1. Introduction
The concept of physical literacy (PL) rapidly gained global attention and led to a research boom [
1,
2]. Given its potential value for transforming society from a physical activity (PA) suppressed culture to a PA rich one [
3], many countries began to promote PL in various environments, including schools, communities, and public health institutes [
4]. The United Nations Educational, Scientific and Cultural Organization (UNESCO) stated that PL is a key component of physical education (PE) [
5]. England Youth Sport Trust set PL as the basis of PE and school sport [
6,
7]. The Society of Health and Physical Educators (SHAPE) America revised the outcome of K-12 PE to cultivate PL people [
8]. The World Health Organization (WHO) also stated that PL should be aligned with health when developing a national action plan on PA [
9]. Although most institutes and researchers referenced the definition from Whitehead claimed: “the motivation, confidence, physical competence, knowledge and understanding to value and take responsibility for engagement in PA for life” [
10], each country and organization promoted their own PL on the basis of cultural difference. Canada adopted the concept early and it flourished in communities [
11]. The Aspen Institute, in the United States, drew up ability, confidence, and desire as key words for people to have PL [
12]. Australia added a social dimension to the concept [
11]. New Zealand, likewise, added a spiritual dimension [
11]. Based on flourishing research, a growing body of work exists that examines the issue of measurement in the PL domain. Canada launched the passport for life program to assess the PL of students and teachers through active participation and living, fitness, and movement skills [
13]. Physical Literacy Assessment for Youth (PLAY) was developed by Canada Sport for Life to examine PL. PLAY has six separate instruments for different groups of people [
14]. The Canadian Assessment of Physical Literacy for children and youth (CAPL) was created to measure PL in four dimensions, namely, physical competence, daily behavior, motivation and confidence, and knowledge and understanding [
15].
Drawing from the current research on PL assessments, many researchers are looking for extensions and applications of PL in cross-sectional fields. Cairney and colleagues introduced a conceptual model that positions PL as a health determinant and concluded that PL can be measured [
16]. Jefferies and colleagues showed a link between PL and resiliency [
17]. Kwan and colleagues also conducted an intervention study to examine the impact of PL on PA behaviors and fitness in university students [
18]. Roetert and colleagues established the bridge between PL and PE. However, non-English speaking regions had difficulty adopting PL because of cultural differences. Raymond and colleagues first invented a perceived physical literacy instrument (PPLI) to measure the PL of students and PE teachers in Hong Kong [
19,
20] and tested the relationship between PL and PA in Hong Kong adolescents [
21]. Then, Ma and colleagues translated PPLI into a simplified Chinese version (PPLI-SC) for testing PL among Chinese undergraduates [
22].
Although Mainland China announced a policy to develop PL among students [
23], the emerging concept still remains at a superficial stage. Most previous studies were limited by translated PL as sports literacy or PE literacy. Researches on the localization of the concept of PL remain slow and without a solid conclusion. Thus, more research should be conducted to improve physical activity (PA) through PL. Exploring the relationship between PL and PA will be the first step. As Whitehead stated, PL is not equal to PA, nor is it a PA substitute. The relationship between PL and PA is one of facilitation. PL does not have to be manifested in PA, people, including those who are not able to perform any PA, can demonstrate and benefit from PL [
11]. PL is the predecessor of PA, which is developed through PL [
24,
25]. Studies exist that focus on the relationship between PL and PA for different age groups in other countries and regions [
21,
26,
27]. Studies found that children’s PL was associated with their PA the authors suggested future research to test the causality of the association [
27]. Other study proved that among Hong Kong adolescents, instead of being forced to participate in PA, individuals will take an active part in PA if they understand PL. Concurrently, individuals will enhance their PL by joining in PA, then forming a beneficial cycle [
21]. Moreover, moderate-to-vigorous PA was proposed to be the combination of PL and PA in school-age children [
26]. For university students, Kwan and colleagues found the PL-based intervention program was effective in helping university students attenuate PA decline in the first year, though the underlying mechanisms remain unclear [
18]. However, Mainland China has yet to conduct such research. This study thus aimed to test the causality of the association between PL and PA in Mainland China. PL was collected by using PPLI-SC [
22]. Different from the original Cantonese version, PPLI-SC re-collected 18 items to test the reliability and validity among Chinese undergraduates and then developed a unique instrument that fits the local culture. For PA, PA level was used to present the behavior and exercise intensities of undergraduates [
28,
29]. The subject was the last stage of the education process from primary school, high school, and college [
30]. Young people at this stage should take PA seriously because it is a significant part of being healthy [
31]. Given that undergraduates are essentially on campus daily, university plays a vital role in developing and maintaining continued PA participation. Understanding the relationship between PA and PL can help improve the school’s PE courses, which enhances students’ health. In addition, Patricia and colleagues proposed that individual factors such as gender, educational factor, and socioeconomic status (SES) might impact the PL [
24]. There was little research testing the relationship between PL and PA in different demographic groups. This study will be the first to explore PL in relation to demographic groups among Chinese undergraduates.
3. Results
Before testing the relationship between PA and PL, demographic characteristics of participants were analyzed. A total of 622 undergraduates were recruited and 536 of them completed the questionnaires (female = 403; male = 133). The participants’ ages ranged from 18 to 21 years old. The average age of participants was 19.40 years old and standard deviation was 0.83, which were all consistent with the research subjects.
Pearson’s product-moment correlation was used to test the correlation between each dimension of PA level and three attributes of perceived PL (
Table 1). The average of total perceived PL of participants was 30.28 (17–40, standard deviation ±4.20) and the average of three domains, (1) motivation, (2) confidence and physical competence, and (3) interaction with the environment were 11.92 (5–15, ±0.25), 11.09 (3–15, ±0.29), and 7.27 (2–10, ±0.23), respectively. The daily average time the participants spent on moderate PA and vigorous PA were 7.34 (2–60, ±9.06) minutes and 4.94 (1–34.42, ±6.9) minutes, respectively. The daily average transportation time of students was 10.57 (1–58.14, ±8.73) minutes, which included 1.71 (0–27.4, ±0.12) minutes riding, 0.14 (0–17.14, ±0.04) minutes cycling, and 8.5 (3–25.7, ±0.32) minutes walking. Daily average housework and leisure time were 4.29 (0–60, ±6.74) minutes and 13.52 (0–62.85, ±12.5) minutes.
The correlation (r) between participants’ perceived PL and PA levels was 0.35, which was significant at the 0.01 level (2-tailed). Although the correlations are not high, three dimensions of PPLI-SC, namely, intensities of PA (walking, moderate PA, and vigorous PA) and three domains of IPAQ-SC (transportation, housework, and leisure time) showed a high significance (r = 0.088–0.336, p < 0.01). No significant correlation existed between sedentary domain of IPAQ-SC and any other attribute.
Multiple linear regression was used to predict the PA levels of Chinese undergraduates on the basis of the three dimensions of perceived PL (
Table 2). The analysis showed a significant regression equation (F = 25.228,
p < 0.01, ΔR
2 = 0.120) and so did the dimensions of perceived PL. The predicted PA level of participants was −3818.582 + 272.535 (motivation) + 249.848 (confidence and physical competence) + 149.899 (interaction with the environment) MET value. The standardized coefficients of (1) motivation, (2) confidence and physical competence, and (3) interaction with the environment were 0.176, 0.184, and 0.087, respectively. Except for interaction with the environment, the probability values of the other two dimensions were all less than 0.05, which showed a positive relationship between perceived PL and PA level. The participants’ MET values would increase by 272.535, 249.848, and 149.899 for each dimension score of perceived PL. The results showed that two dimensions were significant predictors of PA level.
Multiple linear regression was then used to examine males and females separately (
Table 3 and
Table 4). The results showed a significant regression equation both in males (F = 9.219,
p < 0.01, ΔR
2 = 0.157) and females (F = 15.243,
p < 0.001, ΔR
2 = 0.098). For PL dimensions in males, only motivation showed significant probability value (
p = 0.048). For females, the probability values of motivation and physical competence were both less than 0.05.
SES (
p = 0.092) and GPA (r = 0.119,
p = 0.022) factors also examined the association by using Pearson’s product-moment correlation. Except for physical competence (r = 0.039,
p = 0.461), motivation (r = 0.116,
p = 0.022) and interaction with environment (r = 0.131,
p = 0.012) both showed significant correlation with GPA.
Table 5 shows the results of the standard regression analysis to predict PL using GPA (F = 5.294). For each point increase in GPA, the perceived PL score will is enhanced by 0.857. Comparison of the different PL in gender (t = 2.174,
p < 0.01) was explored by student’s
t-test. Perceived PL of males (31.02, ±4.97) were higher than females (29.98, ±3.88). BMI (F = 7.381,
p = 0.001) was classified into three types and analyzed giving their different PL through ANOVA. BMI was classified into three types, namely, low, medium, and high. The medium group showed significant differences with low (
p = 0.001) and high (
p = 0.015) groups. For further analysis, z-scores were calculated to explore the different influences of each group in the relationship between PL and PA (
Table 6). Except for the high GPA group (r = 0.025, z = 0.025), all other groups showed significant positive correlation between PL and PA (r = 0.338–0.529, r = 0.352–0.589).