Table 3

Model fit indices for the SPSI

Scaled χ2 (df)CFITLIRMSEA (90% CI)AIC/BICSRMR
Athletes (n=168)
 Unidimensional model(44) 238.81, p<0.0010.5920.4900.162 (0.142 to 0.183), p<0.0014587.93/4691.020.181
 Second order model(41) 57.39, p=0.0500.9660.9540.049 (0.007 to 0.077), p=0.5004241.31/4353.780.081
 Bifactor model00 (33) 36.68, p=0.3020.9920.9870.026 (0.000 to 0.064), p=0.8214222.76/4360.210.054
Coaches/HPSS (n=238)
 Unidimensional model(44) 340.37, p<0.0010.5840.4810.168 (0.152 to 0.185), p<0.0015843.82/5958.400.148
 Second order model(41) 103.79, p<0.0010.9120.8820.080 (0.061 to 0.100), p=0.0065492.37/5617.370.077
 Bifactor model(34) 53.63, p=0.0170.9720.9550.049 (0.021 to 0.073), p=0.4925434.25/5583.560.048
Combined (n=408)
 Unidimensional model(44) 585.09, p<0010.5740.4670.174 (0.162 to 0.187), p<0.00110541.39/10 673.600.166
 Second order model(41) 123.46, p<0.0010.9350.9130.070 (0.056 to .085), p=0.0109798.93/9943.160.077
 Bifactor model(34) 30.35, p=0.0040.9790.9670.044 (0.025 to 0.061), p=0.7019718.03/9890.300.049
  • AIC/BIC, Akaike information criterion / Bayesian Information Criteria; ARPB, Average relative parameter bias; CFI, Comparative fit index; HPSS, high-performance support staff; RMSEA, Root mean square error of approximation; SPSI, Sport Psychological Safety Inventory; SRMR, Standardized Root Mean Squared Residual; TLI, Tucker-Lewis index.