Table 3.
Multiple linear regression models predicting psychological well-being total score among coaches working with athletes without disabilities vs. athletes with disabilities
| Predictor | No disability group (n = 157) | Disability group (n = 145) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| B (SE) | β | t | p | 95% CI for B | B (SE) | β | t | p | 95% CI for B | |
| Intercept | 21.347 (6.641) | — | 3.214 | 0.002** | 8.225 to 34.470 | 27.386 (7.895) | — | 3.469 | < 0.001*** | 11.775 to 42.998 |
| Age (years) | 0.255 (0.134) | 0.175 | 1.909 | 0.058 | −0.009 to 0.519 | 0.179 (0.139) | 0.134 | 1.289 | 0.200 | −0.096 to 0.454 |
| Sex | −3.100 (1.448) | −0.154 | −2.141 | 0.034** | −5.962 to − 0.239 | −1.599 (1.663) | −0.076 | −0.962 | 0.338 | −4.887 to 1.689 |
| Education level | 2.099 (1.695) | 0.093 | 1.239 | 0.217 | −1.250 to 5.448 | 1.831 (2.217) | 0.067 | 0.826 | 0.410 | −2.553 to 6.214 |
| Relationship status | −0.704 (1.596) | −0.032 | −0.441 | 0.660 | −3.858 to 2.450 | −2.996 (2.032) | −0.118 | −1.475 | 0.143 | −7.014 to 1.021 |
| Income status | 5.111 (1.276) | 0.339 | 4.006 | < 0.001*** | 2.590 to 7.633 | 3.126 (1.347) | 0.191 | 2.321 | 0.022** | 0.463 to 5.789 |
| Years of coaching | 0.038 (0.156) | 0.022 | 0.245 | 0.807 | −0.270 to 0.347 | 0.127 (0.173) | 0.080 | 0.734 | 0.464 | −0.215 to 0.469 |
| Coaching level (rank) | 0.365 (0.875) | 0.035 | 0.417 | 0.678 | −1.365 to 2.094 | 1.617 (0.972) | 0.144 | 1.662 | 0.099 | −0.306 to 3.540 |
Model fit (No disability): R = 0.524; R²=0.275; Adj. R²=0.241; F(7,149) = 8.059; p < 0.001; Durbin–Watson = 1.325; For coaches working with non-disabled athletes, variance inflation factor (VIF) values ranged between 1.06 and 1.73, indicating no evidence of multicollinearity. Model fit (disability): R = 0.420; R²=0.176; Adj. R²=0.134; F(7,137) = 4.184; p < 0.001; Durbin–Watson = 1.330; For coaches working with athletes with disabilities, VIF values ranged between 1.03 and 1.95, also suggesting that multicollinearity was not a concern
B unstandardized regression coefficient, SE Standard error, β standardized coefficient, CI Confidence interval, Adj Adjusted, CI Confidence interval, DW Durbin–Watson, SE Standard error. Categorical predictors were dummy-coded using the following reference categories: sex = female, relationship status = in a relationship, and income level = low income, and education = Secondary school. Coaching level (grade) was entered as an ordinal variable with categories 1–5 (higher values indicate higher coaching grade)
p values: p < 0.05 **; p < 0.001 ***