Demographic data of respondents
Table 1 presents the gender distribution of the survey respondents. Of the total, 149 respondents (39.20%) were male and 232 (60.80%) were female, indicating a significant disparity between the number of men and women in the sample, with a notably higher representation of females. In terms of age, the majority of respondents, 304 (80.00%), were aged between 21 and 22 years. The next largest age group, those younger than 21, comprised 38 respondents (10.00%). There were 16 respondents (4.00%) aged between 23 and 25 years, and 23 respondents (6.00%) were older than 25 years.
Further demographic analysis is provided concerning the educational institutions of the respondents. As detailed in the table, 113 students (35.00%) attend University A, an equal number attend University B, and 115 students (30.00%) are enrolled at University C. This distribution is summarized in Table 1.
Finding based on objective one: to investigate the factors that cause EFL learners’ speaking anxiety in their higher education institutions
The internal consistency and construct reliability of the questionnaire were tested using SPSS. Sauder et al. (2003) define reliability as the degree to which data collection methods agree with the results obtained. The goal of data reliability analysis is to determine whether or not it is trustworthy. Reliability testing determines the degree of consistency among different variables. The Cronbach's Alpha Score in this study was greater than 0.700, which is acceptable.
Table 2 examines Cronbach's alpha coefficients, which indicate the reliability of different constructs. Lack of knowledge (6 items) scored 0.843; lack of self-confidence (7 items) scored 0.757; lack of self-esteem (7 items) scored 0.751; low-class interaction (5 items) scored 0.773; lack of practice (5 items) scored 0.797; poor listening comprehension (4 items) scored 0.747, and English-speaking anxiety (5 items) scored 0.726. Table 2 shows a summary.
The Exploratory Factor Analysis (EFA) was used on the Influence Factor items to classify fundamental factors that represent Influence Factor characteristics. A statistical procedure known as factor analysis collects a set of items to construct appropriate factors (Macaro, 2003). Before analyzing the variables, the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy and Bartlett's Test of Sphericity were used.
The KMO value of.834 and χ2 (171) = 1721.624 (p < 0.000) confirmed the validity of the factor analysis. The EFA yielded a six-factor solution, explaining 49.52% of the total variance. The eigenvalues were greater than one: LK = 5.65, LC = 4.889, LE = 2.989, LI = 2.05, LP = 1.74, and PL = 1.63. Each item was assigned to its original factor, which typically contained four to seven items and had factor loadings ranging from 0.59 to 0.70. Cronbach's alpha values for all factors were satisfactory (α = 0.84 for LK, α = 0.75 for LC, α = 0.75 for LE, α = 0.77 for LI, α = 0.79 for LP, and α = 0.74 for PL). According to the findings of research question one, six factors contribute to anxiety in English-speaking individuals: lack of knowledge, lack of self-confidence, lack of self-esteem, lack of class interaction, lack of practice, and poor listening comprehension.
Finding based on objective two: to examine the influences of anxiety factors on students’ English-speaking anxiety
In response to research question two, our analysis shows that the independent variable has a significant influence on the dependent variable, accounting for approximately 64% of the variance (R² = 0.64, p < 0.001).
Table 3 details the various factors that contribute to ESA. LK makes a significant contribution (β = 0.31, p < 0.001), while LC has a minor impact (β = 0.05, p < 0.05). LE has a significant effect on ESA (β = 0.25, p < 0.001), LI (β = 0.32, p < 0.001), and PL (β = 0.31, p < 0.001). LP has a negative impact (β = -0.172, p < 0.05). Overall, the findings support Hypotheses H1, H3, and H4. In contrast, the empirical evidence did not support Hypotheses H2 and H5.