gains in upper-limb throwing skills (overhand throw and underhand throw). Interestingly, the skating roller group still demonstrated significant gains in forehand strike of self-bounced ball, one hand stationary dribble, and two hand catch, likely due to course elements such as cone weaving and object-retrieval games, which may have enhanced target tracking and control. However, no significant improvement was observed in two-hand strike of a stationary ball. This may be due to the higher difficulty of this task and the relatively low weight assigned to accuracy in the TGMD-3 scoring, limiting the scope for score increases.

Overall, all three courses were associated with improvements in object control skills. However, the roller skating course—with its singular movement focus and limited opportunities for upper-limb practice—appeared to be associated with relatively smaller gains compared to the basketball and physical training courses.

Summary and educational implications

This study uniquely compared the effects of three types of structured physical activity courses within the same research design. While previous studies have typically examined only one activity in isolation, our study highlights how different course contents affect gross motor development both overall and in specific subdomains. Our analysis suggested that physical training and basketball courses, with their varied movement patterns, interactive ball exercises, and balanced upper- and lower-limb training, were associated with broader improvements in motor development. In contrast, the roller skating course, which focused primarily on lower-limb strength and balance with limited opportunities for upper-limb coordination, appeared to be associated with more modest improvements in object control skills. This comparative approach represents a key innovation of our study. It offers new perspectives for understanding how different types of structured activities influence motor development. For educators, coaches, and parents, such evidence-based insights have both theoretical and practical value. From an educational perspective, all three courses can support gross motor development in preschoolers. However, for more comprehensive development, programs that incorporate a wide range of movement types, include object manipulation, and emphasize balanced limb training—such as the basketball and physical training courses—may provide broader opportunities for motor skill enhancement.

Limitations and future directions

Despite careful design, this study has several limitations. First, the sample size was relatively small (15 children per group) and participants were recruited from a single integrated sports training institution, which may have introduced self-selection bias and limited the external validity of the findings. Although the inclusion criteria targeted children from non–sports-specialized kindergartens, the recruited sample likely reflected families more motivated toward extracurricular sports participation. Therefore, the results should be interpreted with caution and considered preliminary. Future research should recruit larger and more diverse samples across multiple settings to enhance representativeness and generalizability. Second, our study used convenience sampling and a quasi-experimental design with non-random assignment to intervention groups, which inevitably introduces potential selection bias and confounding. Because group allocation was based on participants’ course participation rather than randomization, there may have been pre-existing differences among children or their families that influenced the outcomes. Although baseline TGMD-3 scores were similar across groups, other unmeasured factors—such as parental support or individual motivation—could still have affected the results. These inherent differences make it difficult to attribute observed improvements solely to the type of activity. This limitation reflects a common challenge in applied preschool intervention research, where randomization is often impractical in real-world educational contexts. Nevertheless, future studies should adopt randomized controlled trials (RCTs), employ stratified sampling, and recruit participants across multiple sites to strengthen causal inference, minimize potential bias, and enhance the external validity of the findings. Third, the TGMD-3 Object Control Skills subtest mainly assesses upper-limb object control abilities (e.g., throwing, catching, dribbling, striking), which overlap with the practice content of the basketball and physical training courses. Therefore, part of the observed improvement may reflect task-specific learning rather than generalized motor development. Future studies should include more diverse assessment tools, such as the Bruininks–Oseretsky Test of Motor Proficiency (BOTMP) [27] and the Körperkoordinationstest für Kinder (KTK) [28], to more comprehensively evaluate preschool children’s motor competence. Finally, this study lacked a no-intervention control group due to ethical and practical constraints. While we compared three active intervention groups—roller skating, basketball, and physical training—the absence of an untreated control group makes it difficult to distinguish the effects of the training programs from natural maturation. Children aged 3–6 years typically experience rapid motor development over a 12-week period as part of normal growth. It is therefore likely that some portion of the improvement in motor skills observed in all groups would have occurred naturally, even without any structured program. In this sense, the study should be regarded as exploratory and quasi-experimental in nature, providing descriptive and hypothesis-generating evidence rather than definitive causal conclusions. Future studies should include an untreated control group and adopt randomized controlled designs to isolate intervention effects from age-related developmental gains with greater internal validity.

Conclusion and recommendations

Conclusion

A 12-week engagement in roller skating, basketball, or physical training courses was associated with enhancements in locomotor skills, object control skills, and overall gross motor skill levels among 3–6-year-old children. Further comparison indicated that, although no significant group differences were observed in the overall gain in total locomotor scores, the basketball and physical training groups exhibited positive changes across all individual locomotor tasks. In contrast, while the roller skating group was associated with positive changes in most locomotor skills, less pronounced gains were observed in gallop and skip. Regarding object control skills, children in the basketball and physical training groups showed greater gains than those in the roller-skating group in both overall object control performance and specific tasks such as two-hand strike of a stationary ball, two-hand catch, and underhand throw.

Education recommendations

Based on these findings, the following recommendations are proposed for designing and selecting preschool physical activity courses: First, encourage the development of a diverse range of movement skills. During the critical motor development period (ages 3–6), children should be encouraged to learn and experience a wide variety of movement skills. Whether in sport-specific courses (e.g. Roller skating or basketball) or comprehensive fitness courses, emphasis should be placed on varied practice content and coordinated development of upper and lower limbs. In particular, appropriately incorporating ball-handling or equipment-based activities can foster well-rounded development of both locomotor and object control skills. Second, avoid early specialization in a single type of training. When implementing sport-related physical activity courses for preschool children, it is essential to adapt them with playful and gamified approaches. The movement tasks practiced in these courses, including those derived from specific sports, should be diverse and fundamental in nature to align with preschoolers’ developmental needs. Overly monotonous activities pose a risk of imbalanced skill development and reduce children’s opportunities to acquire a broad range of motor skills.

Authors’ contributions

R. was responsible for developing the intervention plan, recruiting participants, collecting and analyzing data, and drafting the initial manuscript. Z. assisted with the design of the study, supervised the implementation of the intervention and the overall progress of the manuscript, and contributed to its review and editing. Q. participated in the study design and contributed to the review and editing of the manuscript. S. participated in the study design and contributed to the review and editing of the manuscript. All authors reviewed and approved the final manuscript.

Funding

This study did not receive any funding.

Data availability

Due to the unique characteristics of the study population, the data will not be publicly shared. However, researchers with reasonable requests can obtain the data from the corresponding author.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Beijing Sport University. Approval number is 2024192 H. All the participants in this study were preschool children, and the informed consent of the guardians of all participants has been obtained. The experiment complied with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.