The notion of talent and its development have been explored in other domains, and while providing some intriguing and relevant elements for those working in sport settings, these domains may be insufficient to capture the complexity of talent in sports (Joseph et al., 2019). Many scholars have carried out much relevant research on TID before including multidisciplinary, longitudinal, prospective, or qualitative studies, such as the investigation and research on the process of sports professional TID (Ford et al., 2020), the current model and future development direction of TID in sports (Vaeyens et al., 2008), and the predictive factors in the process of TID (Williams et al., 2020). In terms of the concept and content of sport talent, it is to point to a career in sports, sports professional knowledge, skills, and technical talents, mainly includes the athletes, coaches, sports scientific research personnel, sports management, sports teacher, entrepreneur, and sports agent for the development of undertakings of physical culture and sports as well as promote social progress, has made certain contribution or have some contribution to the potential of people. TID is a multi-factorial process (Ricard et al., 2018) and relate to identifying people who participate in sports with the potential to become professional athletes through training and cultivating.

According to the theory of subject knowledge, a multidisciplinary approach is used for TID including physical, technical, and psychological predictor variables in players (e.g., adult height, maturity, sprinting, dribbling skills, and intrinsic motivation) (Williams and Reilly, 2000; Ford et al., 2020). Hence, TID is a linear process, where sports competitions source athletes for the existing vacancies. This linear process is reactive in its nature and gift, thus lead to increased time to finding and cost to cultivating. From of the aforementioned discussion, we can know that there is increasing research on sport TID carried out around the concept, method, content of sport TID, and other aspects related to TID. As an important part of cultivating sports talents, PE class plays a significant role in transporting sports reserve forces for national sports, as well as sport TID.

In this process, TID is associated with many aspects including not only anthropology, pedagogy, physiology, and psychology but also sociology and coaching science (Williams and Drust, 2012). In another words, TID is an important stage of talent growth, and in this complex and multidimensional process and is affected by multi-dimensional factors. There have been many research studies on TID in the past decade, including methods of TID (Vladan et al., 2009; Clarke et al., 2018; Kyle et al., 2019; Williams et al., 2020) and identification factors (Norikazu and Taigo, 2016; Gledhill et al., 2017; Dodd and Newans, 2018; Lai and Ishizaka, 2020). Among them, physical factors are more discussed in TID. Genetic factors are considered to play a critical role in athletic performance and related phenotypes (Miah and Rich, 2006; Ahmetov et al., 2016; David et al., 2017). There is a consensus in the scientific and sporting communities that genetic factors contribute to athletic performance. According to research, at least 155 genetic markers were found to be related to elite athlete performance (93 genetic markers related to endurance and 62 genetic markers related to power/strength). Meanwhile, genome-wide association studies (GWASs) are the most commonly used method to identify athletic performance in athletes (Ahmetov et al., 2016). Components like height, body fat percentage, size of palm, dynamic balance, static balance, and hand strength were the key elements in TID of badminton (Mojtaba et al., 2012). Therefore, anthropometric factors index plays a crucial role in TID process.

Psychological factors are also very significant in the process of TID and talent development, and a focus that has grown markedly over the last decade or more. Saby et al. (2020) used a longitudinal method to examine changes in appraisals, emotion regulation, and emotions in adolescent soccer players based at a competitive season in France and found intertwined psychological constructs in a dynamic relationship allowing athletes to continuously adjust to their constantly changing everyday demands. Similarly, Saward et al. (2016) used a mixed-longitudinal prospective method to examine the psychological characteristics associated with elite youth soccer players aged 8–18 years in three seasons. Apart from these, Rongen et al. (2020) used a prospective longitudinal cohort design to track psychosocial outcomes of academy involvement in male youth elite soccer players and suggested that negative psychosocial impacts of soccer academy involvement did not materialize in this context. Enrique and William (2006) through 24 in-depth semi-structured interviews with adolescents who involved in competitive sports concluded that motivational roles are related to the interpersonal influence on adolescents' sports motivation.

Compared with the first two factors, sociological factors are less discussed, but there are still some research on this. For instance, self-regulation and adaptive volitional behaviors appear to be key intra-individual factors associated with talent development (Haugaasen et al., 2014), and a number of talent development environmental factors (e.g., media coverage, sports participation rate, birthplace, long-term development, and quality preparation) also exist in sports (Li et al., 2014). On the whole, according to the current research, there are many factors affecting the sport TID, mainly physiological, psychological, and sociological factors, which play a vital role in both the identify of sports talents and the development of sports talents.

Although there are a great deal of studies on the influencing factors of sport TID, most of them focus on a specific aspect, such as the relative age effect (RAE), athletes' psychological characteristics, and students' growing environment around these factors. This study used the relevant literature to analyze the multidisciplinary factors affecting sport TID as much as possible and verified these influencing factors through structural equation modeling as a way to better improve the success rate of sports talent. In this study, on the basis of a questionnaire survey, a Likert scale was used to assign scores to the listed influencing factors, and the data collected by questionnaire were analyzed. The questionnaires were distributed to PE teachers, sports administrators, and coaches, who were asked to rate the degree of influence of the factors in the questionnaire. The analysis of the data shows that physical, psychological, coach, environmental factors all have a positive impact on the sport TID.

As an integral part of education, PE is a purposeful, planned, and organized educational process conducted through physical activities and other auxiliary means (Mountakis, 2001). In a systematic review, Prieto-Ayuso et al. (2020) showed that methods and instruments for talent identification in PE are changing, so it is important to select appropriate programs to deal with gifted children. In addition, PE teachers in the development of the scale can help speed up the identification of potential talents (Platvoet et al., 2015); however, teachers in school PE still face certain difficulties within the identification process (Bailey et al., 2009).

In another aspect, there are many factors influencing sport TID in the school PE curriculum, due to the complex environment and existing regulations. Thus, based on the literature, expert consultation, and the factors associated with TID, a questionnaire was designed and distributed through online and offline. Researchers should have a reasonable theoretical basis and specific literature support when setting up models (Hoyle and Panter, 1995; Boomsma, 2000; McDonald and Ho, 2002). The setting of the model in this study includes relevant theories of the variables in the research model and the essence of past research. In this research, the authors selected physical, psychological, coach, and environment as four external variables, corresponding to 12 observation variables, respectively, and sport TID is as internal latent variables.

Although the existing literature on sport TID is very extensive, there are still some deficiencies. The development of school PE promotes the sport TID, and the development of sport TID will promote the innovation and progress of the school PE curriculum. However, the sport TID in the school PE curriculum is affected by many factors. For this reason, identifying the factors that affect TID in the school PE curriculum can improve the identification rate of school sports talent and further consolidate the strength of sports reserve talent.

This research aims to reveal factors influencing sport TID in the PE curriculum in China, and the following questions will be discussed in this research:

Physical quality is not only the foundation guarantee of athletes but also the basis of special sports quality. On the one hand, Norikazu and Taigo (2016) through 2-year follow-up measurements of soccer players found that factors of the anthropometric index such as sprint ability is a useful identification index and muscular power is limited and useful in TID of soccer. On the other hand, physiological aspects included anthropometrical, linear speed, change of direction speed (CODS), maximal anaerobic power, repeated sprint ability (RSA), maximal aerobic power, and maximal lower body strength, and they should be taken into account in the TID of soccer in the testing batteries (Dodd and Newans, 2018). Hence, we propose the following hypotheses:

Psychological influence sport TID means the mentality maturity of the personal are advantaged on performance tests (Vaeyens et al., 2008), psychological factors manifesting itself at various levels, having a predominantly unconscious nature. Likewise, a systematic review also indicated that psychological factors (e.g., adaptive perfectionism, task/mastery orientation, delaying gratification, and coping strategies) are associated with talent development in football and suggested that psychological characteristics of self-regulation, resilience, commitment, and discipline appear to be most impactful on TID (Gledhill et al., 2017). Based on this, this research propose the following hypotheses:

In the same way, coaches' knowledge is associated with talent detection, but it is unable to articulate how or in what ways they see talent (Roberts et al., 2019). An increasing number of researchers have explored the value of the “coach's eye” in TID and talent evaluation settings. For instance, a research study through systematic review and meta-analysis describing a model from decision-making concluded that a coach's eye is subjective and experience-based, and these criteria can differ from coach to coach (Lath et al., 2021). In another aspect, in the process of TID, because the coach's eye is based on intuition, coaches often fail to state clear variables and reasons for their selection decisions (Roberts et al., 2019). Therefore, we propose the following hypotheses:

Social and environmental factors that can impact upon TID (Mills et al., 2014): John et al. (2019) through systematic review found that critical life events have an impact on developmental pathways of elite athletes. In the meantime, from another point of view, the organization operation and the human resource management (HRM) system are commonly associated with talent management, whereas talent development can be further enhanced by influencing personal and environmental catalysts (Meyers et al., 2013). Accordingly, this study proposes the following hypotheses:

A total of 330 questionnaires were sent and collected. Invalid questionnaires were removed, and the returned questionnaires were carefully reviewed. Finally, 310 questionnaires were determined to be valid, with a questionnaire effective rate of 93.9%.

The information about the respondents in the confirmed valid questionnaire is as follows: (1) Gender and age composition: The female proportion was 48.38%, and the proportion of people under 30 years is 21.3% and that between 30 and 45 years is 36.7%; (2) The situation of professional title: In this survey, the title of professor accounts for 7%, and the title of associate professor was 9%. (3) Occupational distribution: The percentage of PE teachers was 30.64%, coaches were 24.84%, and the rest were PE administration and education administration personnel. (4) Education level: The proportion of those with doctoral degrees was 17.42%. (5) Type of work unit and work experience: The number of respondents working in colleges and universities accounted for nearly half, 48.39%. (6) Work experience: 13.2% had <3 years of experience, 27.9% had 3–5 years, 40.6% had 5–10 years, and 18.3% had more than 10 years of work experience.

The procedure is to distribute the questionnaires online (e.g., e-mail and Google online questionnaire) and offline and collect them after they are filled in. After repeatedly revising and determining the questionnaire, the researchers conducted a questionnaire survey on the factors affecting the sport TID in China's PE curriculum among PE teachers, coaches, and sports experts, and collected the sample data of the questionnaire survey.

The majority of the survey studies of performance relied on the questionnaire for date collection (Doherty, 1998). The respondents of this study are mainly concentrated in Guangdong province, according to the number of staff in the sports system in the Guangdong Provincial Statistical Yearbook 2021 (http://stats.gd.gov.cn/gdtjnj/), there were 1,874 sports coaches, 927 full-time teachers, and 5,224 administrative personnel. Overall, 330 participants were selected for this research. Then through the method of random sampling, an appropriate number of individuals are randomly selected in each subgroup. Last, based on the principle of stratified sampling, 77 people are selected from sports coaches, 38 teachers, and 215 sports management personnel. Hence, the sample size in this study is 330 sports system practitioners (i.e., administrative personnel, coaches, and school PE teachers). Sports experts, administrative personnel, middle school PE teachers, and coaches from Guangdong province were selected to represent the south of China. In this way, the questionnaire was distributed among the school PE and education system practitioners by random sampling.

For the purpose of ensuring the validity of this questionnaire, it is designed from multiple angles. By reading the relevant literature and expert interviews, combined with the characteristics of school PE classes and talent growth environment, a questionnaire has been designed. After the questionnaire is collected, it is necessary to process information. The design of this questionnaire for this study was structured and multiple-choice type. It was divided into three parts. The first part comprised the basic information of the respondents, including personal information (gender, professional title, age, working unit, and education level) and basic information of the affiliation (affiliation category). These questions ensure the validity of the samples and allow the survey results to be classified and cross-analyzed. The second part is the evaluation of the influencing factors, which includes 12 influencing factors (Table 1), with the degree of importance ranging from high impact, moderate impact, low impact, rare impact to no impact, and 5, 4, 3, 2, and 1 points are given, respectively. The third part comprises other factors affecting the detection of sports talents in the school PE curriculum. This part of special measurement is mainly to collect some specific factors of the sports TID in school PE classes in China of this questionnaire. The aim of the part is to collect respondents' judgments on the degree of influence between variables.

The structural equation model (SEM) is an empirical analysis model method that uses the linear equation system to express the relationship between observation variables and latent variables, as well as the relationship between latent variables. Compared with the traditional linear regression model, the SEM has the advantage of processing multiple variables at a time, allowing latent variables have multiple index variables and resulting in the reliability and validity of index variables, so that the model can adapt to a wider range and be more elastic. In recent years, the SEM has become the main analytical method in many fields such as sports (Marsh, 2007; Rocha and Chelladurai, 2012), education (Violato and Hecker, 2007; Khine, 2013; Green, 2016), and general management (Williams et al., 2009). The measurement model describes the relationship between latent variables and observed variables, while the structural model describes the relationship between latent variables (Figures 1, 2).

Analysis of moment structure (AMOS) can verify various measurement models and different path analysis models. In addition, multi-group analysis and the average structure number test can also be conducted. AMOS has four advantages (Cunningham and Wang, 2005): a graphical interface that helps see the changes that occur in the model and to establish relationships between the surfaces, easy to interpret and understand the meaning of the model diagrams created, a clear and intuitive way to read the results, and a full graphical display of the AMOS results.

In this research, the applicability of the four dimensions of TID was tested by SPSS26.0 according to the principal component reduction data set dimension, in which the KMO test value was 0.902, and the significance probability of the χ² statistical value of the Bartlett sphere test was 0.000 (Table 2), indicating that the feasibility standard of principal component analysis was fully met.

Confirmatory factor analysis (CFA) is a part of SEM analysis and plays a significant role in SEM analysis (MacCallum and Austin, 2000; Brown, 2006). CFA mainly presents the commonality of factor load and measurement model factor variables in tables. The contents should include standardized and non-standardized load, standard error, significance, composition reliability, extraction of mean variance, and appropriate fitting indicators. Thus, CFA can provide details about the setting and evaluation of SEM for this study and increase confidence in the research results.

In this hypothetical model, there are four factors, and each factor has three observed variables and a total of 12 observed variables. All factors are the cause of the observed variable, so they point to the observed variable. When the observed variable is estimated by latent variable, there are two parts: explainable variance (variation) and unexplainable variance (variation). The square of the standardized factor load is the explainable variation, and there are 12 errors (unexplained variation) from e₁ to e₁₂. Meanwhile, there is a correlation between latent variables, so there are six covariances (correlation). The four-factor hypothetical model of sport TID is shown in Figure 3, which includes four facets and a total of 12 measurement indicators. All latent variables are exogenous variables. Exogenous variables affect endogenous variables (measurement indicators), and there is a complete correlation between latent variables. For this reason, the sport TID model has 12 endogenous variables and 16 exogenous variables (including 12 measurement errors and four exogenous latent variables). Because the four dimensions of sport TID are our hypothetical structure and do not exist in the data, in order to measure the hypothetical structure, the factor load of the first question of each structure is set to 1, and its variance is mapped to each facet of sport TID.

After this study determines that the model can be estimated, the next step is to estimate the model parameters. The model in this research is analyzed by the maximum likelihood estimates (MLEs), which make the estimated results of all parameter values have the highest similarity with the actual data, and the MLE is completed by the iterative program. Table 3 shows the correlations matrix of all analyzed variables for TID, with three decimal places retained (Hoyle and Panter, 1995), for a total of 310 samples. Figure 4 shows the non-standardized estimated results and Figure 5 shows the standardized estimated results of CFA for sport TID.

Table 4 lists the main fit indexes obtained from the structural model test in detail. After comparing with the given recommended values of the fitness indicators, the indices of the fit indicators fall within the recommended values, except for the AGFI value, which is just above the recommended value of 0.9. It can be seen that the setting of the theoretical model in this study is acceptable.

Meanwhile, Table 5 shows the parameter estimates after model testing, including unstandardized estimates, standard errors, significance, and standardized estimates. As shown in Table 5, all estimated parameters are significant with standardized estimates ranging from 0.63 to 0.74 and squared multiple correlations (SMC) ranging from 0.39 to 0.54, in accordance with the requirements suggested by Fornell and Larcker (1981).

Totally, four factors were selected by principal component analysis in SPSS, as shown in Table 6. A1, A2, and A3 were clustered as one factor and named physical factor. B1, B2, and B3 were clustered as one factor and named psychological factor. C1, C2, and C3 were clustered as one factor and named coaches factor. D1, D2, and D3 were clustered as one factor and named environmental factor. In the meantime, Table 7 shows the interpretation of the total variance interpretation after rotation. Among them, the cumulative variance contribution rate of the rotation of the first four factors reaches 67.16% (Table 8). In general social investigation and research, the contribution rate of more than 60% is also in line with the requirements, which can better explain that these factors can represent the main content and information.

The aforementioned data and models showed that the path coefficient is significant. Physical, psychological, coach, and environmental factors are the four main aspects that affect the sport TID in the PE curriculum in China.

As an important organizer and executor of sport TID, coaches play an inseparable role in the process of sport TID. It is always said that swift horses are usually found but not the same as the person who has good judgments to spot them. A professional and experienced coach can fully understand and tap a student's sports potential, as well as can predict whether a student has the ability to become an excellent athlete in future. The professional quality and ability of coaches have laid a theoretical and practical foundation for them to identify sports talents. At the same time, the more familiar coaches are with students, the more they can understand students' inner world and whether students truly like sports, so as to better know which students have the qualification to become athletes. Additionally, there is a certain relationship between coaches' management consciousness and sport TID. A coach with strong management consciousness and organizational ability can have good control over students and discover, cultivate, and select talents for the country.

Under the background of the policy of “Sports Power” and “Talents Rejuvenate the Country,” policy plays a critical role in promoting the work of school PE and TID. In particular, the reform of the school PE curriculum and emphasis on the training of sports talents have been greatly promoted, which will further accelerate the identification of sports talents. A strong campus sports atmosphere can stimulate more students to participate in sports; with more participation base the more talent can be identified. Hence, for sport TID, the overall external environment is also very important, and it will affect school PE development, talent training, and guarantee from all aspects, which is inseparable with TID. In this way, with the implementation and guidance of the policy, the identification of school sports talents will be more favorable to conduct.