Associations between physical activity, basic motor competencies and automatic evaluations of exercise

ABSTRACT

Dual process theories suggest that the decision to be physically active is influenced by reflective and automatic processes. However, associations of automatic (affective) evaluations of exercise with physical activity and underlying basic motor competencies have not yet been investigated in children and young adolescents. Ninety-one participants (52 male; age: 10–14 years) were recruited from academic high schools in Germany and Switzerland. Automatic evaluations of exercise were measured with the Single-Target Implicit Association Test (ST-IAT) and a D-score was calculated. Moderate-to-vigorous physical activity (MVPA) and vigorous physical activity (VPA) per day were determined via wrist-worn actigraphy over the course of seven days. Basic motor competencies were measured using the MOBAK-5 test battery. Pearson correlations showed non-significant associations of automatic evaluations of exercise with MVPA, but significant associations with VPA. Basic motor competencies were associated with automatic evaluations of exercise, and the MOBAK subscale of object movement was associated with both MVPA and VPA. Our results underscore the relevance of affective processes for physical activity behaviour. This could potentially be relevant for interventions targeting physical activity promotion. Longitudinal investigations and intervention studies are necessary to verify causal relationships and potential underlying mechanisms.

KEYWORDS:

• Implicit attitudes
• implicit associations
• gross motor skills
• fundamental movement skills
• accelerometry

Introduction

Physical inactivity is one of the biggest health problems of the 21^st century (Blair, 2009; Kohl et al., 2012). To tackle this issue, numerous programmes and interventions to promote health through physical activity and exercise have been introduced (Ekkekakis & Zenko, 2016). However, effect sizes and long term success rates often remained low (Ekkekakis & Zenko, 2016), and so far, exercise interventions were not able to generally reduce the still rising rates of physical inactivity in Western societies (Ekkekakis, 2017).

To enhance the effectiveness of intervention programmes, it is important to understand how physical activity and exercise behaviour is regulated, and which psychological processes impede people from becoming and staying physically active. Most commonly, physical activity and exercise interventions are based on theories explaining physical activity and exercise behaviour through rational and reflective regulatory processes (e.g. plans, beliefs, expected value), such as the self-efficacy theory (Bandura, 1997) or the theory of planned behaviour (Ajzen, 1991). However, scholars have criticized that these theories might exaggerate people’s “capacity and willingness to make rational decisions in order to achieve desired goals” (Brand & Ekkekakis, 2018). Acknowledging the contribution of affects, feelings and emotions to decision making processes, dual process or dual system theories account for both type 1 (implicit/automatic, non-conscious, affective) processes and type 2 (explicit, conscious, reflective) processes (Evans & Stanovich, 2013) to explain behaviour. More specifically, type 1 processes are defined as autonomous processes (e.g. in decision making) that do not require working memory and are therefore non-conscious, fast, associative and often experience-based. Type 2 processes involve working memory, cognitive decoupling or mental simulation. They are characterized as conscious, slow and abstract, and involve consequential decision making (for a detailed overview, see Evans & Stanovich, 2013). When adapted to processes underlying the decision to be physically active or inactive, both automatic evaluations of exercise (type 1 processes, e.g. aversion due to past unpleasant experiences) and reflective reasoning (type 2 processes, e.g. conscious decision to be active because of known health benefits of regular exercise) have been claimed to play a role. Automatic evaluations of exercise relate to the non-conscious assignment of a positive or negative value to exercising (e.g. associating it with feelings of pleasantness or unpleasantness), which can be based on previous emotional or sensory experiences related to exercise (e.g. pride, embarrassment, or physical strain or pain) (Brand & Ekkekakis, 2018). These automatic evaluations can create strong approach or avoidance tendencies towards physical activity and exercising (Chen & Bargh, 1999). The most common theoretical approach is that those automatic evaluations yield default responses to exercise-related stimuli, unless they are overruled by reasoning processes (Brand & Ekkekakis, 2018; Evans & Stanovich, 2013). Therefore, researchers argue that physical activity and exercise interventions might be more successful in changing behaviour by shifting focus from reflective to automatic evaluations of exercise (Brand & Ekkekakis, 2018; Gerber et al., 2018; Rebar et al., 2016).

Because of their unconscious nature, measuring automatic evaluations of exercise (type 1 processes) is difficult. Compared to methods based on self-report, response latency methods such as the Single-Target Implicit Association Test (ST-IAT) have the advantage of avoiding potential bias through social desirability. The ST-IAT is a computerized test allowing the measurement of an individual’s valuation of a target concept (here: exercise) based on their response latencies when the target construct is presented in connection with positive or negative stimuli (Bluemke & Friese, 2008; Greenwald et al., 1998). A recent review by Rebar et al. (2016) on the effects of type 1 processes on physical activity showed that a) evidence gained through such response latency methods is still inconclusive, and b) most of the samples were university students and there is a general lack of data with regard to automatic evaluations of exercise and their association with actual physical activity behaviour in children and adolescents (cp. Gerber et al., 2019). During childhood, first experiences with physical activity and exercise are made which can shape automatic evaluations of exercise in the one or the other direction. These automatic evaluations, on the other hand, might influence future decisions to be physically active or inactive. Learning more about these mechanisms in young people is important as a basis for interventions that promote an active and healthy lifestyle in earlier life stages (Ortega et al., 2008).

Apart from automatic and explicit psychological processes, physical activity and exercise behaviour is determined by a variety of different factors at intrapersonal, social, environmental, cultural and policy levels (Condello et al., 2017; Craggs et al., 2011; Trost et al., 2001). At the intrapersonal level, it has been highlighted that physical preconditions (e.g. health state, motor competencies) constitute important factors. In order to be able to participate actively in the culture of sport and exercise, children and adolescents depend on certain basic motor competencies (Herrmann & Seelig, 2017). Other frequently and often interchangeably used terms are fundamental/gross movement/motor skills (for a review on the terminology see Logan et al., 2018). Without such basic functional performance dispositions, such as jumping, throwing, catching or balancing, children and adolescents are limited in their repertoire of practical possibilities for sport- and exercise related movements and have limited opportunities for successful engagement in physical activity later in life (Hulteen et al., 2018; Loprinzi et al., 2012). These competencies are not acquired naturally, but have to be learned and practiced during childhood (O’ Brien et al., 2016). According to dual process theories, previous experiences with exercising shape our non-conscious attitudes towards exercise. Consequently, children who are more successful at advancing their basic motor competencies might be more likely to show positive automatic evaluations of exercise. While most studies focused on associations between automatic evaluations of exercise and physical activity, potential associations with underlying basic motor competencies still remain unknown.

Given this background, in this cross-sectional, correlative analysis we investigated associations between automatic evaluations of exercise, objectively assessed physical activity and basic motor competencies in children and adolescents. Based on previous research and theoretical considerations, it is hypothesized that automatic evaluations of exercise are associated with both physical activity and basic motor competencies.

Materials and methods

Participants

The present study used cross-sectional data obtained from baseline assessments of two prospective studies that were conducted in Leipzig (Germany) and Basel (Switzerland), respectively. 91 participants (52 male; age: 10–14 years) were recruited class-wise; further details are provided in (Ludyga et al., 2018) and (Ludyga et al., 2019). The data from these two studies were combined to increase statistical power, and the observed statistical power of the main comparisons, as presented in the supplementary material (Supplement 1), shows that the calculations were indeed sufficiently powered. Because of neurophysiological measurements that were also performed within this study, only right-handed students, as controlled by the Edinburgh Handedness Inventory (Oldfield, 1971), were included. Exclusion criteria were regular drug intake and prevalence of chronic or acute diseases, which could possibly restrict physical activity during everyday life. Oral assent and written informed consent were provided by the children and their legal guardians, respectively. All procedures were in line with the Declaration of Helsinki and ethical approval was granted by the local ethics committees (EKNZ, 2016–01059; EK Leipzig, 246/16 ek).

Procedures

To ensure high ecological validity, all measurements took place within the school setting. All participants’ body height and weight was measured with a stadiometer and an electronic scale (Tanita BC-601, Tokyo, Japan), respectively, and an ST-IAT was administered. The ST-IAT was performed with one participant at a time and in a separate room, where noise was kept to a minimum. Physical activity was monitored over 7 consecutive days using accelerometers. At the end of the recording period, participants completed the MOBAK-5 test items in a group setting during a PE lesson. All assessments were completed within 10 days and during school-time.

Measures

Physical activity. Participants’ physical activity was measured using triaxial accelerometers (ActiGraph, Pensacola, FL, USA). The devices were worn on the wrist of the non-dominant (left) hand over the course of 7 consecutive days. Most publications on associations between basic motor competencies and physical activity relied on the assessment of moderate-to-vigorous physical activity (MVPA) (Holfelder & Schott, 2014). However, while MVPA might include non-exercise related physical activity above the threshold (e.g. bicycle commuting), vigorous physical activity (VPA) serves as a more accurate representation of exercise-related activities. For this reason, we calculated the amount of both parameters using an algorithm suitable for 12-year-old children distinguishing between sedentary, light, moderate and vigorous physical activity (Mattocks et al., 2007). Non-wear time was determined using the Troiano algorithm (Troiano et al., 2008). Measured data were considered valid if wear time was ≥70% per day and if at least three weekdays and one weekend day with sufficient wear time were detected (Barreira et al., 2015).

Basic motor competencies. The MOBAK-5 test battery was used to assess basic motor competencies. This battery includes a total of eight test items and is designed to measure the basic motor competencies self-movement (items: balancing, rolling, jumping, rope skipping) and object movement (items: throwing, catching, bouncing, dribbling) in fifth-grade children aged between eleven and twelve years. The test items consist of standardized tasks and evaluation criteria (see Herrmann & Seelig, 2017). The data was collected in classes during a regular 90-minute lesson. The classes were split up into small groups of three to four children each. The groups were guided and assessed by trained research assistants from the Universities of Basel and Leipzig, respectively. After a brief explanation and a one-off demonstration of the individual test items by the research assistant, the children were allowed two attempts (no trial run) to complete each of the test items. Each individual attempt was assessed on a dichotomous scale (0 = failed, 1 = successful). To obtain the final scores per item, we added up the number of successful attempts per test item (0 points = no successful attempts, 1 point = one successful attempt, 2 points = two successful attempts). The test items throwing and catching were an exception to this rule. In these cases, the children had six attempts per item, and the number of successful attempts was recorded. Afterwards, 0–2 successful attempts were scored as 0 points, 3–4 successful attempts as 1 point, and 5–6 successful attempts as 2 points. Accordingly, in both MOBAK-5 subscales (self-movement, object movement) maximum scores of 8 points can be reached, resulting in a combined MOBAK-5 sum score ranging from 0 (lowest basic motor competencies) to 16 points (highest basic motor competencies).

Automatic evaluations of exercise. Using E-Prime 2.0 (PST, USA), automatic evaluations of exercise were measured with a computer-based Single-Target Implicit Association Test (ST-IAT), comprising the target concept “exercise” and the evaluative categories “good” and “bad”. As visual stimuli representing exercise, neutral photographs showing young adults at different exercise scenarios (e.g. running, cycling) were presented. The evaluative categories were represented by emoticons (8 different smileys and frowneys). Pictures and emoticons have been obtained from the ST-IAT script employed in a previous study (Brand & Antoniewicz, 2016). Two test blocks of 32 trials were completed, consisting of target images and emoticons in a randomized order. They were preceded by 16 practice trials to reduce learning effects. The participants evaluated the emoticons by pressing a key corresponding to the categories “good” or “bad” with the left or right index finger, respectively. In one test block, the target images had to be assigned to the “bad”, in the other test block to the “good” category. The order was counterbalanced across participants and controlled for in the statistical analysis. The fixation period before stimulus presentation was 250 ms; participants were instructed to respond as quickly and accurately as possible as soon as the stimulus was present. Reaction time (response latency; time from stimulus presentation to pressing the response key) was measured on all trials. Trials with falsely assigned categories were repeated at the end of the test block until a correct response was given. Compared to repeating falsely assigned trials directly after the trial, this approach captures the natural latency unaffected by processing of feedback, and has been used in other studies before (Gerber et al., 2019, 2018). Throughout the experiment, participants were seated comfortably and surrounding noise was kept to a minimum. The ST-IAT has been found to be reliable and to exhibit discriminant validity (Bluemke & Friese, 2008).

The ST-IAT was analysed using a D-score, which was calculated by dividing the ST-IAT raw scores (difference in mean reaction time between the two blocks) by the within-individual standard deviation of response latencies calculated across the compatible and incompatible trials (Greenwald et al., 2003). D-scores from −2 to +2 are possible, with higher values representing more positive associations with exercise. As suggested by Blanton et al. (2015), values of 0.15, 0.35 and 0.64 were considered a slight, moderate or strong preference, respectively.

Statistical analyses

Eleven participants did not reach sufficient accelerometer wear time and were excluded from further MVPA and VPA analyses. Two children did not participate in the MOBAK-5 measurements due to illness on the day of testing. For the following correlational analyses, only the remaining valid data were used. Pearson correlations were calculated between the variables representing physical activity (MVPA and VPA), basic motor competencies (sum score, self- and object movement) and automatic evaluations of exercise (ST-IAT D-Score). In a second step, correlational analyses were repeated controlling for sex, age and body mass index using bootstrapping (number of samples: 5000), because these variables have often been shown to influence physical activity and basic motor competencies (Bolger et al., 2018; Herrmann & Seelig, 2017; Trost et al., 2001). Pearson coefficients of 0.1–0.29, 0.3–0.49 and ≥0.5 were considered small, medium or large, respectively (Cohen, 1988). Significance level was defined as <0.05. All calculations were performed with SPSS 26 (IBM Corporation, Armonk, NY, USA).

Results

Description of the sample

The main characteristics of the sample are depicted in Table 1. Participants’ mean D-score was 0.08, indicating neutral to very small positive automatic evaluations of exercise across the sample. Forty-four participants (48%) showed a small or stronger positive automatic evaluation of exercise, while 25 (27%) exhibited a small or stronger negative automatic evaluation of exercise.

Table 1. Sample characteristics

	N	Possible range	Mean	Standard Deviation
Age in years	91	10 to 14	11.16	1.25
BMI in kg/m^2	91		18.41	3.31
ST-IAT (D-score)	91	−2 to 2	0.08	0.30
MVPA in min/day	80		154.20	43.86
VPA in min/day	80		50.12	25.34
MOBAK overall score	89	0 to 16	6.60	2.93
MOBAK self-movement	89	0 to 8	3.62	1.63
MOBAK object movement	89	0 to 8	2.98	2.05

MOBAK = basic motor competencies. MVPA = moderate-to-vigorous physical activity. ST-IAT = Single-Target Implicit Association Test. VPA = vigorous physical activity. Differences in N for different variables are due to missing values in actigraphy (technical issues, exclusion because of insufficient wear time) and basic motor competencies (absence on testing day due to illness).

Bivariate correlations

Table 2 shows the correlations between the main outcomes of the study and the corresponding 95% confidence intervals. While significant medium correlations of the ST-IAT D-scores with the MOBAK sum score and significant small correlation with both MOBAK subscales were found, the ST-IAT D-score showed small non-significant correlations with MVPA and small significant correlations with VPA. Both MVPA and VPA showed a significant medium correlation with MOBAK “object-movement”, but only small non-significant correlations with MOBAK “self-movement”. Controlling for sex, age and body mass index did not change the general patterns of result, with one exception: the association between ST-IAT and MOBAK self-movement became statistically non-significant. However, the 95% confidence interval remained above zero [.00, .42].

Table 2. Correlations between implicit evaluations of exercise, physical activity and basic motor competencies before (below principal) and after controlling for sex, age and BMI (above principal)

	1.	2.	3.	4.	5.	6.
1. ST-IAT (D-score)		.21 [−.01;.43]	.28* [.07;.48]	.30** [.10;.49]	.22 [.00;.42]	.26* [.03;.48]
2. MVPA in min/day	.15 [−.09;.39]		.92*** [.87;.95]	.36** [.10;.58]	.16 [−.12;.40]	.40*** [.19;.57]
3. VPA in min/day	.23* [.01;.44]	.90*** [.85;.94]		.44*** [.20;.62]	.22 [−.05;.43]	.46*** [.36;.61]
4. MOBAK overall score	.32** [.14;.49]	.31** [.03;.53]	.40*** [.17;.58]		.74*** [.64;.83]	.84*** [.76;.90]
5. MOBAK self-movement	.24* [.07;.43]	.09 [−.18;.34]	.16 [−.08;.37]	.74*** [.65;.83]		.26* [.04;.47]
6. MOBAK object movement	.27* [.04;.46]	.37** [.15;.54]	.45*** [.26;.61]	.84*** [.77;.90]	.25* [.06;.47]
Sex (1 = male, 2 = female)	−.01 [−.24;.22]	−.07 [−.29;.16]	−.25* [−.45;-.04]	−.06 [−.28;.15]	.06 [−.17;.27]	−.14 [−.36;.08]
Age	.21 [−.01;.41]	−.22 [−.40;-.00]	−.10 [−.31;.13]	.20 [.01;.39]	.23* [.03;.42]	.11 [−.09;.31]
BMI	.02 [−.22;.23]	−.04 [−.23;.17]	−.02 [−.21;.21]	−.20 [−.39;.04]	−.11 [−.32;.15]	−.20 [−.39;.02]

BMI = Body mass index in kg/m²; MOBAK = basic motor competencies. MVPA = moderate-to-vigorous physical activity. ST-IAT = Single-Target Implicit Association Test. VPA = vigorous physical activity.

*p < .05. **p < .001. ***p < .001. Upper and lower 95% confidence intervals are displayed in square brackets

Discussion

In the present study, we examined whether automatic evaluations of exercise are associated with measures of regular physical activity and basic motor competencies in children and adolescents aged 10–14 years. Our results show that in our sample, a) basic motor competencies are associated with automatic evaluations of exercise, b) vigorous, but not moderate-to-vigorous physical activity is associated with automatic evaluations of exercise, and c) physical activity is associated with basic motor competencies in the domain of object movement, but not in the domain of self-movement.

Over the last decade, research regarding implicit processes in relation to exercise and physical activity increased rapidly (Chevance et al., 2019). In adult populations, results from prospective studies suggested that automatic evaluations of exercise were related to future objectively measured physical activity (Chevance et al., 2018; Conroy et al., 2010; Hyde et al., 2012). In a study by Rebar et al. (2015), not the traditional D-score, but information processing efficiency as calculated using the EZ-diffusion model, was correlated with objectively measured physical activity in a sample of 91 university students. Overall, recent meta-analytical findings indicated a small, but significant summary effect over 26 independent studies (Chevance et al., 2019), allowing the conclusion that in adults, intervention programmes targeting the promotion of physical activity might benefit from including implicit processes as well.

However, none of the studies that are currently available included children and young adolescents. Gaining knowledge on automatic evaluations of exercise in this age group is particularly important because a) surveys show that the majority of children and adolescents across the globe do not meet the WHO minimum recommendations of physical activity (Fan & Cao, 2017; National Physical Activity Plan Alliance, 2018; Finger et al., 2018; Sharara et al., 2018), and b) physical activity habits and health outcomes manifested in adulthood might be in part influenced by physical activity patterns during childhood (Cleland et al., 2012; Hayes et al., 2019; Loprinzi et al., 2012; Telama, 2009; Telama et al., 2014).

In our study with generally healthy children and adolescents aged 10–14 years, only small associations were observed between physical activity and automatic evaluations of exercise, and only associations with VPA, as opposed to MVPA, were statistically significant. Although the differences in the correlation coefficient were only minor (.23 versus .15), this might be a first indication that in this age group, this association might depend on the intensity of physical activity, meaning that children and young adolescents who engage more in age-typical vigorous physical activity develop more positive automatic evaluations of exercise. As visual stimuli in the ST-IAT, pictures of people exercising were used. Since activities of sufficient intensity to be classified as VPA are more likely to be exercise-related than activities that fall within the wider range of MVPA, this is a logical finding. However, according to a recent meta-analysis on studies with adults, the type of measured physical activity (e.g. MVPA versus light physical activity) was not a significant moderator of the relationship between physical activity and automatic associations of exercise (Chevance et al., 2019). Thus, this effect might be more pronounced in children compared to adults. Our finding of generally small associations between physical activity outcomes and automatic evaluations of exercise is in line with results which other researchers found for older age groups (Brand & Ekkekakis, 2018; Chevance et al., 2019). According to the Affective–Reflective Theory of physical inactivity and exercise (Brand & Ekkekakis, 2018), automatic evaluations of exercise constitute the first step in the process of the decision to become physically active or remain inactive. They are followed by affective valuations and reflective evaluations, which may have greater influence on exercise and physical activity behaviour. Other factors such as beliefs, traits or socio-cultural influences might be similarly important. Another explanation for small correlations might be that the association between physical activity and implicit evaluations of exercise is only indirect and mediated by other variables. Further studies with longitudinal design are necessary to investigate potential pathways (Schinkoeth & Antoniewicz, 2017).

With regard to overall basic motor competencies, we observed medium associations with automatic evaluations of exercise, and the subdomain of object movement was moderately associated with physical activity. Although our results represent cross-sectional data and do not allow direct conclusions with regard to causal relations, several explanations for the results seem possible. Basic motor competencies are usually considered to be a prerequisite to physical activity (Hulteen et al., 2018). This means, in order to participate in a minimum of regular physical activity, such basic skills are essential (Herrmann & Seelig, 2017; Loprinzi et al., 2012). Furthermore, research demonstrated the importance of basic motor competencies for physical, cognitive and social development (Lubans et al., 2010). Our data suggest, that predominantly those skills related to object movement (i.e. throwing, catching, bouncing and dribbling) might be necessary for children to actively participate in regular physical activity, while self-movement competencies such as balancing, rolling, rope skipping or moving variably were not associated with physical activity in our sample. This is supported by longitudinal analyses showing that object movement, but not locomotor skills during childhood are predictors of both physical fitness (Barnett et al., 2008) and time spent in moderate-to-vigorous and organized physical activity (Barnett et al., 2009) during adolescence. The authors speculated that perhaps object movement proficiency was associated with physical activity because these competencies are fundamental to involvement in various games and sports that depend on object control-related performance, which might be tendentially preferred by children at that age (Barnett et al., 2009). Furthermore, results of other studies also showed that coherence might be higher in children playing team sport comparing to individual sport (Eime et al., 2013; Elbe et al., 2017), which might contribute to a better manifestation of object movement compared to self-movement skills.

With regard to the observed association between basic motor competencies and automatic evaluations of exercise, bi-directional relationships are possible. On the one hand, children with more positive (implicit/automatic) attitudes towards exercise might be more prone to try different types of exercise and might be more motivated to develop corresponding motor competencies. On the other hand, higher motor competencies could elicit higher automatic evaluations of activities related to exercise because of generally more positive experiences in relation to exercise (Boyd & Yin, 1996; Field & Temple, 2017; Simpson et al., 2017). According to Rebar et al. (2016), a “potential technique for enhancing non-conscious regulation of physical activity is to strengthen automatic associations through evaluative conditioning” (p. 403), which can be achieved “by repeatedly pairing the presence of a stimulus with pleasant (or unpleasant) cues” (p. 403). High or low basic motor competencies might constitute such a pleasant (or unpleasant, respectively) exercise-related stimulus in everyday life, with higher competencies eliciting more positive implicit associations by repeatedly facilitating positive experiences in different exercise scenarios. Within the Affective-Reflective Theory of physical inactivity and exercise (Brand & Ekkekakis, 2018), in situations when people have to decide between exercising or sedentary behaviour, these positive associations might lead to a higher probability of a decision in favour of exercising.

Our study has certain strengths and limitations. One important strength is the measurement of both physical activity and basic motor competencies, showing that both variables differ in their association with automatic evaluations of exercise. Further strengths of our study are a thorough design and methodology including the use of response-latency-based rather than self-report methods to measure implicit processes, the objective measurement of physical activity over the course of seven days and the comprehensive assessment of basic motor competencies with the MOBAK-5 test battery. However, several limitations are important to notice. The interpretation of our results is limited by the cross-sectional nature of the study. Causal inferences need to be verified with longitudinal data. Furthermore, other potential determinants of physical activity behaviour such as explicit processes, self-efficacy, motivation, or socio-cultural influences were not measured, and thus not controlled for. In future studies, it would be interesting to examine whether the relationship between automatic evaluations of exercise and basic motor competencies persists even after adjusting for children’s and adolescents’ explicit, conscious attitudes towards exercise. Finally, our study includes data from three different schools in the neighbouring countries of Germany and Switzerland. However, since all included schools were academic high schools, all participants spoke German as their first language and no systematic differences between schools and countries were observed in any of the variables, it is unlikely that this caused a substantial bias.

Conclusions

Our results are novel and could potentially be relevant for interventions targeting the promotion of physical activity in children. Our results provide limited support for the notion that interventions to promote physical activity and exercise could be more successful if they considered strategies targeting non-conscious regulating processes. However, our findings suggest that incorporating basic motor competencies especially in the domain of object movement in interventions might be worthwhile. On the one hand, the improvement of basic motor competencies might benefit physical activity by extending children’s motor repertoire. On the other hand, improved basic motor competencies might elicit more positive automatic evaluations of exercise in children and adolescents. Nevertheless, such conclusions must be interpreted with caution, as our study provides only preliminary findings which are (a) based on cross-sectional data and (b) have not yet been replicated. Further longitudinal investigations and intervention studies are necessary to verify causal relationships and to test potential underlying mechanism that connect implicit processes, physical activity and basic motor competencies in children and adolescents.

Supplementary material

Supplemental data for this article can be accessed online https://doi.org/10.1080/02640414.2021.1907902

Disclosure statement

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