EFA var-covar NPD warning

[Bryan Stiles]

Hello,

I am running an EFA within lavaan using the efa() command. This is specifying a 5-factor solution, MLR estimator, and an oblimin rotation. N = 1782 and the questionnaire is 19 items. Output is presented below for reference.

I received the following warning in lavaan: 

Warning: lavaan->lav_model_vcov():  
   The variance-covariance matrix of the estimated parameters (vcov) does not appear to be positive definite! The smallest eigenvalue (=
   -5.570291e-33) is smaller than zero. This may be a symptom that the model is not identified.

Nearly identical NPD warnings are returned when other factor solutions (3 or 4) are called. I'm having trouble isolating the issue. Considering my sample size and the corresponding number of items, I'm wondering if this is more an issue with the machine/computational precision and possibly ignorable. 

Thanks for any feedback and ideas.

This is lavaan 0.6-19 -- running exploratory factor analysis

  Estimator                                         ML
  Rotation method                      OBLIMIN OBLIQUE
  Oblimin gamma                                      0
  Rotation algorithm (rstarts)                GPA (30)
  Standardized metric                             TRUE
  Row weights                                     None

  Number of observations                          1782

Fit measures:
                    aic      bic    sabic   chisq df pvalue   cfi rmsea
  nfactors = 5 108119.5 108689.9 108359.5 358.171 86      0 0.974 0.047

Eigenvalues correlation matrix:

    ev1     ev2     ev3     ev4     ev5     ev6     ev7     ev8     ev9    ev10    ev11    ev12    ev13    ev14    ev15    ev16    ev17
  5.599   2.938   1.545   1.207   0.893   0.820   0.743   0.638   0.627   0.518   0.479   0.467   0.448   0.430   0.407   0.365   0.339
   ev18    ev19
  0.299   0.238

Standardized loadings: (* = significant at 1% level)

               f1      f2      f3      f4      f5       unique.var   communalities
T1_PAS_1R   0.357* -0.124*  0.319*  0.029*  0.083*           0.693           0.307
T1_PAS_2R   0.914*  0.019* -0.021* -0.012*  0.007            0.174           0.826
T1_PAS_3    0.025   0.217*  0.036   0.411*  0.093*           0.633           0.367
T1_PAS_4R   0.180*  0.209*  0.321*  0.205* -0.264*           0.629           0.371
T1_PAS_5R   0.061*  0.076*  0.648*  0.102* -0.104*           0.501           0.499
T1_PAS_6   -0.050   0.512*  0.091*  0.232* -0.017            0.551           0.449
T1_PAS_7R   0.335* -0.114*  0.134*  0.165*  0.055            0.804           0.196
T1_PAS_8    0.022   0.811*  0.027  -0.009   0.032            0.300           0.700
T1_PAS_9    0.017   0.814* -0.019  -0.024   0.038            0.336           0.664
T1_PAS_10  -0.028   0.220*  0.097* -0.021   0.355*           0.743           0.257
T1_PAS_11   0.031   0.085* -0.002   0.749*  0.051            0.330           0.670
T1_PAS_12   0.014  -0.052* -0.033   0.893* -0.015            0.245           0.755
T1_PAS_13R -0.004  -0.050   0.695* -0.143*  0.130*           0.479           0.521
T1_PAS_14R -0.036   0.014   0.764*  0.007   0.014            0.425           0.575
T1_PAS_15  -0.010  -0.021   0.036   0.796*  0.082*           0.346           0.654
T1_PAS_16  -0.025   0.033  -0.048   0.767* -0.037            0.408           0.592
T1_PAS_17R  0.017   0.033   0.744* -0.032   0.002            0.416           0.584
T1_PAS_18  -0.009   0.012   0.085*  0.145*  0.633*           0.496           0.504
T1_PAS_19   0.069*  0.105* -0.027   0.017   0.649*           0.503           0.497

                              f4    f3    f2    f1    f5 total
Sum of sq (obliq) loadings 3.105 2.458 2.005 1.246 1.174 9.988
Proportion of total        0.311 0.246 0.201 0.125 0.118 1.000
Proportion var             0.163 0.129 0.106 0.066 0.062 0.526
Cumulative var             0.163 0.293 0.398 0.464 0.526 0.526

Factor correlations: (* = significant at 1% level)

       f1     f2     f3     f4     f5
f1  1.000                            
f2  0.303  1.000                    
f3  0.433  0.405  1.000              
f4  0.271  0.495  0.031  1.000      
f5  0.101  0.356  0.227  0.244  1.000