Output Guide
This page guides you through the interpretation of Mplus output. It shows key, but not all, sections of the output. Hover your mouse over either output labels or numerical entries and an explanation or interpretation appears as a 'tool tip.' If a tip does not appear, then there is none for the entry. There is a slight delay when you position your mouse over the entry and tip appearance. I use the example study described on the 'Syntax I' tab of this website. If you have not read the description of it, click on the 'Syntax' tab and do so.
Double
Global Fit Indices
Chi-Square Test of Model Fit
Value 4.728*
Degrees of Freedom 7
P-Value 0.6932
Scaling Correction Factor 0.9554
for MLR
RMSEA (Root Mean Square Error Of Approximation)
Estimate 0.000
90 Percent C.I. 0.000 0.042
Probability RMSEA <= .05 0.977
CFI/TLI
CFI 1.000
TLI 1.000
Chi-Square Test of Model Fit for the Baseline Model
Value 2093.586
Degrees of Freedom 15
P-Value 0.0000
SRMR (Standardized Root Mean Square Residual)
Value 0.013
Value 4.728*
Degrees of Freedom 7
P-Value 0.6932
Scaling Correction Factor 0.9554
for MLR
RMSEA (Root Mean Square Error Of Approximation)
Estimate 0.000
90 Percent C.I. 0.000 0.042
Probability RMSEA <= .05 0.977
CFI/TLI
CFI 1.000
TLI 1.000
Chi-Square Test of Model Fit for the Baseline Model
Value 2093.586
Degrees of Freedom 15
P-Value 0.0000
SRMR (Standardized Root Mean Square Residual)
Value 0.013
Modification Indices
MODEL MODIFICATION INDICES
Minimum M.I. value for printing the modification index 4.000
M.I. E.P.C. Std E.P.C. StdYX E.P.C.
No modification indices above the minimum value.
Minimum M.I. value for printing the modification index 4.000
M.I. E.P.C. Std E.P.C. StdYX E.P.C.
No modification indices above the minimum value.
Standardized Residuals (z-scores) for Covariances
WLOSS SR1 SR2 EFF1 EFF2
________ ________ ________ ________ ________
WLOSS -0.827
SR1 -0.939 0.000
SR2 -0.556 999.000 0.000
EFF1 -1.134 -1.110 -1.023 0.000
EFF2 -0.151 -0.447 -0.409 999.000 0.000
TREAT -0.852 1.182 -0.329 0.709 -0.586
Standardized Residuals (z-scores) for Covariances
TREAT
________
TREAT 0.000
WLOSS SR1 SR2 EFF1 EFF2
________ ________ ________ ________ ________
WLOSS -0.827
SR1 -0.939 0.000
SR2 -0.556 999.000 0.000
EFF1 -1.134 -1.110 -1.023 0.000
EFF2 -0.151 -0.447 -0.409 999.000 0.000
TREAT -0.852 1.182 -0.329 0.709 -0.586
Standardized Residuals (z-scores) for Covariances
TREAT
________
TREAT 0.000
Standardized Residuals
Residuals for Correlations
WLOSS SR1 SR2 EFF1 EFF2
________ ________ ________ ________ ________
WLOSS 0.000
SR1 -0.010 0.000
SR2 -0.004 0.000 0.000
EFF1 -0.018 -0.038 -0.037 0.000
EFF2 0.001 -0.015 -0.015 0.000 0.000
TREAT -0.010 0.006 -0.004 0.005 -0.006
Residuals for Correlations
TREAT
________
TREAT 0.000
WLOSS SR1 SR2 EFF1 EFF2
________ ________ ________ ________ ________
WLOSS 0.000
SR1 -0.010 0.000
SR2 -0.004 0.000 0.000
EFF1 -0.018 -0.038 -0.037 0.000
EFF2 0.001 -0.015 -0.015 0.000 0.000
TREAT -0.010 0.006 -0.004 0.005 -0.006
Residuals for Correlations
TREAT
________
TREAT 0.000
Correlation Residuals
Unstandardized Coefficients
MODEL RESULTS
Two-Tailed
Estimate S.E. Est./S.E. P-Value
LSREG BY
SR1 1.000 0.000 999.000 999.000
SR2 0.865 0.030 29.143 0.000
LEFF BY
EFF1 1.000 0.000 999.000 999.000
EFF2 0.791 0.030 25.954 0.000
LSREG ON
TREAT 0.999 0.069 14.381 0.000
LEFF ON
TREAT 0.964 0.073 13.278 0.000
WLOSS ON
LSREG 2.267 0.120 18.884 0.000
LEFF 1.715 0.129 13.319 0.000
Intercepts
WLOSS 0.034 0.184 0.184 0.854
SR1 0.066 0.048 1.359 0.174
SR2 0.010 0.044 0.220 0.826
EFF1 0.016 0.053 0.307 0.759
EFF2 0.048 0.043 1.097 0.272
Residual Variances
WLOSS 4.094 0.292 14.041 0.000
SR1 0.113 0.020 5.583 0.000
SR2 0.109 0.015 7.147 0.000
EFF1 0.103 0.022 4.720 0.000
EFF2 0.122 0.016 7.630 0.000
LSREG 0.507 0.038 13.232 0.000
LEFF 0.589 0.046 12.864 0.000
Two-Tailed
Estimate S.E. Est./S.E. P-Value
LSREG BY
SR1 1.000 0.000 999.000 999.000
SR2 0.865 0.030 29.143 0.000
LEFF BY
EFF1 1.000 0.000 999.000 999.000
EFF2 0.791 0.030 25.954 0.000
LSREG ON
TREAT 0.999 0.069 14.381 0.000
LEFF ON
TREAT 0.964 0.073 13.278 0.000
WLOSS ON
LSREG 2.267 0.120 18.884 0.000
LEFF 1.715 0.129 13.319 0.000
Intercepts
WLOSS 0.034 0.184 0.184 0.854
SR1 0.066 0.048 1.359 0.174
SR2 0.010 0.044 0.220 0.826
EFF1 0.016 0.053 0.307 0.759
EFF2 0.048 0.043 1.097 0.272
Residual Variances
WLOSS 4.094 0.292 14.041 0.000
SR1 0.113 0.020 5.583 0.000
SR2 0.109 0.015 7.147 0.000
EFF1 0.103 0.022 4.720 0.000
EFF2 0.122 0.016 7.630 0.000
LSREG 0.507 0.038 13.232 0.000
LEFF 0.589 0.046 12.864 0.000
Standardized Coefficients
STANDARDIZED MODEL RESULTS
STDYX Standardization
Two-Tailed
Estimate S.E. Est./S.E. P-Value
LSREG BY
SR1 0.933 0.013 72.532 0.000
SR2 0.915 0.013 72.019 0.000
LEFF BY
EFF1 0.943 0.013 74.614 0.000
EFF2 0.899 0.016 57.892 0.000
LSREG ON
TREAT 0.574 0.030 19.036 0.000
LEFF ON
TREAT 0.532 0.033 16.162 0.000
WLOSS ON
LSREG 0.563 0.025 22.657 0.000
LEFF 0.444 0.028 15.995 0.000
Residual Variances
WLOSS 0.334 0.025 13.147 0.000
SR1 0.130 0.024 5.411 0.000
SR2 0.162 0.023 6.959 0.000
EFF1 0.112 0.024 4.682 0.000
EFF2 0.192 0.028 6.877 0.000
LSREG 0.671 0.035 19.380 0.000
LEFF 0.717 0.035 20.517 0.000
R-SQUARE
Observed Two-Tailed
Variable Estimate S.E. Est./S.E. P-Value
WLOSS 0.666 0.025 26.247 0.000
SR1 0.870 0.024 36.266 0.000
SR2 0.838 0.023 36.010 0.000
EFF1 0.888 0.024 37.307 0.000
EFF2 0.808 0.028 28.946 0.000
Latent Two-Tailed
Variable Estimate S.E. Est./S.E. P-Value
LSREG 0.329 0.035 9.518 0.000
LEFF 0.283 0.035 8.081 0.000
STDYX Standardization
Two-Tailed
Estimate S.E. Est./S.E. P-Value
LSREG BY
SR1 0.933 0.013 72.532 0.000
SR2 0.915 0.013 72.019 0.000
LEFF BY
EFF1 0.943 0.013 74.614 0.000
EFF2 0.899 0.016 57.892 0.000
LSREG ON
TREAT 0.574 0.030 19.036 0.000
LEFF ON
TREAT 0.532 0.033 16.162 0.000
WLOSS ON
LSREG 0.563 0.025 22.657 0.000
LEFF 0.444 0.028 15.995 0.000
Residual Variances
WLOSS 0.334 0.025 13.147 0.000
SR1 0.130 0.024 5.411 0.000
SR2 0.162 0.023 6.959 0.000
EFF1 0.112 0.024 4.682 0.000
EFF2 0.192 0.028 6.877 0.000
LSREG 0.671 0.035 19.380 0.000
LEFF 0.717 0.035 20.517 0.000
R-SQUARE
Observed Two-Tailed
Variable Estimate S.E. Est./S.E. P-Value
WLOSS 0.666 0.025 26.247 0.000
SR1 0.870 0.024 36.266 0.000
SR2 0.838 0.023 36.010 0.000
EFF1 0.888 0.024 37.307 0.000
EFF2 0.808 0.028 28.946 0.000
Latent Two-Tailed
Variable Estimate S.E. Est./S.E. P-Value
LSREG 0.329 0.035 9.518 0.000
LEFF 0.283 0.035 8.081 0.000
CONFIDENCE INTERVALS OF MODEL RESULTS
Lower .5% Lower 2.5% Lower 5% Estimate Upper 5% Upper 2.5% Upper .5%
LSREG BY
SR1 1.000 1.000 1.000 1.000 1.000 1.000 1.000
SR2 0.789 0.807 0.816 0.865 0.914 0.923 0.941
LEFF BY
EFF1 1.000 1.000 1.000 1.000 1.000 1.000 1.000
EFF2 0.713 0.732 0.741 0.791 0.841 0.851 0.870
LSREG ON
TREAT 0.820 0.863 0.884 0.999 1.113 1.135 1.178
LEFF ON
TREAT 0.777 0.822 0.845 0.964 1.083 1.106 1.151
WLOSS ON
LSREG 1.957 2.031 2.069 2.267 2.464 2.502 2.576
LEFF 1.383 1.463 1.503 1.715 1.927 1.967 2.047
Lower .5% Lower 2.5% Lower 5% Estimate Upper 5% Upper 2.5% Upper .5%
LSREG BY
SR1 1.000 1.000 1.000 1.000 1.000 1.000 1.000
SR2 0.789 0.807 0.816 0.865 0.914 0.923 0.941
LEFF BY
EFF1 1.000 1.000 1.000 1.000 1.000 1.000 1.000
EFF2 0.713 0.732 0.741 0.791 0.841 0.851 0.870
LSREG ON
TREAT 0.820 0.863 0.884 0.999 1.113 1.135 1.178
LEFF ON
TREAT 0.777 0.822 0.845 0.964 1.083 1.106 1.151
WLOSS ON
LSREG 1.957 2.031 2.069 2.267 2.464 2.502 2.576
LEFF 1.383 1.463 1.503 1.715 1.927 1.967 2.047
Confidence Intervals
Confidence intervals are provided by Mplus using the above format for all unstandardized parameters, standardized parameters and direct and indirect effects. I do not show the output for each option to conserve space. I provide 'tool tips' for the row of labels at the top of the table and the individuals entries. See my book for qualifications about CIs for standardized coefficients.
TOTAL, TOTAL INDIRECT, SPECIFIC INDIRECT, AND DIRECT EFFECTS
Two-Tailed
Estimate S.E. Est./S.E. P-Value
Effects from TREAT to WLOSS
Total 3.917 0.235 16.661 0.000
Total indirect 3.917 0.235 16.661 0.000
Specific indirect 1
WLOSS
LSREG
TREAT 2.264 0.182 12.462 0.000
Specific indirect 2
WLOSS
LEFF
TREAT 1.653 0.171 9.678 0.000
Two-Tailed
Estimate S.E. Est./S.E. P-Value
Effects from TREAT to WLOSS
Total 3.917 0.235 16.661 0.000
Total indirect 3.917 0.235 16.661 0.000
Specific indirect 1
WLOSS
LSREG
TREAT 2.264 0.182 12.462 0.000
Specific indirect 2
WLOSS
LEFF
TREAT 1.653 0.171 9.678 0.000
Mediation Analysis:Unstandardized Coefficients
Predicted Correlations Between All Variables
ESTIMATED CORRELATION MATRIX FOR THE LATENT VARIABLES
LSREG LEFF WLOSS SR1 SR2
________ ________ ________ ________ ________
LSREG 1.000
LEFF 0.305 1.000
WLOSS 0.698 0.616 1.000
SR1 0.933 0.285 0.651 1.000
SR2 0.915 0.279 0.639 0.854 1.000
EFF1 0.288 0.943 0.580 0.268 0.263
EFF2 0.274 0.899 0.553 0.256 0.251
TREAT 0.574 0.532 0.559 0.535 0.525
ESTIMATED CORRELATION MATRIX FOR THE LATENT VARIABLES
EFF1 EFF2 TREAT
________ ________ ________
EFF1 1.000
EFF2 0.847 1.000
TREAT 0.501 0.478 1.000
LSREG LEFF WLOSS SR1 SR2
________ ________ ________ ________ ________
LSREG 1.000
LEFF 0.305 1.000
WLOSS 0.698 0.616 1.000
SR1 0.933 0.285 0.651 1.000
SR2 0.915 0.279 0.639 0.854 1.000
EFF1 0.288 0.943 0.580 0.268 0.263
EFF2 0.274 0.899 0.553 0.256 0.251
TREAT 0.574 0.532 0.559 0.535 0.525
ESTIMATED CORRELATION MATRIX FOR THE LATENT VARIABLES
EFF1 EFF2 TREAT
________ ________ ________
EFF1 1.000
EFF2 0.847 1.000
TREAT 0.501 0.478 1.000
Mplus does not always report the estimated correlations between model variables in the main portions of the output. The correlations reliably appear in the Tech4 output section, along with a matrix of estimated standard errors, critical ratios and p values (where you can determine if each correlation is statistically significant). The matrix includes both latent variables (LSREG and LEFF) and observed variables. I only report here the matrix of the predicted correlations. I omit 'tool tips' for the diagonal elements.
Mplus provides a table of standardized coefficients that parallel the unstandardized coefficients, but in this case, they are not relevant. This is because we typically do not interpret standardized coefficients for binary or nominal predictors. TREAT is binary. Some researchers use a partially standardized solution in such cases; see my book for details
Missing Data
For the web example, there is no missing data. If you include PATTERNS on the output line, you obtain information about missing data, if it exists. I use another example to illustrate the output. The model has four variables, X1, X2, X3, and X4. The first output is the covariance coverage matrix. It specifies the proportion of cases that did NOT have missing data for each variance and covariance in the 4X4 covariance matrix.
PROPORTION OF DATA PRESENT
Covariance Coverage
X1 X2 X3 X4
________ ________ ________ ________
X1 1.000
X2 1.000 1.000
X3 0.734 0.734 0.734
X4 0.503 0.503 0.383 0.503
Covariance Coverage
X1 X2 X3 X4
________ ________ ________ ________
X1 1.000
X2 1.000 1.000
X3 0.734 0.734 0.734
X4 0.503 0.503 0.383 0.503
Mplus also provides the patterns of missing data and the frequency of each pattern. In the current case, there were four patterns. The first pattern is in Column 1 and consisted of people with no missing data. The second pattern is in Column 2 and consisted of people who had complete data on X1, X2, and X3 but missing data on X4. The third pattern is in Column 3 and consisted of people who had complete data on X1, X2, and X4, but missing data on X3. The fourth pattern is in Column 4 and consisted of people who had complete data on X1 and X2 but missing data on X3 and X4. See Chapter 27 for nuances in the output.
MISSING DATA PATTERNS (x = not missing)
1 2 3 4
X1 x x x x
X2 x x x x
X3 x x
X4 x x
MISSING DATA PATTERN FREQUENCIES
Pattern Frequency Pattern Frequency
1 111 3 35
2 102 4 42
1 2 3 4
X1 x x x x
X2 x x x x
X3 x x
X4 x x
MISSING DATA PATTERN FREQUENCIES
Pattern Frequency Pattern Frequency
1 111 3 35
2 102 4 42
Descriptive Statistics
In the early part of the output, Mplus provides descriptive statistics for model variables. Information for each variable is presented in two rows. Mplus uses a column heading format that lists the name of the statistic in the first row before a slash (/) and the name of the statistic in the second row after the slash. For example, the heading Mean/Variance indicates the mean value for the row variable is in the first row and the variance is in the second row, beneath it.
UNIVARIATE HIGHER-ORDER MOMENT DESCRIPTIVE STATISTICS
Variable/ Mean/ Skewness/ Minimum/ % with Percentiles
Sample Size Variance Kurtosis Maximum Min/Max 20%/60% 40%/80% Median
WLOSS 1.942 -0.213 -7.840 0.20% -0.921 1.189 2.114
501.000 12.101 0.039 11.417 0.20% 2.893 4.836
SR1 0.552 0.065 -1.999 0.20% -0.305 0.294 0.561
501.000 0.869 -0.382 3.008 0.20% 0.816 1.355
SR2 0.430 0.070 -1.962 0.20% -0.309 0.204 0.383
501.000 0.676 -0.204 3.167 0.20% 0.627 1.149
EFF1 0.486 0.044 -1.862 0.20% -0.348 0.268 0.496
501.000 0.925 -0.262 3.434 0.20% 0.786 1.271
EFF2 0.419 0.070 -1.954 0.20% -0.271 0.209 0.423
501.000 0.637 -0.120 3.155 0.20% 0.621 1.114
TREAT 0.487 0.052 0.000 51.30% 0.000 0.000 0.000
501.000 0.250 -1.997 1.000 48.70% 1.000 1.000