Department of Industrial Psychology
Stellenbosch University
South Africa
library(lavaan)
library(psych)
library(semTools)
mydata <- read.csv(file.choose())
### Choose an estimator from "ML", "MLM", "MLR"
estimator <- "MLM"
### Type the name of the grouping variable
group <- "Country_habitat"mymodel <- '
Stress =~ GWS1 + GWS2 + GWS3 + GWS4 + GWS5 + GWS6 + GWS7 + GWS8 + GWS9
'measurement.invariance <- function(fit.config, fit.weak, fit.strong, fit.strict, fit.means) {
config.fit <- fitmeasures(fit.config, c("cfi.robust", "tli.robust", "rmsea.robust", "srmr", "BIC", "chisq.scaled", "df.scaled", "pvalue.scaled"))
weak.fit <- fitmeasures(fit.weak, c("cfi.robust", "tli.robust", "rmsea.robust", "srmr", "BIC", "chisq.scaled", "df.scaled", "pvalue.scaled"))
strong.fit <- fitmeasures(fit.strong, c("cfi.robust", "tli.robust", "rmsea.robust", "srmr", "BIC", "chisq.scaled", "df.scaled", "pvalue.scaled"))
strict.fit <- fitmeasures(fit.strict, c("cfi.robust", "tli.robust", "rmsea.robust", "srmr", "BIC", "chisq.scaled", "df.scaled", "pvalue.scaled"))
means.fit <- fitmeasures(fit.means, c("cfi.robust", "tli.robust", "rmsea.robust", "srmr", "BIC", "chisq.scaled", "df.scaled", "pvalue.scaled"))
fit.models <- rbind(config.fit, weak.fit, strong.fit, strict.fit, means.fit)
delta.weak <- config.fit - weak.fit
delta.strong <- weak.fit - strong.fit
delta.strict <- strong.fit - strict.fit
delta.means <- strict.fit - means.fit
delta.models <- rbind(delta.weak, delta.strong, delta.strict, delta.means)
fit.models
delta.models
anova.models <- anova(fit.config, fit.weak, fit.strong, fit.strict, fit.means)
compare.models <- list(anova = anova.models, model.fit = fit.models, model.delta = delta.models[ , c(1, 3, 5)])
compare.models
}Users indicate the parameter equality constraints that should be
lifted with the group.partial argument. These constraints
can be identified by means of Lagrange multiplier tests.
How to indicate a parameter
Example 1 - if it is desired that the factor loading
of item i1 should be freely estimated across groups, one would add the
following argument to the weak invariance model and all
subsequent invariance models:
group.partial = c("F1 =~ i1")
Example 2 - if it is desired that the intercept of
item i1 should be freely estimated across group, one would add the
following argument to the strong invariance model and all
subsequent models: group.partial = c("i1 ~ 1")
Example 3 - if it is desired that the residual
variance of item i1 should be freely estimated across group,
one would add the following argument to the strict invariance
model and all subsequent models:
group.partial = c("i1 ~~ i1")
Example 4: if it is desired that the factor loading of item i1 and
the intercept of item i2 and the residual variance of i3 should be
freely estimated one would add the following argument for the weak
invariance model: group.partial = c("F1 =~ i1"). Then one
would add the following argument for the strong invariance model:
group.partial = c("F1 =~ i1", "i1 ~ 1"). Finally, one would
add the following argument to the strict invariance model:
group.partial = c("F1 =~ i1", "i2 ~ 1", "i3 ~~ i3")
##### Configural invariance
fit.config <- cfa(mymodel, data = mydata, group = group,
estimator = estimator)
#summary(fit.config)
####### Weak invariance
fit.weak <- cfa(mymodel, data = mydata, group = group,
estimator = estimator,
group.equal ="loadings",
group.partial = NULL)
#summary(fit.weak)
####### Strong invariance
fit.strong <- cfa(mymodel, data = mydata, group = group,
estimator = estimator,
group.equal = c("loadings", "intercepts"),
group.partial = c("GWS7 ~ 1", "GWS8 ~ 1", "GWS6 ~ 1"))
#summary(fit.strong)
####### Strict invariance
fit.strict <- cfa(mymodel, data = mydata, group = group,
estimator = estimator,
group.equal = c("loadings", "intercepts", "residuals"),
group.partial = c("GWS7 ~ 1", "GWS8 ~ 1", "GWS6 ~ 1"))
#summary(fit.strict)
####### Equal factor means
fit.means <- cfa(mymodel, data = mydata, group = group,
estimator = estimator,
group.equal = c("loadings", "intercepts", "residuals", "means"),
group.partial = c("GWS7 ~ 1", "GWS8 ~ 1", "GWS6 ~ 1"))
#summary(fit.means)
###################################################################
# Run the measurement invariance function in the MeasInvar.R file #
###################################################################
measurement.invariance(fit.config, fit.weak, fit.strong, fit.strict, fit.means)## $anova
##
## Scaled Chi-Squared Difference Test (method = "satorra.bentler.2001")
##
## lavaan NOTE:
## The "Chisq" column contains standard test statistics, not the
## robust test that should be reported per model. A robust difference
## test is a function of two standard (not robust) statistics.
##
## Df AIC BIC Chisq Chisq diff Df diff Pr(>Chisq)
## fit.config 54 29471 29754 842.56
## fit.weak 62 29464 29704 850.86 8.292 8 0.405518
## fit.strong 67 29468 29682 865.40 15.473 5 0.008521 **
## fit.strict 76 29512 29679 927.14 54.882 9 1.283e-08 ***
## fit.means 77 29558 29720 975.61 70.185 1 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## $model.fit
## cfi.robust tli.robust rmsea.robust srmr bic chisq.scaled
## config.fit 0.8807417 0.8409889 0.1435070 0.05366521 29753.55 591.7743
## weak.fit 0.8806962 0.8614536 0.1339543 0.05570739 29704.03 621.4061
## strong.fit 0.8791627 0.8701450 0.1296846 0.05698057 29682.44 647.1674
## strict.fit 0.8711234 0.8779064 0.1257493 0.06289092 29679.13 706.6205
## means.fit 0.8636819 0.8725337 0.1284863 0.08676744 29720.37 748.1621
## df.scaled pvalue.scaled
## config.fit 54 0
## weak.fit 62 0
## strong.fit 67 0
## strict.fit 76 0
## means.fit 77 0
##
## $model.delta
## cfi.robust rmsea.robust bic
## delta.weak 4.546861e-05 0.009552626 49.520277
## delta.strong 1.533467e-03 0.004269700 21.592692
## delta.strict 8.039326e-03 0.003935299 3.309571
## delta.means 7.441490e-03 -0.002736971 -41.239457Type here the name of the fitted model that you want to inspect. This will typically be the best fitting model that was identified in the previous step.
summary(fit.strong, fit.measures = TRUE, standardized = TRUE)## lavaan 0.6.15 ended normally after 38 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 55
## Number of equality constraints 14
##
## Number of observations per group:
## 2 793
## 1 584
##
## Model Test User Model:
## Standard Scaled
## Test Statistic 865.405 647.167
## Degrees of freedom 67 67
## P-value (Chi-square) 0.000 0.000
## Scaling correction factor 1.337
## Satorra-Bentler correction
## Test statistic for each group:
## 2 431.144 322.418
## 1 434.261 324.749
##
## Model Test Baseline Model:
##
## Test statistic 6523.400 4555.388
## Degrees of freedom 72 72
## P-value 0.000 0.000
## Scaling correction factor 1.432
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.876 0.871
## Tucker-Lewis Index (TLI) 0.867 0.861
##
## Robust Comparative Fit Index (CFI) 0.879
## Robust Tucker-Lewis Index (TLI) 0.870
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -14693.053 -14693.053
## Loglikelihood unrestricted model (H1) -14260.351 -14260.351
##
## Akaike (AIC) 29468.107 29468.107
## Bayesian (BIC) 29682.441 29682.441
## Sample-size adjusted Bayesian (SABIC) 29552.200 29552.200
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.132 0.112
## 90 Percent confidence interval - lower 0.124 0.105
## 90 Percent confidence interval - upper 0.139 0.119
## P-value H_0: RMSEA <= 0.050 0.000 0.000
## P-value H_0: RMSEA >= 0.080 1.000 1.000
##
## Robust RMSEA 0.130
## 90 Percent confidence interval - lower 0.121
## 90 Percent confidence interval - upper 0.139
## P-value H_0: Robust RMSEA <= 0.050 0.000
## P-value H_0: Robust RMSEA >= 0.080 1.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.057 0.057
##
## Parameter Estimates:
##
## Standard errors Robust.sem
## Information Expected
## Information saturated (h1) model Structured
##
##
## Group 1 [2]:
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## Stress =~
## GWS1 1.000 0.854 0.786
## GWS2 (.p2.) 1.116 0.025 44.507 0.000 0.953 0.829
## GWS3 (.p3.) 0.970 0.033 29.744 0.000 0.828 0.746
## GWS4 (.p4.) 0.902 0.039 23.414 0.000 0.770 0.682
## GWS5 (.p5.) 0.774 0.031 24.805 0.000 0.661 0.669
## GWS6 (.p6.) 0.851 0.030 28.231 0.000 0.727 0.772
## GWS7 (.p7.) 0.874 0.034 25.681 0.000 0.746 0.715
## GWS8 (.p8.) 0.870 0.033 26.126 0.000 0.743 0.754
## GWS9 (.p9.) 0.773 0.033 23.252 0.000 0.660 0.645
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .GWS1 (.20.) 2.736 0.036 76.917 0.000 2.736 2.517
## .GWS2 (.21.) 2.295 0.039 58.675 0.000 2.295 1.997
## .GWS3 (.22.) 2.356 0.036 65.927 0.000 2.356 2.121
## .GWS4 (.23.) 2.367 0.037 64.595 0.000 2.367 2.099
## .GWS5 (.24.) 2.137 0.031 68.058 0.000 2.137 2.162
## .GWS6 2.135 0.034 63.692 0.000 2.135 2.267
## .GWS7 2.569 0.037 69.041 0.000 2.569 2.462
## .GWS8 1.963 0.035 55.680 0.000 1.963 1.993
## .GWS9 (.28.) 2.098 0.033 64.435 0.000 2.098 2.050
## Stress 0.000 0.000 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .GWS1 0.452 0.026 17.258 0.000 0.452 0.383
## .GWS2 0.414 0.029 14.458 0.000 0.414 0.313
## .GWS3 0.548 0.044 12.355 0.000 0.548 0.444
## .GWS4 0.680 0.040 17.077 0.000 0.680 0.534
## .GWS5 0.540 0.031 17.474 0.000 0.540 0.552
## .GWS6 0.358 0.024 15.036 0.000 0.358 0.404
## .GWS7 0.532 0.033 16.161 0.000 0.532 0.488
## .GWS8 0.419 0.030 13.959 0.000 0.419 0.431
## .GWS9 0.612 0.038 16.035 0.000 0.612 0.584
## Stress 0.729 0.050 14.706 0.000 1.000 1.000
##
##
## Group 2 [1]:
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## Stress =~
## GWS1 1.000 0.694 0.695
## GWS2 (.p2.) 1.116 0.025 44.507 0.000 0.775 0.764
## GWS3 (.p3.) 0.970 0.033 29.744 0.000 0.674 0.703
## GWS4 (.p4.) 0.902 0.039 23.414 0.000 0.626 0.626
## GWS5 (.p5.) 0.774 0.031 24.805 0.000 0.538 0.657
## GWS6 (.p6.) 0.851 0.030 28.231 0.000 0.591 0.710
## GWS7 (.p7.) 0.874 0.034 25.681 0.000 0.607 0.578
## GWS8 (.p8.) 0.870 0.033 26.126 0.000 0.604 0.696
## GWS9 (.p9.) 0.773 0.033 23.252 0.000 0.537 0.639
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .GWS1 (.20.) 2.736 0.036 76.917 0.000 2.736 2.737
## .GWS2 (.21.) 2.295 0.039 58.675 0.000 2.295 2.263
## .GWS3 (.22.) 2.356 0.036 65.927 0.000 2.356 2.458
## .GWS4 (.23.) 2.367 0.037 64.595 0.000 2.367 2.369
## .GWS5 (.24.) 2.137 0.031 68.058 0.000 2.137 2.612
## .GWS6 2.271 0.040 56.742 0.000 2.271 2.730
## .GWS7 3.077 0.047 65.183 0.000 3.077 2.930
## .GWS8 2.253 0.042 53.773 0.000 2.253 2.596
## .GWS9 (.28.) 2.098 0.033 64.435 0.000 2.098 2.496
## Stress -0.317 0.045 -7.022 0.000 -0.456 -0.456
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .GWS1 0.517 0.034 15.184 0.000 0.517 0.517
## .GWS2 0.429 0.036 11.825 0.000 0.429 0.417
## .GWS3 0.465 0.034 13.740 0.000 0.465 0.506
## .GWS4 0.607 0.039 15.613 0.000 0.607 0.608
## .GWS5 0.380 0.026 14.846 0.000 0.380 0.568
## .GWS6 0.343 0.025 13.457 0.000 0.343 0.495
## .GWS7 0.734 0.045 16.387 0.000 0.734 0.666
## .GWS8 0.388 0.033 11.874 0.000 0.388 0.515
## .GWS9 0.418 0.033 12.751 0.000 0.418 0.592
## Stress 0.482 0.041 11.712 0.000 1.000 1.000Inspect the chi-square and p-values of the Lagrange multiplier tests. These tests are used to identify parameter equality constraints that contribute toward model misfit. Identify the parameter equality constraint with the highest chi-square. Fit a new series of measurement invariance models where this constraint is lifted, which yields a model of partial measurement invariance. Note whether this leads to an improvement of fit. Repeat the process by conducting new Lagrange multiplier tests until satisfactory fit is obtained.
lavTestScore(fit.strong)## Warning in lavTestScore(fit.strong): lavaan WARNING: se is not `standard'; not
## implemented yet; falling back to ordinary score test## $test
##
## total score test:
##
## test X2 df p.value
## 1 score 21.859 14 0.082
##
## $uni
##
## univariate score tests:
##
## lhs op rhs X2 df p.value
## 1 .p2. == .p31. 0.016 1 0.899
## 2 .p3. == .p32. 4.950 1 0.026
## 3 .p4. == .p33. 1.719 1 0.190
## 4 .p5. == .p34. 0.001 1 0.979
## 5 .p6. == .p35. 0.077 1 0.781
## 6 .p7. == .p36. 0.271 1 0.603
## 7 .p8. == .p37. 0.764 1 0.382
## 8 .p9. == .p38. 0.340 1 0.560
## 9 .p20. == .p49. 0.015 1 0.903
## 10 .p21. == .p50. 0.088 1 0.767
## 11 .p22. == .p51. 0.047 1 0.827
## 12 .p23. == .p52. 9.033 1 0.003
## 13 .p24. == .p53. 7.267 1 0.007
## 14 .p28. == .p57. 0.073 1 0.787parTable(fit.strong)## id lhs op rhs user block group free ustart exo label plabel start
## 1 1 Stress =~ GWS1 1 1 1 0 1 0 .p1. 1.000
## 2 2 Stress =~ GWS2 1 1 1 1 NA 0 .p2. .p2. 1.163
## 3 3 Stress =~ GWS3 1 1 1 2 NA 0 .p3. .p3. 0.889
## 4 4 Stress =~ GWS4 1 1 1 3 NA 0 .p4. .p4. 0.841
## 5 5 Stress =~ GWS5 1 1 1 4 NA 0 .p5. .p5. 0.667
## 6 6 Stress =~ GWS6 1 1 1 5 NA 0 .p6. .p6. 0.757
## 7 7 Stress =~ GWS7 1 1 1 6 NA 0 .p7. .p7. 0.756
## 8 8 Stress =~ GWS8 1 1 1 7 NA 0 .p8. .p8. 0.804
## 9 9 Stress =~ GWS9 1 1 1 8 NA 0 .p9. .p9. 0.725
## 10 10 GWS1 ~~ GWS1 0 1 1 9 NA 0 .p10. 0.581
## 11 11 GWS2 ~~ GWS2 0 1 1 10 NA 0 .p11. 0.662
## 12 12 GWS3 ~~ GWS3 0 1 1 11 NA 0 .p12. 0.591
## 13 13 GWS4 ~~ GWS4 0 1 1 12 NA 0 .p13. 0.651
## 14 14 GWS5 ~~ GWS5 0 1 1 13 NA 0 .p14. 0.487
## 15 15 GWS6 ~~ GWS6 0 1 1 14 NA 0 .p15. 0.445
## 16 16 GWS7 ~~ GWS7 0 1 1 15 NA 0 .p16. 0.549
## 17 17 GWS8 ~~ GWS8 0 1 1 16 NA 0 .p17. 0.493
## 18 18 GWS9 ~~ GWS9 0 1 1 17 NA 0 .p18. 0.530
## 19 19 Stress ~~ Stress 0 1 1 18 NA 0 .p19. 0.050
## 20 20 GWS1 ~1 0 1 1 19 NA 0 .p20. .p20. 2.734
## 21 21 GWS2 ~1 0 1 1 20 NA 0 .p21. .p21. 2.299
## 22 22 GWS3 ~1 0 1 1 21 NA 0 .p22. .p22. 2.359
## 23 23 GWS4 ~1 0 1 1 22 NA 0 .p23. .p23. 2.313
## 24 24 GWS5 ~1 0 1 1 23 NA 0 .p24. .p24. 2.183
## 25 25 GWS6 ~1 0 1 1 24 NA 0 .p25. 2.135
## 26 26 GWS7 ~1 0 1 1 25 NA 0 .p26. 2.569
## 27 27 GWS8 ~1 0 1 1 26 NA 0 .p27. 1.963
## 28 28 GWS9 ~1 0 1 1 27 NA 0 .p28. .p28. 2.093
## 29 29 Stress ~1 0 1 1 0 0 0 .p29. 0.000
## 30 30 Stress =~ GWS1 1 2 2 0 1 0 .p30. 1.000
## 31 31 Stress =~ GWS2 1 2 2 28 NA 0 .p2. .p31. 1.064
## 32 32 Stress =~ GWS3 1 2 2 29 NA 0 .p3. .p32. 0.812
## 33 33 Stress =~ GWS4 1 2 2 30 NA 0 .p4. .p33. 0.553
## 34 34 Stress =~ GWS5 1 2 2 31 NA 0 .p5. .p34. 0.459
## 35 35 Stress =~ GWS6 1 2 2 32 NA 0 .p6. .p35. 0.555
## 36 36 Stress =~ GWS7 1 2 2 33 NA 0 .p7. .p36. 0.490
## 37 37 Stress =~ GWS8 1 2 2 34 NA 0 .p8. .p37. 0.566
## 38 38 Stress =~ GWS9 1 2 2 35 NA 0 .p9. .p38. 0.529
## 39 39 GWS1 ~~ GWS1 0 2 2 36 NA 0 .p39. 0.516
## 40 40 GWS2 ~~ GWS2 0 2 2 37 NA 0 .p40. 0.511
## 41 41 GWS3 ~~ GWS3 0 2 2 38 NA 0 .p41. 0.489
## 42 42 GWS4 ~~ GWS4 0 2 2 39 NA 0 .p42. 0.496
## 43 43 GWS5 ~~ GWS5 0 2 2 40 NA 0 .p43. 0.323
## 44 44 GWS6 ~~ GWS6 0 2 2 41 NA 0 .p44. 0.343
## 45 45 GWS7 ~~ GWS7 0 2 2 42 NA 0 .p45. 0.542
## 46 46 GWS8 ~~ GWS8 0 2 2 43 NA 0 .p46. 0.367
## 47 47 GWS9 ~~ GWS9 0 2 2 44 NA 0 .p47. 0.349
## 48 48 Stress ~~ Stress 0 2 2 45 NA 0 .p48. 0.050
## 49 49 GWS1 ~1 0 2 2 46 NA 0 .p20. .p49. 2.421
## 50 50 GWS2 ~1 0 2 2 47 NA 0 .p21. .p50. 1.937
## 51 51 GWS3 ~1 0 2 2 48 NA 0 .p22. .p51. 2.045
## 52 52 GWS4 ~1 0 2 2 49 NA 0 .p23. .p52. 2.147
## 53 53 GWS5 ~1 0 2 2 50 NA 0 .p24. .p53. 1.848
## 54 54 GWS6 ~1 0 2 2 51 NA 0 .p54. 2.002
## 55 55 GWS7 ~1 0 2 2 52 NA 0 .p55. 2.800
## 56 56 GWS8 ~1 0 2 2 53 NA 0 .p56. 1.978
## 57 57 GWS9 ~1 0 2 2 54 NA 0 .p28. .p57. 1.858
## 58 58 Stress ~1 0 2 2 55 NA 0 .p58. 0.000
## 59 59 .p2. == .p31. 2 0 0 0 NA 0 0.000
## 60 60 .p3. == .p32. 2 0 0 0 NA 0 0.000
## 61 61 .p4. == .p33. 2 0 0 0 NA 0 0.000
## 62 62 .p5. == .p34. 2 0 0 0 NA 0 0.000
## 63 63 .p6. == .p35. 2 0 0 0 NA 0 0.000
## 64 64 .p7. == .p36. 2 0 0 0 NA 0 0.000
## 65 65 .p8. == .p37. 2 0 0 0 NA 0 0.000
## 66 66 .p9. == .p38. 2 0 0 0 NA 0 0.000
## 67 67 .p20. == .p49. 2 0 0 0 NA 0 0.000
## 68 68 .p21. == .p50. 2 0 0 0 NA 0 0.000
## 69 69 .p22. == .p51. 2 0 0 0 NA 0 0.000
## 70 70 .p23. == .p52. 2 0 0 0 NA 0 0.000
## 71 71 .p24. == .p53. 2 0 0 0 NA 0 0.000
## 72 72 .p28. == .p57. 2 0 0 0 NA 0 0.000
## est se
## 1 1.000 0.000
## 2 1.116 0.025
## 3 0.970 0.033
## 4 0.902 0.039
## 5 0.774 0.031
## 6 0.851 0.030
## 7 0.874 0.034
## 8 0.870 0.033
## 9 0.773 0.033
## 10 0.452 0.026
## 11 0.414 0.029
## 12 0.548 0.044
## 13 0.680 0.040
## 14 0.540 0.031
## 15 0.358 0.024
## 16 0.532 0.033
## 17 0.419 0.030
## 18 0.612 0.038
## 19 0.729 0.050
## 20 2.736 0.036
## 21 2.295 0.039
## 22 2.356 0.036
## 23 2.367 0.037
## 24 2.137 0.031
## 25 2.135 0.034
## 26 2.569 0.037
## 27 1.963 0.035
## 28 2.098 0.033
## 29 0.000 0.000
## 30 1.000 0.000
## 31 1.116 0.025
## 32 0.970 0.033
## 33 0.902 0.039
## 34 0.774 0.031
## 35 0.851 0.030
## 36 0.874 0.034
## 37 0.870 0.033
## 38 0.773 0.033
## 39 0.517 0.034
## 40 0.429 0.036
## 41 0.465 0.034
## 42 0.607 0.039
## 43 0.380 0.026
## 44 0.343 0.025
## 45 0.734 0.045
## 46 0.388 0.033
## 47 0.418 0.033
## 48 0.482 0.041
## 49 2.736 0.036
## 50 2.295 0.039
## 51 2.356 0.036
## 52 2.367 0.037
## 53 2.137 0.031
## 54 2.271 0.040
## 55 3.077 0.047
## 56 2.253 0.042
## 57 2.098 0.033
## 58 -0.317 0.045
## 59 0.000 0.000
## 60 0.000 0.000
## 61 0.000 0.000
## 62 0.000 0.000
## 63 0.000 0.000
## 64 0.000 0.000
## 65 0.000 0.000
## 66 0.000 0.000
## 67 0.000 0.000
## 68 0.000 0.000
## 69 0.000 0.000
## 70 0.000 0.000
## 71 0.000 0.000
## 72 0.000 0.000de Bruin, G. P. & Taylor, N. (2005). Development of the Sources of Work Stress Inventory. South African Journal of Psychology, 35, 748-765.
de Bruin, G. P. (2006). Dimensionality of the General Work Stress Scale. SA Journal of Industrial Psychology, 32(4), 68-75.