Skip to main content

Table 4 Fit indices of the 5-factor BSEM models with different priors for the FFMQ-SF

From: Psychometric properties of the Chinese version of Five Facet Mindfulness Questionnaire—short form in cancer patients: a Bayesian structural equation modeling approach

Model Prior specification # pD pppp 2.5% PPL 97.5% PPL PPP DIC RMSEA CFI
1 Uninformative 70 72 100.4 204.3 .000 6403 0.091 0.84
Informative priors on cross-loadings
2 var = 0.01 150 100 .12 12.5 132.7 .008 6352 0.072 0.92
3 var = 0.02 150 108 .61 − 3.0 117.5 .030 6343 0.066 0.93
4 var = 0.04 150 115 .94 − 8.0 115.7 .044 6347 0.066 0.94
5 Revised Model 3 150 107 .70 − 4.8 111.7 .037 6339 0.065 0.94
Informative priors on residual covariance
6 d = 10 340 211 .44 − 75.5 50.9 .662 6377 0.010 1.00
7 d = 20 340 209 .20 − 78.9 42.2 .726 6369 0.000 1.00
8 d = 30 340 214 .06 − 73.6 50.0 .644 6381 0.020 0.99
Informative priors on cross-loadings with one residual covariance
9 Revised Model 5 151 108 .75 − 14.9 97.6 .073 6329 0.060 0.95
  1. N = 127; # = number of free parameters; pD = estimated number of parameters; pppp = prior posterior predictive p value; PPL = posterior predictive limit; PPP = posterior predictive p value; DIC = Deviance information criterion; RMSEA = Root mean square error of approximation; CFI = comparative fit index; d = degree of freedom