Skip to main content

Table 3 Overall fitted results of structural equation modelling using the maximum likelihood method with missing values1,2 (n = 21,439 expatriates)

From: Loneliness and trust issues reshape mental stress of expatriates during early COVID-19: a structural equation modelling approach

Fit statistics

Estimates

 

Reference5

Likelihood ratio

 Chi-square MS (177)

7517.49

  

 p > chi2

0.000

  

 Chi-square/df

47.18

 

 < 3.00

 Chi-square BS (210)

197,829.15

  

Population error

   

 RMSEA

0.04

 

 < ≈0.06

 90% CI

0.043–0.045

  

 Pclose

1.000

P(RMSEA) <  = 0.05

 

Information criterion

 AIC3

1,391,196.27

  

 BIC3

1,391,794.24

  

Baseline comparison

 CFI

0.96

 

 > 0.95

 TLI

0.96

 

 > 0.95

 GFI

0.96

  

Size of residuals

 CD

0.972

  

 SRMR4

Not reported

  
  1. MS- model vs saturated; BS- baseline vs saturated; RMSEA- root mean squared error of approximation; CI- confidence interval; CFI- Comparative fit index; TLI- Tucker-Lewis index; GFI- goodness of fit index; CD- coefficient of determination and SRMR- standardized root-mean-square residuals
  2. 1–Method: mlmv; Log-likelihood: –695,523.13
  3. 2Endogenous latent variable: PSS and exogenous latent variables: PLS, IPT and IT
  4. 3Akaike's and Bayesian information criteria (AIC and BIC) were ignored (Huang, 2017)
  5. 4SRMR is not reported due to missing values
  6. 5Reference values were provided in the third column (Hu & Bentler, 1999; Kline, 2015)
Back to article page

BMC Psychology

ISSN: 2050-7283

Contact us