Introduction

Grounded in existentialism and phenomenology, physical literacy refers to an integrated state of being physically cultured to value and take responsibility to engage in lifelong physical activity1,2. Being physically literate is foundational to lifelong physical activity participation2 and can facilitate cognitive, mental, and physical health outcomes3. Developing physical literacy should start early in life for individuals to achieve adequate readiness for optimal health and wellness. However, most children across the world have been reported with inadequate physical literacy4,5, warranting purposeful educational and health promotion efforts from a young age. To develop physical literacy, it is important to involve parents at home as well as educators and sport coaches so children would become physically literate and active6,7. Home is one of the settings where children initially exhibit, acquire, and practice fundamental motor skills (FMS; e.g., walking, jumping, or sliding) and develop competence and confidence, before their later exposure to sophisticated physical activities such as sports and physical exercises2,8. While prior research has substantiated the importance of parenting for fostering physically active children9, however, little is known as to how parental perceptions relate to their children’s physical literacy, a crucial correlate of lifetime physical activity participation.

Physical literacy development should address cognitive (i.e. knowledge and understanding of physical activity and fitness), affective (i.e. confidence and motivation toward physical activity), physical (i.e. fundamental motor skill and physical fitness), and behavioral (i.e. daily physical activity) domains through relevant in- (e.g., physical education) and out-of-school opportunities (e.g., youth sports and recreational activities)2. Only through engaging in a broad variety of physical activities across settings can children develop the needed competence and confidence for sustained voluntary physical activity participation10,11. Specifically, at home, children look up to their parents as role-models to initiate active living behaviors and habits12. Parents’ functional social support such as encouragement13,14, companionship7,15,16,17 and instrumental support15,18 (e.g., sports equipment) are indispensable for children to participate in physical activity and develop physical literacy. Parents who hold positive perceptions toward physical activity (e.g., greater awareness, higher level of knowledge, more valuing, and positive beliefs of physical activity) will likely nurture their children to become physically literate and active19,20,21,22. In fact, parents’ valuing of and motivation for physical activity23,24 contribute to their own physical literacy development, making inter-generational impact on their children22 through improved parenting style25 and parenting practice19,26.

Parents’ perception of physical activity (PPPA) is a multi-dimensional construct that considers one’s value (i.e. perceived importance of the benefits associated with physical activity), appreciation (i.e. appreciation of aesthetics associated with athletics [e.g., sport competition and/or physique]), understanding (i.e. knowledge and understanding of physical activity and health promotion), attitude (i.e. attitude towards physical exercise and willingness to support), and awareness (i.e. recognition of physically active lifestyle) concerning physical activity27,28. Based on the framework of Epistemology, PPPA offers a thorough lens to analyze parents’ psychological perceptions when fostering children’s physical literacy. Our review of prior research has found that parental valuing9,29,30knowledge and awareness19social support13,31enjoyment and behavior of physical activity24,32,33and explicit and implicit attitudes34 are influential factors underlying children’s physical literacy and physical activity. These parental factors collectively amplify that the intergenerational transmission of parents’ physical activity-related promotors on their children’s physical literacy and physical activity behavior is considerable24. However, these relationships have only been partially substantiated by sporadic evidence. No prior research has investigated PPPA, as a multi-dimensions construct, and its predictive role in children’s physical literacy development. Furthermore, prior research indicated that children’s physical literacy varied by gender5,35,36,37, age5,36,37, body mass index5,38 (BMI), and socioeconomic status5 (SES). Additionally, previous studies have shown that parents’ gender14,15,16 moderated the association between parenting and children’s physical activity or physical literacy. Therefore, it is important to consider these factors when investigating the predictive relationship between PPPA and children’s physical literacy.

Given the current research gaps identified above, the purpose of this study was to ascertain the predictive effects of PPPA on children’ physical literacy and its four domains, after taking into account gender, age, BMI, SES, and gender of participating parent. As informed by prior research, we expected to observe significant predictive effects of PPPA (and its dimensions) on children’s physical literacy and its four domains (i.e. cognitive, affective, physical, and behavioral domains) after controlling for the sociodemographic and anthropometric factors.

Methods

Participants

A sample of 195 children (Girls: n = 91 or 46.7%; Mage = 9.09 ± 1.08) were recruited from a public primary school located in Central China. The recruitment was conducted during physical education classes where the purpose, procedure, and benefits/risks of participating in the program were described in detail, followed by the distribution of written consent and assent forms. At the time of data collection, the participating school enrolled 837 students (Han nationality = 95.1%) in 22 classes with a 13.95 student to teacher ratio. Official approval for all study protocols was granted by the human research ethics committee of Wuhan Sports University. All methods used in this study were carried out in accordance with the approved protocols. Informed consent and assent were obtained from all participants and their legal guardians with assistance of the physical education teachers. Participants were also informed of their rights to decline or withdraw from the study at any time without impact on their academic grades or school standing.

Variables and measures

Physical literacy

The simplified Chinese version of Canadian Assessment of Physical Literacy Version II4,39 (CAPL-2; χ2 = 70.16 [df = 43, p < 0.05], RMSEA = 0.04, CFI = 0.94, TLI = 0.90; Cronbach’s α = 0.82) was used to capture children’ physical literacy, which includes specific test batteries for each physical literacy domain (i.e., cognitive [10 points], affective [30 points], physical [30 points], and behavioral [30 points]). To complete the tests of cognitive and affective domains, children responded to a written survey with five questions assessing knowledge and understanding of physical activity and fitness, and 12 questions (5-points Likert scale) assessing competency, confidence, motivation, and predilection to physical activity participation. The first four knowledge questions were based on multiple-choice response (one point per question; four points in total), and the fifth question was based on fill the blank format (one point per blank space; six points in total), with one point scored per correct response. Tests of physical domain included the Progressive Aerobic Cardiovascular Endurance Run (PACER), Plank, and Canadian Agility and Movement Skill Assessment (CAMSA), with each test weighted 10 points. The behavioral domain was assessed both objectively and subjectively using pedometer (Yamax Digi Walker SW-200, Yamax Corporation, Tokyo, Japan) and self-reported instrument (weekly active days), respectively. To capture the average daily step count, participants wore the pedometer on their waist in waking hours for seven consecutive days and recorded daily steps using step log sheets. The self-reported instrument was incorporated in the survey along with the cognitive and affective questionnaire to collect number of days that self-perceived as physically active for at least 60 min per day across seven days (i.e., weekly active days). These raw data were subsequently converted using the standard-criteria approach. Scores of four domains were aggregated to arrive at a composite total score (100 points in total), to represent a child’s total physical literacy.

PPPA

PPPA was assessed using a validated questionnaire (Cronbach’s α across five dimensions: 0.80 ~ 0.86; CFI = 0.93, AGFI = 0.85, GFI = 0.90, RMSEA = 0.08) developed by Xu28. The questionnaire has 21 question items dimensional to attitude (five items), awareness (four items), value (four items), cognitive (four items), and appreciation (four items). A question example of the parental awareness dimension is phrased as, “Doing sports and/or exercising is an indispensable part of my life.” Parent participants (either father or mother) are then asked to circle a response on a Likert scale from five (strongly agree) to one (strongly disagree). Responses were summed up to obtain aggregate scores for each dimension (four-five points to each dimension) and the total PPPA score (105 points in total).

Demographic and anthropometric factors

Informed by prior research5,14,35,36, three demographic factors (i.e. gender, age, and SES), one anthropometric variable (i.e. BMI), and one ID variable (gender of participating parent) were regarded as covariates in data analyses. Participants self-reported their demographic information on a written survey, including gender, age, SES, and identity of their parent completing the survey (i.e., mother or father). The collection of SES information adopted Han’s approach40considering the occupational, educational, and economic statuses of each family. Each of the three SES aspects was rated against a 7-level Likert scale (i.e., occupational [1 = unemployed or partially employed, 7 = senior government executives or large private business owners], educational [1 = primary or below, 7 = doctoral], and economic [1 = 1300RMB or below, 7 = 7600RMB]), with a total of 21 points maximum possible. The total SES scores were subsequently categorized into five SES groups: very high SES = 18–21, high SES = 15–17, moderate SES = 11–14, low SES = 8–10, and very low SES = 1–7. Weight and height were assessed using a physician scale (i.e. device model # LK-T1016; Shenzhen Lingkang Electronics Co., Ltd) to capture child participants’ height and weight, which were subsequently converted to BMI scores (formula: weight / height2 [weight: kg, height: meter]). The World Health Organization (WHO) Child Growth Standards41 were applied to convert BMI raw data to BMI Z scores which were categorized into normal (-2SD to 1SD), underweight (below − 2SD), overweight (1SD to 2SD), and obese (above 2SD).

Data collection

Batteries of CAPL-2 (i.e., written survey [demographic information, tests of cognitive and affective domains, and self-reported weekly frequency of active days], and physical tests [CAMSA, PACER, and plank]) were administered sequentially in three physical education classes during one week, followed by daily step assessment using pedometers in the following week. Sufficient warmup of five to 10 min was required prior to all physical tests, which took place at the school’s open track stadium. Children’s physical literacy was assessed by two certified physical education teachers (9 years of teaching experiences on average) who had received sufficient research training one week prior to data collection.

In the first physical education lesson, the children completed the paper-based survey (including cognitive, affective, and behavioral domains) in a quiet classroom at the beginning of the class, averaging 13 to 19 min. After the written survey, children were organized to take the plank test (10 children each round of test). The CAMSA (using two tool kits) and PACER (15 children each round of test) were administered in the second lesson on a different day of the week with a 10-minute interval between different tests. To complete CAMSA, each child was given one free trial, and then formally tested twice from which the better performance score was recorded for data analysis. The average performance time per trial ranged from 17 to 34 s. Subsequently, PACER was conducted on marked flat lanes of 20 m, with five to seven minutes spent for each round of PACER test. The third lesson was for makeup, as some children missed some tests in the first two lessons.

Parents were asked to assist children’ daily steps assessment using pedometers for seven consecutive days. On the day before the formal step count assessment, each child participant received a pedometer and a daily step log sheet, and were instructed by their physical education teachers about the assessment protocol, including instructions on the appropriate wearing of pedometer, when to take on/off the monitor, how to record the daily step count, and steps to submit the completed log sheets and pedometers as well as trouble-shooting tips. To minimize disruption to schooling, the children resumed to their regular physical education activities as planned in the curriculum after the assessment session.

Data analysis

Descriptive analysis was used to capture group characteristics (i.e., mean [M] and standard deviation [SD]) of PPPA, children’s physical literacy, and their domains by gender, age, BMI category, and SES. Inferential analyses were sequentially used to address the main research questions: (a) one-way analysis of variance (ANOVA) to examine grouping effects, (b) correlational analysis to determine the strength and direction of the bivariate relationships between PPPA and children’s physical literacy as well as their domains, and (c) a series of hierarchical linear models (HLM) to examine the predictive effects of PPPA/PPPA dimensions on children’s physical literacy and its four domains. HLM was used, because the children nested in different classes and were taught by different head teachers and physical education teachers, which might have yielded the clustering effect (i.e. students within the same class share more similarities than those from other classes). Intra-class correlation (ICC = σα2/[σα2 + σε2]; σα2: variance between clusters, and σε2: the variance within clusters) coefficients were calculated to further determine the need for conducting the HLM analysis (criteria: ICC coef. ≥ 0.05)42. To account for the potential clustering effect, HLM was designed to have two levels, with children (level one) nested in classes (level two). Specifically, demographic variables (i.e., gender [dummy variable: boy = 1, girl = 0], age, and SES), BMI, and gender of the participating parent (dummy variable: father = 1, mother = 0) were entered as covariates, while total physical literacy and the four domains (i.e., affective, cognitive, behavioral, and physical domains) were regressed on PPPA (model A) and the five dimensions of PPPA (model B), respectively (with random slopes and intercept to each model). Eventually, ten models were built following the models specified below. Partial R2 were calculated for each predictor in the linear models. Cohen’s d was reported to reflect group difference. Significant level was set at 0.05. All statistical analyses were conducted using SPSS and R.4.2.1.

Level 1 of Model A (predictor: attitude, awareness, value, cognition, and appreciation):

\(\begin{aligned} {Y_{{\text{ij}}}} & ={b_{0{\text{j}}}}+{b_{{\text{1j}}}}*{\text{Gender}}+{b_{{\text{2j}}}}*{\text{Age}}+{b_{{\text{3j}}}}*{\text{BMI}}+{b_{{\text{4j}}}}*{\text{SES}}+{b_{{\text{5j}}}}*{\text{PPPA}}.{\text{Altitude}}+{b_{{\text{6j}}}}*{\text{PPPA}}.{\text{Awareness}}+{b_{{\text{7j}}}}*{\text{PPPA}}.{\text{Value}} \\ & \quad +{b_{{\text{8j}}}}*{\text{PPPA}}.{\text{Cog}}+{b_{{\text{9j}}}}*{\text{PPPA}}.{\text{Appre}}+{b_{{\text{1}}0{\text{j}}}}*{\text{Parent}}.{\text{Gender }}+{\varepsilon _{{\text{ij}}}} \\ \end{aligned}\)

Level 2 of Model A:

\(\begin{aligned} & {\text{Intercept}}:{b_{0{\text{j}}}}\,=\,{{\gamma_{00}}}\,+\,{{\mu_{0{\text{j}}}}} \\ & {\text{Slope}}:{b_{{\text{5j}}}}\,=\,{{\gamma_{{\text{5}}0}}}\,+\,{{\mu_{{\text{5j}}}}} \\ & {\text{Slope}}:{b_{{\text{6j}}}}\,=\,{{\gamma_{{\text{6}}0}}}\,+\,{{\mu_{{\text{6j}}}}} \\ & {\text{Slope}}:{b_{{\text{7j}}}}\,=\,{{\gamma_{{\text{7}}0}}}\,+\,{{\mu_{{\text{7j}}}}} \\ & {\text{Slope}}:{b_{{\text{8j}}}}\,=\,{{\gamma_{{\text{8}}0}}}\,+\,{{\mu_{{\text{8j}}}}} \\ & {\text{Slope}}:{b_{{\text{9j}}}}\,=\,{{\gamma_{{\text{9}}0}}}\,+\,{{\mu_{{\text{9j}}}}} \\ \end{aligned}\)

Level 1 of Model B (predictors: PPPA):

\({Y_{{\text{ij}}}}\,=\,{b_{0{\text{j}}}}\,+\,{b_{{\text{1j}}}}*{\text{Gender}}\,+\,{b_{{\text{2j}}}}*{\text{Age}}\,+\,{b_{{\text{3j}}}}*{\text{BMI}}\,+\,{b_{{\text{4j}}}}*{\text{SES}}\,+\,{b_{{\text{5j}}}}*{\text{PPPA}}\,+\,{b_{{\text{6j}}}}*{\text{Parent}}.{\text{Gender}}+{\varepsilon _{{\text{ij}}}}\)

Level 2 of Model B:

\(\begin{aligned} & {\text{Intercept}}:{b_{0{\text{j}}}}\,=\,{{\gamma_{00}}}\,+\,{{\mu_{0{\text{j}}}}} \\ & {\text{Slope}}:{b_{{\text{5j}}}}\,=\,{{\gamma_{{\text{5}}0}}}\,+\,{{\mu_{{\text{5j}}}}} \\ \end{aligned}\)

Results

Group characteristics

Tables 1 and 2 report the descriptive results for PPPA/PPPA dimensions and children’s physical literacy /its four domains by gender, age, BMI category, and SES category. One-way ANOVA showed significant group differences by gender (physical domain [F(1, 193) = 4.91, p = 0.03, η2 = 0.03]), grade (total physical literacy score [F(3, 191) = 3.88, p = 0.01, η2partial = 0.06]; cognitive domain [F3, 191) = 4.91, p < 0.01, η2partial = 0.07]; affective domain [F(3, 191) = 11.33, p < 0.01, η2partial = 0.15]; behavioral domain [F(3, 191) = 4.15, p = 0.01, η2partial = 0.06]; and physical domain [F(3, 191) = 3.71, p = 0.01, η2partial = 0.06]), BMI (physical domain [F(3, 191) = 2.86, p = 0.04, η2partial = 0.04), and SES (total physical literacy score [F(3, 191) = 82.83, p < 0.01, η2partial = 0.57]; affective domain [F(3, 191) = 15.17, p < 0.01, η2partial = 0.19]; behavioral domain [F(3, 191) = 42.25, p < 0.01, η2partial = 0.40]; and physical domain [F(3, 191) = 11.96, p < 0.01, η2partial = 0.16]). We did not conduct group difference testing on PPPA and its dimensions using inferential analyses, because rarely has prior research examined the differences in PPPA by the demographic and anthropometric variables of children.

Table 1 Descriptive results of parents’ perceptions of physical activity (PPPA).
Full size table

For group characteristics, boys scored higher in the physical domain than girls (Cohen’s d = 0.32). While younger children scored higher in affective (grade 4 vs. grade 5, Cohen’s d = 1.42; grade 2 vs. grade 5, Cohen’s d = 1.07; grade 2 vs. grade 3, Cohen’s d = 0.57; but lower in grade 3 than grade 4 [Cohen’s d = 0.85]), and behavioral (grade 2 vs. grade 5, Cohen’s d = 0.74) domains, older children scored favorably higher in cognitive (grade 2 vs. grade 5, Cohen’s d = 0.85) and physical (grade 3 vs. grade 5, Cohen’s d = 0.77; grade 2 vs. grade 5, Cohen’s d = 0.70) domains. Physical domain also favored children with normal weight compared to those with obesity (Cohen’s d = 0.49). Total physical literacy score favored very high SES families over high SES (Cohen’s d = 1.08) and moderate (Cohen’s d = 2.68) SES families, and high SES over moderate SES families (Cohen’s d = 1.51). In addition, the affective and behavioral domains favored the very high SES (Cohen’s d = 1.13; Cohen’s d = 1.79) and high SES families (Cohen’s d = 0.69; Cohen’s d = 1.28) over moderate SES families. Lastly, the physical domain of physical literacy favored the very high SES families over high SES (Cohen’s d = 0.81) and moderate SES families (Cohen’s d = 0.98).

Table 2 Group characteristics of children’s physical literacy.
Full size table

Results of bivariate correlation

Table 3 shows the correlational matrix for the variables. Correlation analyses demonstrated significant bivariate correlation coefficients (girls: 0.22; boys: 0.20–0.34) between PPPA/PPPA dimensions and children’s physical literacy / its domains (see Table 3 for details).

Table 3 Correlation matrix of parents’ perceptions of physical activity (PPPA) and children’s physical literacy by gender.
Full size table

Results of HLMs

We calculated the ICC to gauge the degree of nesting effect. The calculation revealed that 6.82% –15.37% of the variance was attributable to the nesting nature of the data (Table 4), warranting the use of HLM. The HLM resulted in seven significant associations between PPPA/PPPA dimensions and children’s physical literacy /its domains (Table 5). Significant predictive effects of PPPA were found on total physical literacy (β = 0.61, t = 3.48, p < 0.01, R2partial < 0.01), cognitive domain (β = 0.11, t = 2.66, p = 0.03, R2partial < 0.01), and physical domain (β = 0.17, t = 2.88, p < 0.01, R2partial < 0.01). Parental valuing showed a significant predictive effect on total physical literacy (β = 0.88, t = 2.50, p = 0.01, R2partial < 0.01), physical domain (β = 0.27, t = 2.17, p = 0.03, R2partial = 0.02), and affective domain (β = 0.32, t = 2.31, p = 0.02, R2partial = 0.01). Moreover, parental understanding showed a significant predictive effect on total physical literacy (β = 0.91, t = 2.06, p = 0.04, R2partial < 0.01).

Table 4 Estimates of covariance parameters and intra-class correlation (ICC) across five intercept-only models (students nested in class).
Full size table
Table 5 Statistical results for the HLMs: Estimates of fixed effects of parents’ perceptions of physical activity (PPPA) on children physical literacy and its four domains.
Full size table

Discussion

The purpose of this study was to describe the demographic and anthropometric differences in children’s physical literacy and its four domains, and to ascertain the predictive effects of PPPA/PPPA dimensions on children’ physical literacy and its four domains. The results of the statistical analyses showed a number of group differences in children’s physical literacy and predictive relationships between PPPA/PPPA dimensions and children’s physical literacy / its four domains. These main findings are discussed below.

Group differences of gender, age, and BMI category were only found in physical literacy domains rather than total physical literacy. By gender, boys’ higher achievement in physical domain was in line with previous studies5,35,37. However, other studies5,43,44 reporting girls’ higher achievements over boys in cognitive domain and some items of physical domain (i.e., sit-and-reach) suggest inconsistency of gender-based differences. Similarly, mixed findings were observed by grade level. While older children’s higher scores in cognitive and physical domains of physical literacy (vs. younger children) aligns with previous studies5,36,37, the observation that behavioral and effective domains favoring younger children over older children conflicts with some5,37,44 but not all36,44,45,46 previous findings. The mixed findings on cognitive, physical, behavioral, and affective domains across gender and age groups warrant more future investigations to ascertain these differences. By BMI, children with normal BMI demonstrated higher physical domain than those with obesity, which is consistent with the observations made by two previous studies38,47.

Family SES was found to be an important determinant of total physical literacy and three physical literacy domains. Children from very high SES families demonstrated higher scores in total physical literacy and affective, behavioral and physical domains than these from high and moderate SES families; and children from high SES families were more physically literate than children from moderate SES families. Similar findings were reported by Kozera41 where they found higher motor competence in children from high SES schools than those from low SES schools. Another study5 also reported the similar group pattern where total physical literacy and affective and behavioral domains favored children from higher SES families.

These findings concerning group differences in physical literacy domains across gender, age, BMI category, and SES further supported the need for considering sociodemographic and anthropometric factors in designing children’ physical literacy programming across settings (e.g., home, school, youth sports).

Our study further demonstrated meaningful results supporting that parents with favorable attitude, awareness, value, understanding, and appreciation of physical activity are more likely to nurture physically literate children, regardless of children’ gender, age and BMI, and family SES. Of these parental perceptions, parental valuing and understanding of physical activity are particularly important to children’ physical literacy development. This observation is consistent with prior research14,18,19,24,48,49. For example, previous studies have found that parents with higher valuing and better understanding of the principles and benefits of physical activity are more likely to engage in sports and/or physical exercises14,19. These parents are also more likely to embrace a sports-friendly parenting style at home and nurture active children18,24,48,49. In addition, the positive predictive relationship of PPPA on the cognitive and physical domains of children’ physical literacy further indicates that parents with higher PPPA are critically important to children’s physical literacy development (e.g., fostering more knowledgeable and physically fit kids)2,50.

Lastly, parental valuing, a dimension of the PPPA, further demonstrated significant predictive effect on the affective and physical domains of children’s physical literacy. Perceived value encompasses usefulness, interest, importance, and cost51. In the physical activity context, perceived value may contribute to parents’ cognitive and affective attitudes toward the necessity and importance of regularly participating in physical activity (e.g., sports, physical exercise)24,51. Likewise, parents who attach greater importance to physical activity and appreciate the benefits of physical activity may have higher expectation on their children’s knowledge, skill, performance, and behavior related to physical activity. This finding highlights the pivotal role of parents’ valuing of physical activity in shaping a physically supportive parenting style and sequentially promoting children’s confidence and competence (psychologically and physically) for physical activity participation9,21,22,23 and physical literacy development23,24,26.

Limitations

We acknowledge two limitations of this study. First, the sample consisted of participants recruited from one primary school located in central China and their parents. The findings may not be generalizable to broader populations and settings. We did consider recruiting a more diverse sample from multiple schools across geographic areas. However, conducting the simplified Chinese CAPL-2 and collecting parental survey data were time-consuming. Limited by resources, we eventually decided to focus our research on comprehensively assessing a large sample of participants from one school rather than expanding the research to other sites. Future research with more resources should consider recruiting more participants from multiple schools, regions, countries or cultures. Secondly, we employed the correlational research design to ascertain the predictive relationship between PPPA and children’s physical literacy. Given that, the research findings from correlational analyses should not be interpreted as cause-and-effect in nature.

Conclusions

The findings suggested that children’s physical literacy differ by sociodemographic and anthropometric characteristics and that parents’ perceptions (i.e. PPPA) and its dimensions (e.g., parental valuing) are important factors to consider when fostering children’s physical literacy (e.g., cognitive and physical domains, in particular). While physical literacy development requires concerted efforts across settings (e.g., school, community, and home), this study has highlighted the significance of having positive parental perceptions (and parenting styles and practices) in nurturing physically literate children.