3.1 Statistics of questionnaire results
3.1.1 Descriptive statistics of questionnaire results
A total of 270 young sprinters participated in this study by completing online questionnaires regarding their sources of pre-competition anxiety. Questionnaires with less than 1 year of training experience, no competition experience, incomplete answers, incorrect answers to verification questions or confusion were excluded. After excluding invalid responses that did not meet criteria, a total of 124 valid questionnaires were obtained from participants including 92 male athletes and 32 female athletes, the majority of the respondents were male (74.2%),with an average age of 17.04 years and an average training period of 2.02 years. The mean score for overall anxiety level was found to be 2.64 (Table 3-1).
Table3-1 Basic information about survey respondents
Name
|
Value (mean ± standard deviation)
|
Male athletes
|
92
|
Female athletes
|
32
|
Average age
|
17.04±0.83
|
Average years of training
|
2.02±0.86
|
Total mean of anxiety levels
|
2.64±1.18
|
In this study, a correlation analysis was conducted to examine the association between the four dimensions and their respective questionnaire items. The results demonstrated a significant correlation between the four dimensional factors and the items within each dimension (p < 0.01), as outlined in Table 2-2. Upon analyzing the questionnaire responses in Table 3-2, it was observed that the findings were consistent. Specifically, the overall pre-competition anxiety index for amateur adolescent athletes averaged 2.64, suggesting a moderate level of anxiety, falling between minimal and moderate anxiety. Among the various factors, the one with the highest score was the over-emphasis on results-oriented, with an average of 3.81, indicating an anxiety level ranging from moderate to severe. Conversely, the competition environment factor received the lowest score, averaging 1.98, suggesting that athletes' anxiety levels related to this factor were between minimal and slight anxiety.
Table 3-2 Statistical table of basic information of questionnaire results
Names
|
Sample Size
|
Missing values
|
95%CI
|
M±SD
|
Competition environment impact
|
124
|
0
|
1.764
|
1.98±1.19
|
Low self-efficacy
|
2.382
|
2.64±1.45
|
Over-emphasis on results orientation
|
3.613
|
3.81±1.09
|
Influence of competitive opponent
|
1.932
|
2.15±1.22
|
Total mean of anxiety levels
|
2.432
|
2.64±1.18
|
In this study, a T-test was conducted to examine the potential impact of gender on the research outcomes (Table 3-3). The findings revealed that, across all four factors, there was no significant difference between samples of different genders (p>0.05). This indicates a consistency in the responses regardless of gender. Furthermore, an Analysis of Variance (ANOVA) test was performed on the training years (Table 3-4). Given the controlled conditions of data collection in this study, the results showed no significant difference between training years and any of the dimensions (p>0.05). This suggests a consistency in the data, thereby negating the need for further post-hoc testing.
Table 3-3 Results of gender and influencing factors analysis
|
Gender (M±SD)
|
t
|
p
|
Female(n=32)
|
Male(n=92)
|
Competition environment impact
|
2.34±1.23
|
2.08±1.20
|
1.08
|
0.28
|
Low self-efficacy
|
3.78±1.30
|
3.81±1.01
|
-0.12
|
0.90
|
Over-emphasis on results orientation
|
2.87±1.46
|
2.56±1.44
|
1.04
|
0.30
|
Influence of competitive opponent
|
2.20±1.27
|
1.90±1.16
|
1.22
|
0.22
|
*p<0.05 ** p<0.01
Table 3-4 Analysis results of training years and influencing factors
|
Training years(M±SD)
|
F
|
p
|
1.0(n=44)
|
2.0(n=4)
|
3.0(n=46)
|
Competition environment impact
|
2.01±1.23
|
1.88±1.14
|
2.47±1.20
|
2.86
|
0.06
|
Low self-efficacy
|
3.64±1.08
|
3.77±0.84
|
3.99±1.24
|
1.21
|
0.30
|
Over-emphasis on results orientation
|
2.37±1.47
|
2.43±1.38
|
3.04±1.41
|
2.99
|
0.05
|
Influence of competitive opponent
|
1.85±1.18
|
1.71±1.10
|
2.29±1.23
|
2.84
|
0.06
|
*p<0.05 ** p<0.01
3.2 Questionnaire reliability and validity test
3.2.1 Questionnaire reliability analysis
In this paper, the reliability and validity of the questionnaire were assessed using Cronbach's α coefficient. Specifically, the second section of the questionnaire was analyzed as an independent variable (X) to evaluate the rationality of the questions and determine if there was any misinterpretation by the respondents. Cronbach's α coefficient serves as a measure of internal consistency, where values above 0.9 indicate excellent reliability, 0.8-0.9 indicate good reliability, 0.7-0.8 indicate acceptable reliability, 0.6-0.7 indicate normal reliability, 0.5-0.6 indicate suboptimal reliability, and values below 0.5 suggest the need for reformatting the questionnaire. In the present study, a total of 16 questions were analyzed for reliability using Cronbach's α coefficient. The research findings revealed that the Cronbach's α coefficient for the audience and environmental impact dimension was 0.979 (Table 3-5). For the dimension of low self-efficacy, the Cronbach's α coefficient was 0.987; for over-emphasis on outcome orientation, it was 0.965; and for the influence of competitors, the Cronbach's α coefficient was 0.952. The overall reliability of the questionnaire, represented by the Cronbach's α coefficient, was 0.966. These results indicate that all reliability coefficient values of the questionnaire exceeded 0.8, demonstrating good reliability across all dimensions of the questionnaire.
Table 3-5 Results of questionnaire reliability test
Dimensions
|
Titles
|
Cronbach's α
|
Overall
Cronbach's α
|
Competition environment impact
|
A1 You feel anxious about the audience's gaze or Shouting
|
0.979
|
0.966
|
A2 The changing environment of the game often makes you feel anxious
|
A3 Playing in the right weather can help reduce your anxiety
|
A4 Playing on familiar turf can help ease your anxiety levels
|
Low self-efficacy
|
B1 Every time you compete, you don't think you're going to have a good result
|
0.987
|
B2 When you are faced with a competition, you often have the idea of withdrawing
|
B3 If you don't have a coach or a friend to encourage you before the game, you will be anxious
|
B4 Every time you get a good grade, you always think you're just lucky
|
Over-emphasis on results orientation
|
C1 You can be stressed out by focusing too much on the result before the game
|
0.965
|
C2 You focus on results and lose sight of your actual performance and progress
|
C3 The expected outcome of a competition can motivate your efforts, but it can also unnerve you
|
C4 Instead of focusing on the process, you focus on the results and rankings at the end of the competition
|
Influence of competitive opponent
|
D1 You often attribute a loss to the opponent's influence
|
0.952
|
D2 You will feel anxious before the game because of your opponent
|
D3 You prefer matches with fewer opponents
|
D4 You don't always do well against the best
|
3.2.2 Questionnaire validity test
In this study, the validity of the questionnaire was assessed using the Kaiser-Meyer-Olkin (KMO) test, as presented in Table 3-6. The test involved 16 question items organized within four distinct dimensions, serving as variables. The resulting KMO value was 0.920, exceeding the threshold of 0.8. This indicates that the research data is highly suitable for information extraction and possesses a sufficient degree of correlation among the question item variables, fulfilling the prerequisites for factor analysis. The analysis of the item-factor relationship revealed that the variance explanation rates of the four factors are 34.085%, 33.010%, 27.976%, and 1.573%, respectively. Following rotation, the cumulative variance explanation rate reached 96.644%, exceeding the benchmark of 50%. This suggests that the information content of the research items can be effectively extracted. However, the communality values for factor 4 were all below 0.4, with a variance explanation rate of 1.573%. This significant deviation from the expected relationship among the research items led to the exclusion of factor 4 from further analysis. Conversely, the relationships between factors 1, 2, and 3 aligned with the expected patterns and will be included in the subsequent analyses. The results of the Bartlett's test of sphericity indicated a significant P value of 0.000, thereby rejecting the null hypothesis and confirming the presence of correlation among the variables. This validates the effectiveness of the factor analysis and its fit to the data.
Table 3-6 Validity test results
Name
|
Factor Load factor
|
Factor 1
|
Factor 2
|
Factor 3
|
Factor 4
|
|
0.918
|
0.515
|
0.523
|
0.106
|
|
0.420
|
0.515
|
0.606
|
0.019
|
|
0.423
|
0.823
|
0.417
|
0.011
|
|
0.557
|
0.515
|
0.612
|
-0.142
|
KMO
|
0.920
|
Bartlett
|
4792.852
|
P
|
0.000***
|
*p<0.05 ** p<0.01
3.3 Results of regression analysis
The linear regression analysis conducted on the three factors of the questionnaire and the total anxiety index revealed that the variance inflation factor (VIF) of the low self-efficacy factor stood at 7.4, exceeding the threshold of 5, suggesting potential collinearity issues within the data. Consequently, ridge regression analysis was employed as a solution. This analysis aimed to quantify the influence of the independent variables on the dependent variable, specifically examining the relationship between the three factors and the total anxiety level. Ridge regression analysis serves as a methodological tool to address collinearity among independent variables in linear regression models. By introducing a k identity matrix, the regression coefficients can be accurately estimated. Prior to the analysis, the optimal k value was determined through the examination of the ridge trace plot (Figure 3-1). The selection criterion for the k value was based on the minimum value at which the standardized regression coefficients of each independent variable stabilized, with lower k values indicating reduced deviations.
In this study, low self-efficacy, the influence of the competition environment, and over-emphasis on outcome orientation were designated as independent variables, while the total anxiety level served as the dependent variable for ridge regression analysis. The resulting ridge trace plot was evaluated based on the criterion of VIF ≤ 10, with preference given to smaller k values. Ultimately, a k value of 0.01 was chosen. The model fitting outcomes are presented in Table 3-7. The table indicates that Ridge regression analysis was performed with the three factors as independent variables and the total anxiety level as the dependent variable. The model's R-squared value of 0.998 suggests that low self-efficacy, the influence of the competition environment, and over-emphasis on outcome orientation account for 99.85% of the variation in the total anxiety level, indicating a high degree of fit. However, the possibility of overfitting cannot be discounted. To assess the significance of the model, a Ridge regression ANOVA test (F test) was conducted. As evident from Table 3-7, the model successfully passes the F test (F=25985.161, p=0.000 < 0.05), indicating its meaningfulness. This suggests that a low level of self-efficacy, the influence of the competition environment, and overemphasis on at least one of the results orientations have a significant impact on the total anxiety level. The formula for the model is as follows: Overall Anxiety Level = 0.009 + 0.320 * Low Self-Efficacy + 0.411 * Competition Environment Impact + 0.257 * Overemphasis on Results Orientation.
Figure 3-1 Ridge trace map
The regression coefficient for low self-efficacy is 0.320 (t=36.112, p<0.01), indicating a significant positive impact on the overall mean value of pre-competition anxiety. This indicates that athletes with lower self-efficacy tend to experience higher levels of pre-competition anxiety. This result underscores the significance of enhancing athletes' self-efficacy. According to research result, by strengthening athletes' confidence in their abilities, anxiety can be effectively mitigated, leading to improvements in their performance. Similarly, the regression coefficient for the influence of the competition environment is 0.411 (t=49.142, p<0.01), suggesting a significant positive effect of the audience and environment on the overall average of pre-competition anxiety. This result indicates that factors such as the audience, place, and opponent have a significant and direct effect on the psychology of athletes during competitions. It may necessitates that coaches assist athletes in addressing factors that may cause anxiety or help them adapt to them, thereby improving their performance. Additionally, the regression coefficient for over-emphasis on result-orientation is 0.257 (t=34.149, p<0.01), indicating a significant positive influence on the overall mean of pre-competition anxiety. This data indicates that an over-emphasis on results can intensify athletes' anxiety levels prior to competition. The summary analysis reveals that low self-efficacy, the influence of the competition environment, and excessive emphasis on results-orientation all have significant positive effects on the total anxiety level.
Table 3-7 Results of Ridge regression analysis
|
Nonnormalized coefficient
|
Coefficient of standardization
|
t
|
p
|
VIF
|
B
|
Standard Error
|
Beta
|
Constant
|
0.009
|
0.018
|
-
|
0.471
|
0.638
|
-
|
Competition environment impact
|
0.411
|
0.008
|
0.414
|
49.142
|
0.000**
|
5.543
|
Low self-efficacy
|
0.320
|
0.009
|
0.392
|
36.112
|
0.000**
|
9.198
|
Over-emphasis on results orientation
|
0.257
|
0.008
|
0.236
|
34.149
|
0.000**
|
3.721
|
R²
|
0.998
|
Adjust R²
|
0.998
|
F
|
F (3,120)=25985.161,p =0.000
|
3.4 Results of grey association analysis
In this paper, the linear positive correlation between various items and dimensions was determined using the Pearson correlation coefficient (Table 2-1), revealing a high correlation between each item and each dimension (p < 0.01). Subsequently, a grey correlation analysis was performed to investigate the factors that contribute to pre-competition anxiety among amateur young athletes (Table 3-8). This analysis calculated the nonlinear relationships and interaction effects among the various dimensions. The grey correlation degree analysis focused on three evaluation items: audience and environment, low self-efficacy, and over-emphasis on result-orientation, along with the total anxiety level. Taking the total anxiety level as the "reference sequence," the correlation (correlation degree) between these three evaluation items and the total anxiety level was examined, providing an analytical reference based on the computed correlation degrees. During the application of grey correlation degree analysis, a resolution coefficient of 0.5 was utilized. The correlation value was computed by employing the correlation coefficient calculation formula. This calculated correlation value served as the basis for evaluation and judgment. Based on the aforementioned correlation coefficient results, weighted processing was conducted to determine the correlation degree value. This value was then utilized to evaluate and rank the three evaluation objects. The correlation value ranges from 0 to 1, where a higher value indicates a stronger correlation with the "reference sequence" (parent sequence), thereby indicating a higher evaluation. Grey correlation analysis aims to quantify the correlation degree between the feature sequence and the parent sequence. Specifically, the correlation between the influence of the competition environment and the total anxiety level is 0.679, the correlation between low self-efficacy and the total anxiety level is 0.676, and the correlation between overemphasis on result-orientation and the total anxiety level is 0.759. Among these factors, overemphasis on result-oriented exhibited the highest correlation with the total anxiety level, whereas low self-efficacy displayed the lowest correlation.
Table 3-8 Results of grey relational degree analysis
Items
|
Relevance
|
Ranking
|
Over-emphasis on results orientation
|
0.759
|
1
|
Competition environment impact
|
0.679
|
2
|
Over-emphasis on results orientation
|
0.676
|
3
|
The results indicate that excessive attention and pursuit of outcomes during the pre-competition preparation phase may significantly influence the anxiety levels of athletes. An overemphasis on results may lead to increased pressure perceived by athletes, subsequently augmenting their anxiety. Among various factors, the correlation between audience and environmental aspects is ranked second, suggesting that audience expectations, gaze, arena atmosphere, and venue comfort may impact the psychological state of participants, thus enhancing their anxiety. The lowest correlation is observed for self-efficacy, indicating that athletes are prone to experiencing pre-competition anxiety when they harbor doubts about their own abilities.