Thank you for your suggestion. I just constrained the model like so
lavaan 0.6-7 ended normally after 10 iterations
Estimator ML
Optimization method NLMINB
Number of free parameters 6
Number of equality constraints 2
Number of observations 151
Number of missing patterns 3
Model Test User Model:
Standard Robust
Test Statistic 99.330 NA
Degrees of freedom 1 1
P-value (Chi-square) 0.000 NA
Scaling correction factor NA
Yuan-Bentler correction (Mplus variant)
Model Test Baseline Model:
Test statistic 159.268 116.971
Degrees of freedom 1 1
P-value 0.000 0.000
Scaling correction factor 1.362
User Model versus Baseline Model:
Comparative Fit Index (CFI) 0.379 NA
Tucker-Lewis Index (TLI) 0.379 NA
Robust Comparative Fit Index (CFI) NA
Robust Tucker-Lewis Index (TLI) NA
Loglikelihood and Information Criteria:
Loglikelihood user model (H0) -306.734 -306.734
Scaling correction factor 0.935
for the MLR correction
Loglikelihood unrestricted model (H1) -257.069 -257.069
Scaling correction factor 1.116
for the MLR correction
Akaike (AIC) 621.469 621.469
Bayesian (BIC) 633.538 633.538
Sample-size adjusted Bayesian (BIC) 620.878 620.878
Root Mean Square Error of Approximation:
RMSEA 0.807 0.000
90 Percent confidence interval - lower 0.677 0.000
90 Percent confidence interval - upper 0.945 0.000
P-value RMSEA <= 0.05 0.000 NaN
Robust RMSEA NA
90 Percent confidence interval - lower 0.000
90 Percent confidence interval - upper 0.000
Standardized Root Mean Square Residual:
SRMR 0.385 0.385
Parameter Estimates:
Standard errors Sandwich
Information bread Observed
Observed information based on Hessian
and I had the following error message for lavInspect: