search

simsem / semTools

Useful tools for structural equation modeling
75 stars 36 forks source link

measurementInvarianceCat() wrong estimator call #33

Closed jorgesinval closed 6 years ago

jorgesinval commented 6 years ago

I have this problem, when I run:

measurementInvarianceCat(model, data=data, estimator = "WLSMV", group = "group", parameterization="theta",ordered=ordered.items)

I get:

Error in lav_options_set(opt) : lavaan ERROR: estimator ML for ordered data is not supported yet. Use WLSMV instead.

TDJorgensen commented 6 years ago

Not sure why this occurred, but 2 others have had this problem, and it was fixed after they installed the latest software:

install.packages("lavaan", repos = "http://www.da.ugent.be", type = "source")
devtools::install_github("simsem/semTools/semTools")

Let me know if there is still an error in the latest versions.

jorgesinval commented 6 years ago

Now I get this error:

`

Error in lavaan::lavTestLRT(...) : argument "object" is missing, with no default

`

On Tue, Jun 19, 2018 at 4:56 PM Terrence notifications@github.com wrote:

Not sure why this occurred, but 2 others have had this problem, and it was fixed after they installed the latest software:

install.packages("lavaan", repos = "http://www.da.ugent.be", type = "source") devtools::install_github("simsem/semTools/semTools")

Let me know if there is still an error in the latest versions.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/simsem/semTools/issues/33#issuecomment-398451338, or mute the thread https://github.com/notifications/unsubscribe-auth/AH5QN8fNs-L7oTvL0ZsFcLfS7mHzmKp7ks5t-R83gaJpZM4UtwBr .

TDJorgensen commented 6 years ago

The help page example works fine, so it must be a problem with your analysis. Does your configural model run without an error? Make sure you name all your arguments:

measurementInvarianceCat(model = model, ...)
chriscastille6 commented 6 years ago

I'm having a similar issue. The full error message is:

The following model(s) did not converge: fit.configuralfit.loadingsfit.thresholdsfit.means Error in lavaan::lavTestLRT(...) : argument "object" is missing, with no default In addition: Warning message: In lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some cases are empty and will be ignored: 94

My model runs with regard to the lavaan cfa() function. It is below for reference (nothing complex as you can see):

initial.cfa <- ' PA =~ PA1 + PA2 + PA3 + PA4 + PA5 + PA6 + PA7 + PA8 + PA9 + PA10 '

There is one error that is generated, however, that is common across both analyses: Warning message: In lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some cases are empty and will be ignored: 94

Please advise.

chriscastille6 commented 6 years ago

Also, there does not appear to be an issue with the model syntax.

TDJorgensen commented 6 years ago

My help page example works fine. Can you check your sessionInfo() output to verify the version numbers match?

If you have the latest software, could you please email me (TJorgensen314 "at" gmail "dot" com) enough of your data and the R syntax that reproduces the error?

jorgesinval commented 6 years ago

my session:

version 3.3.3 (2017-03-06) Platform: x86_64-pc-linux-gnu (64-bit) Running under: Debian GNU/Linux 9 (stretch)

​

semTools_0.4-15.930 lavaan_0.6-2.1261

​

On Tue, Jun 19, 2018 at 9:09 PM Terrence notifications@github.com wrote:

My help page example works fine. Can you check your sessionInfo() output to verify the version numbers match?

  • R version 3.5.0 (2018-04-23)
  • lavaan_0.6-2.1261
  • semTools_0.4-15.930

If you have the latest software, could you please email me (TJorgensen314 "at" gmail "dot" com) enough of your data and the R syntax that reproduces the error?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/simsem/semTools/issues/33#issuecomment-398528766, or mute the thread https://github.com/notifications/unsubscribe-auth/AH5QN_Qu9QHU8aA_vrU_GAVQutZ5xX7Jks5t-VqDgaJpZM4UtwBr .

chriscastille6 commented 6 years ago

All but lavaan match:

So I updated lavaan: *lavaan_0.6-2.1261

I'm still getting the same error.

I'll send you an email. Thanks in advance for you taking a look.

TDJorgensen commented 6 years ago

Thanks for sending your data. As I mentioned to @jorgesinval above (and is demonstrated in the help page example), you need to name all your lavaan() arguments (i.e., including the first model= argument). Also, missing = "partial" is not an option for handling missing data. Your example syntax runs fine after you fix these 2 issues.

mi <- measurementInvarianceCat(model = initial.cfa,
                               ordered = c("PA1", "PA2", "PA3", "PA4", "PA5", 
                                           "PA6", "PA7", "PA8", "PA9", "PA10"), 
                               data = data2, group="COND", missing ="pairwise", 
                               parameterization = "theta", estimator = "dwls",
                               information = "expected")

I will add a check for a model= argument and issue an error with this instruction, to clarify this point for users in the future.

jorgesinval commented 6 years ago

Thanks, Jorgensen

On Thu, Jun 21, 2018 at 9:07 AM Terrence notifications@github.com wrote:

Thanks for sending your data. As I mentioned to @jorgesinval https://github.com/jorgesinval above (and is demonstrated in the help page example), you need to name all your lavaan() arguments (i.e., including the first model= argument). Also, missing = "partial" is not an option for handling missing data. Your example syntax runs fine after you fix these 2 issues.

mi <- measurementInvarianceCat(model = initial.cfa, ordered = c("PA1", "PA2", "PA3", "PA4", "PA5", "PA6", "PA7", "PA8", "PA9", "PA10"), data = data2, group="COND", missing ="pairwise", parameterization = "theta", estimator = "dwls", information = "expected")

I will add a check for a model= argument and issue an error with this instruction, to clarify this point for users in the future.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/simsem/semTools/issues/33#issuecomment-399013943, or mute the thread https://github.com/notifications/unsubscribe-auth/AH5QN2Ywg5maTS7FTvZ5mYMN3Xs_d6Haks5t-1RGgaJpZM4UtwBr .

chriscastille6 commented 6 years ago

Hi Terrence,

Thanks for helping out! I really appreciate you clearly pointing out the mistakes that I made.

I do have a few more questions for you:

  1. Research suggests that DWLS-based CFI and TLI are often so close to 1.00 that they are difficult to use for assessing model misspecification (Ny & Drasgow, 2011). In addition to the chi-square, they recommend that SRMSR and RMSEA be consulted in the decision making process. Does the measurementInvarianceCat() function allow one to easily calculate SRMSR? Do you have any guidance to offer in calculating SRMSR?
  2. In many cases, models 1-4 are statistically indistinguishable from one another (i.e., no statistically significant differences emerged in the model comparison process). However, the fit.means model (i.e., model 5) is statistically significant in many cases. Am I correct in interpreting the results as implying that mean differences exist on the construct as a function of group membership? If so, then how might I estimate the size of this difference? To give you a bit of context, I’ve administered the same scale just under different methodological conditions (e.g., items are randomized vs. not randomized).

References: Nye, C. D., & Drasgow, F. (2011). Assessing Goodness of Fit: Simple Rules of Thumb Simply Do Not Work. Organizational Research Methods, 14(3), 548–570. https://doi.org/10.1177/1094428110368562 https://doi.org/10.1177/1094428110368562

Thanks in advance for your help!

Chris

Christopher M. Castille, Ph.D. Assistant Professor of Management College of Business Administration Nicholls State University email: christopher.castille@nicholls.edu

On Jun 21, 2018, at 3:07 AM, Terrence notifications@github.com wrote:

mi <- measurementInvarianceCat(model = initial.cfa, ordered = c("PA1", "PA2", "PA3", "PA4", "PA5", "PA6", "PA7", "PA8", "PA9", "PA10"), data = data2, group="COND", missing ="pairwise", parameterization = "theta", estimator = "dwls", information = "expected")

TDJorgensen commented 6 years ago
  1. You can request any fit measures you want using the fit.measures= argument.
  2. Yes, and you can investigate the standardized model parameters to see how different the means are. The first group's mean == 0, so any other group means are therefore already the difference from the first (reference) group. In a standardized metric (the std.all column in the summary(fit, std = TRUE) output), that is interpreted as Cohen's d. Likewise, the difference between any other groups' (e.g., groups 2 and 3) standardized means are the Cohen's d for that group comparison. You can obtain the fitted model from the list of results returned by the function. 
    mi <- measurementInvarianceCat(...)
summary(mi$fit.means, std = TRUE)
chriscastille6 commented 6 years ago

Thanks for replying, Terrence.

I’m still not quite seeing ini the output what you are saying. You said:

The first group's mean == 0, so any other group means are therefore already the difference from the first (reference) group. In a standardized metric (the std.all column in the summary(fit, std = TRUE) output), that is interpreted as Cohen's d. Under which part of the output am I to find the group means? My output is below (I’m sorry about dumping it here, but I figured this would help you to help me).

lavaan 0.6-2.1261 optimization ended normally (153 iterations)

Optimization method NLMINB Number of free parameters 290 Number of equality constraints 135 Number of observations per group
0 117 1 184

Estimator DWLS Model Fit Test Statistic 2060.261 Degrees of freedom 991 P-value (Chi-square) 0.000

Chi-square for each group:

0 1099.841 1 960.420

Parameter Estimates:

Information Expected Information saturated (h1) model Unstructured Standard Errors Standard

Group 1 [0]:

Latent Variables: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PR =~
PP1 1.000 1.090 0.737 PP2 1.066 0.081 13.111 0.000 1.162 0.758 PP3 1.039 0.079 13.174 0.000 1.133 0.750 PP4 1.215 0.096 12.611 0.000 1.325 0.798 PP5 1.195 0.094 12.777 0.000 1.303 0.793 PP6 0.958 0.070 13.702 0.000 1.044 0.722 PP7 0.909 0.067 13.597 0.000 0.991 0.704 PP8 1.394 0.116 11.996 0.000 1.519 0.835 PP9 1.132 0.082 13.829 0.000 1.234 0.777 PP10 1.045 0.078 13.366 0.000 1.140 0.752 IRB =~
IRB1 1.000 1.754 0.869 IRB2 0.956 0.109 8.761 0.000 1.677 0.859 IRB3 1.205 0.168 7.155 0.000 2.113 0.904 IRB4 1.313 0.197 6.676 0.000 2.302 0.917 IRB5 0.210 0.020 10.371 0.000 0.367 0.345 IRB6 -0.795 0.089 -8.936 0.000 -1.395 -0.813 IRB7 -0.601 0.058 -10.421 0.000 -1.054 -0.726 OCBI =~
OCBI1 1.000 1.221 0.774 OCBI2 0.839 0.072 11.703 0.000 1.025 0.716 OCBI3 0.550 0.043 12.679 0.000 0.672 0.558 OCBI4 0.969 0.084 11.583 0.000 1.183 0.764 OCBI5 0.994 0.084 11.818 0.000 1.214 0.772 OCBI6 0.966 0.080 12.025 0.000 1.179 0.763 OCBI7 0.930 0.080 11.598 0.000 1.135 0.750 OCBO =~
OCBO1 1.000 0.683 0.564 OCBO2 1.128 0.084 13.462 0.000 0.771 0.610 OCBO3 -1.074 0.073 -14.784 0.000 -0.734 -0.592 OCBO4 -0.914 0.068 -13.358 0.000 -0.625 -0.530 OCBO5 -1.154 0.083 -13.881 0.000 -0.788 -0.619 OCBO6 1.138 0.083 13.667 0.000 0.777 0.614 OCBO7 0.878 0.064 13.739 0.000 0.600 0.514

Covariances: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PR ~~
IRB 0.627 0.068 9.160 0.000 0.328 0.328 OCBI 0.711 0.059 12.017 0.000 0.534 0.534 OCBO 0.313 0.027 11.475 0.000 0.420 0.420 IRB ~~
OCBI 0.850 0.095 8.905 0.000 0.397 0.397 OCBO 0.971 0.096 10.096 0.000 0.810 0.810 OCBI ~~
OCBO 0.548 0.047 11.580 0.000 0.656 0.656

Intercepts: Estimate Std.Err z-value P(>|z|) Std.lv Std.all .PP1 0.000 0.000 0.000 .PP2 0.000 0.000 0.000 .PP3 0.000 0.000 0.000 .PP4 0.000 0.000 0.000 .PP5 0.000 0.000 0.000 .PP6 0.000 0.000 0.000 .PP7 0.000 0.000 0.000 .PP8 0.000 0.000 0.000 .PP9 0.000 0.000 0.000 .PP10 0.000 0.000 0.000 .IRB1 0.000 0.000 0.000 .IRB2 0.000 0.000 0.000 .IRB3 0.000 0.000 0.000 .IRB4 0.000 0.000 0.000 .IRB5 0.000 0.000 0.000 .IRB6 0.000 0.000 0.000 .IRB7 0.000 0.000 0.000 .OCBI1 0.000 0.000 0.000 .OCBI2 0.000 0.000 0.000 .OCBI3 0.000 0.000 0.000 .OCBI4 0.000 0.000 0.000 .OCBI5 0.000 0.000 0.000 .OCBI6 0.000 0.000 0.000 .OCBI7 0.000 0.000 0.000 .OCBO1 0.000 0.000 0.000 .OCBO2 0.000 0.000 0.000 .OCBO3 0.000 0.000 0.000 .OCBO4 0.000 0.000 0.000 .OCBO5 0.000 0.000 0.000 .OCBO6 0.000 0.000 0.000 .OCBO7 0.000 0.000 0.000 PR 0.000 0.000 0.000 IRB 0.000 0.000 0.000 OCBI 0.000 0.000 0.000 OCBO 0.000 0.000 0.000

Thresholds: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PP1|t1 -3.049 0.281 -10.864 0.000 -3.049 -2.061 PP1|t2 -0.857 0.111 -7.745 0.000 -0.857 -0.579 PP1|t3 0.020 0.101 0.194 0.846 0.020 0.013 PP1|t4 2.108 0.178 11.855 0.000 2.108 1.425 PP2|t1 -2.574 0.221 -11.627 0.000 -2.574 -1.679 PP2|t2 -1.528 0.137 -11.149 0.000 -1.528 -0.996 PP2|t3 0.913 0.116 7.897 0.000 0.913 0.596 PP3|t1 -3.337 0.334 -10.005 0.000 -3.337 -2.208 PP3|t2 -2.405 0.194 -12.404 0.000 -2.405 -1.592 PP3|t3 -1.011 0.117 -8.667 0.000 -1.011 -0.669 PP3|t4 1.485 0.133 11.157 0.000 1.485 0.983 PP4|t1 -3.868 0.418 -9.248 0.000 -3.868 -2.330 PP4|t2 -2.735 0.227 -12.042 0.000 -2.735 -1.647 PP4|t3 -1.836 0.160 -11.477 0.000 -1.836 -1.106 PP4|t4 0.986 0.126 7.840 0.000 0.986 0.594 PP5|t1 -1.568 0.144 -10.865 0.000 -1.568 -0.955 PP5|t2 -0.351 0.113 -3.091 0.002 -0.351 -0.214 PP5|t3 1.788 0.155 11.542 0.000 1.788 1.089 PP6|t1 -2.386 0.189 -12.646 0.000 -2.386 -1.650 PP6|t2 -1.426 0.126 -11.281 0.000 -1.426 -0.986 PP6|t3 0.824 0.108 7.630 0.000 0.824 0.570 PP7|t1 -2.921 0.266 -10.962 0.000 -2.921 -2.075 PP7|t2 -1.337 0.121 -11.079 0.000 -1.337 -0.950 PP7|t3 -0.510 0.100 -5.086 0.000 -0.510 -0.362 PP7|t4 1.281 0.118 10.814 0.000 1.281 0.910 PP8|t1 -3.664 0.357 -10.266 0.000 -3.664 -2.014 PP8|t2 -1.821 0.169 -10.776 0.000 -1.821 -1.001 PP8|t3 -0.423 0.125 -3.388 0.001 -0.423 -0.233 PP8|t4 1.864 0.172 10.823 0.000 1.864 1.025 PP9|t1 -2.928 0.248 -11.794 0.000 -2.928 -1.843 PP9|t2 -1.160 0.125 -9.278 0.000 -1.160 -0.730 PP9|t3 -0.052 0.108 -0.484 0.629 -0.052 -0.033 PP9|t4 1.859 0.154 12.097 0.000 1.859 1.170 PP10|t1 -1.330 0.127 -10.494 0.000 -1.330 -0.877 PP10|t2 0.008 0.104 0.076 0.940 0.008 0.005 PP10|t3 2.085 0.169 12.358 0.000 2.085 1.375 IRB1|t1 -5.550 0.679 -8.179 0.000 -5.550 -2.749 IRB1|t2 -4.065 0.408 -9.954 0.000 -4.065 -2.014 IRB1|t3 0.437 0.172 2.533 0.011 0.437 0.216 IRB2|t1 -4.730 0.490 -9.656 0.000 -4.730 -2.423 IRB2|t2 -3.328 0.320 -10.396 0.000 -3.328 -1.705 IRB2|t3 0.228 0.164 1.393 0.164 0.228 0.117 IRB3|t1 -6.353 0.821 -7.743 0.000 -6.353 -2.718 IRB3|t2 -5.154 0.612 -8.426 0.000 -5.154 -2.205 IRB3|t3 0.164 0.198 0.831 0.406 0.164 0.070 IRB4|t1 -6.224 0.833 -7.467 0.000 -6.224 -2.479 IRB4|t2 -4.515 0.565 -7.992 0.000 -4.515 -1.799 IRB4|t3 0.377 0.216 1.745 0.081 0.377 0.150 IRB5|t1 -2.065 0.163 -12.662 0.000 -2.065 -1.938 IRB5|t2 -1.319 0.108 -12.227 0.000 -1.319 -1.238 IRB5|t3 -0.711 0.086 -8.236 0.000 -0.711 -0.667 IRB5|t4 0.803 0.088 9.107 0.000 0.803 0.754 IRB6|t1 -0.142 0.142 -1.000 0.317 -0.142 -0.083 IRB6|t2 2.390 0.233 10.274 0.000 2.390 1.393 IRB6|t3 3.013 0.284 10.623 0.000 3.013 1.756 IRB7|t1 0.311 0.119 2.617 0.009 0.311 0.214 IRB7|t2 2.166 0.180 12.055 0.000 2.166 1.491 IRB7|t3 2.372 0.198 11.987 0.000 2.372 1.632 IRB7|t4 3.047 0.299 10.184 0.000 3.047 2.097 OCBI1|t1 -2.130 0.180 -11.854 0.000 -2.130 -1.349 OCBI1|t2 -1.266 0.137 -9.220 0.000 -1.266 -0.802 OCBI1|t3 1.177 0.134 8.759 0.000 1.177 0.745 OCBI2|t1 -2.119 0.169 -12.554 0.000 -2.119 -1.480 OCBI2|t2 -1.041 0.119 -8.717 0.000 -1.041 -0.727 OCBI2|t3 0.988 0.117 8.438 0.000 0.988 0.690 OCBI3|t1 -2.273 0.179 -12.673 0.000 -2.273 -1.887 OCBI3|t2 -0.907 0.099 -9.203 0.000 -0.907 -0.753 OCBI3|t3 -0.096 0.088 -1.091 0.275 -0.096 -0.079 OCBI3|t4 1.220 0.108 11.309 0.000 1.220 1.012 OCBI4|t1 -3.558 0.365 -9.746 0.000 -3.558 -2.296 OCBI4|t2 -2.401 0.200 -11.979 0.000 -2.401 -1.549 OCBI4|t3 -1.444 0.143 -10.126 0.000 -1.444 -0.932 OCBI4|t4 1.070 0.129 8.283 0.000 1.070 0.691 OCBI5|t1 -3.613 0.368 -9.808 0.000 -3.613 -2.296 OCBI5|t2 -2.156 0.178 -12.078 0.000 -2.156 -1.370 OCBI5|t3 -1.020 0.129 -7.935 0.000 -1.020 -0.648 OCBI5|t4 0.911 0.126 7.248 0.000 0.911 0.579 OCBI6|t1 -2.863 0.281 -10.206 0.000 -2.863 -1.852 OCBI6|t2 -1.730 0.153 -11.271 0.000 -1.730 -1.119 OCBI6|t3 -0.646 0.118 -5.457 0.000 -0.646 -0.418 OCBI6|t4 1.426 0.140 10.176 0.000 1.426 0.922 OCBI7|t1 -2.797 0.245 -11.400 0.000 -2.797 -1.848 OCBI7|t2 -2.035 0.171 -11.924 0.000 -2.035 -1.345 OCBI7|t3 0.802 0.119 6.720 0.000 0.802 0.530 OCBO1|t1 -2.701 0.237 -11.391 0.000 -2.701 -2.230 OCBO1|t2 -1.675 0.133 -12.554 0.000 -1.675 -1.383 OCBO1|t3 -0.830 0.100 -8.261 0.000 -0.830 -0.685 OCBO1|t4 0.494 0.095 5.208 0.000 0.494 0.408 OCBO2|t1 -2.432 0.195 -12.497 0.000 -2.432 -1.927 OCBO2|t2 -1.582 0.131 -12.048 0.000 -1.582 -1.253 OCBO2|t3 0.207 0.097 2.141 0.032 0.207 0.164 OCBO3|t1 -0.479 0.097 -4.932 0.000 -0.479 -0.386 OCBO3|t2 1.234 0.114 10.787 0.000 1.234 0.995 OCBO3|t3 1.718 0.136 12.660 0.000 1.718 1.385 OCBO3|t4 2.851 0.258 11.058 0.000 2.851 2.299 OCBO4|t1 0.261 0.090 2.911 0.004 0.261 0.222 OCBO4|t2 1.567 0.125 12.540 0.000 1.567 1.329 OCBO4|t3 1.895 0.149 12.696 0.000 1.895 1.607 OCBO5|t1 -0.511 0.101 -5.076 0.000 -0.511 -0.401 OCBO5|t2 0.921 0.108 8.500 0.000 0.921 0.723 OCBO5|t3 1.633 0.135 12.116 0.000 1.633 1.283 OCBO5|t4 2.793 0.271 10.312 0.000 2.793 2.193 OCBO6|t1 -2.241 0.176 -12.748 0.000 -2.241 -1.770 OCBO6|t2 -0.920 0.108 -8.520 0.000 -0.920 -0.726 OCBO6|t3 1.033 0.111 9.317 0.000 1.033 0.815 OCBO7|t1 -2.124 0.164 -12.943 0.000 -2.124 -1.821 OCBO7|t2 -1.038 0.102 -10.142 0.000 -1.038 -0.890 OCBO7|t3 1.139 0.105 10.823 0.000 1.139 0.977

Variances: Estimate Std.Err z-value P(>|z|) Std.lv Std.all .PP1 1.000 1.000 0.457 .PP2 1.000 1.000 0.425 .PP3 1.000 1.000 0.438 .PP4 1.000 1.000 0.363 .PP5 1.000 1.000 0.371 .PP6 1.000 1.000 0.478 .PP7 1.000 1.000 0.505 .PP8 1.000 1.000 0.302 .PP9 1.000 1.000 0.396 .PP10 1.000 1.000 0.435 .IRB1 1.000 1.000 0.245 .IRB2 1.000 1.000 0.262 .IRB3 1.000 1.000 0.183 .IRB4 1.000 1.000 0.159 .IRB5 1.000 1.000 0.881 .IRB6 1.000 1.000 0.339 .IRB7 1.000 1.000 0.474 .OCBI1 1.000 1.000 0.401 .OCBI2 1.000 1.000 0.488 .OCBI3 1.000 1.000 0.689 .OCBI4 1.000 1.000 0.417 .OCBI5 1.000 1.000 0.404 .OCBI6 1.000 1.000 0.418 .OCBI7 1.000 1.000 0.437 .OCBO1 1.000 1.000 0.682 .OCBO2 1.000 1.000 0.627 .OCBO3 1.000 1.000 0.650 .OCBO4 1.000 1.000 0.719 .OCBO5 1.000 1.000 0.617 .OCBO6 1.000 1.000 0.623 .OCBO7 1.000 1.000 0.735 PR 1.189 0.124 9.582 0.000 1.000 1.000 IRB 3.075 0.525 5.856 0.000 1.000 1.000 OCBI 1.491 0.185 8.055 0.000 1.000 1.000 OCBO 0.467 0.055 8.541 0.000 1.000 1.000

Scales y*: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PP1 0.676 0.676 1.000 PP2 0.652 0.652 1.000 PP3 0.662 0.662 1.000 PP4 0.602 0.602 1.000 PP5 0.609 0.609 1.000 PP6 0.692 0.692 1.000 PP7 0.710 0.710 1.000 PP8 0.550 0.550 1.000 PP9 0.630 0.630 1.000 PP10 0.659 0.659 1.000 IRB1 0.495 0.495 1.000 IRB2 0.512 0.512 1.000 IRB3 0.428 0.428 1.000 IRB4 0.398 0.398 1.000 IRB5 0.939 0.939 1.000 IRB6 0.583 0.583 1.000 IRB7 0.688 0.688 1.000 OCBI1 0.634 0.634 1.000 OCBI2 0.698 0.698 1.000 OCBI3 0.830 0.830 1.000 OCBI4 0.645 0.645 1.000 OCBI5 0.636 0.636 1.000 OCBI6 0.647 0.647 1.000 OCBI7 0.661 0.661 1.000 OCBO1 0.826 0.826 1.000 OCBO2 0.792 0.792 1.000 OCBO3 0.806 0.806 1.000 OCBO4 0.848 0.848 1.000 OCBO5 0.785 0.785 1.000 OCBO6 0.790 0.790 1.000 OCBO7 0.858 0.858 1.000

Group 2 [1]:

Latent Variables: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PR =~
PP1 1.000 0.912 0.674 PP2 1.066 0.081 13.111 0.000 0.972 0.697 PP3 1.039 0.079 13.174 0.000 0.947 0.688 PP4 1.215 0.096 12.611 0.000 1.108 0.742 PP5 1.195 0.094 12.777 0.000 1.090 0.737 PP6 0.958 0.070 13.702 0.000 0.873 0.658 PP7 0.909 0.067 13.597 0.000 0.829 0.638 PP8 1.394 0.116 11.996 0.000 1.271 0.786 PP9 1.132 0.082 13.829 0.000 1.032 0.718 PP10 1.045 0.078 13.366 0.000 0.953 0.690 IRB =~
IRB1 1.000 2.417 0.924 IRB2 0.956 0.109 8.761 0.000 2.311 0.918 IRB3 1.205 0.168 7.155 0.000 2.913 0.946 IRB4 1.313 0.197 6.676 0.000 3.174 0.954 IRB5 0.210 0.020 10.371 0.000 0.507 0.452 IRB6 -0.795 0.089 -8.936 0.000 -1.923 -0.887 IRB7 -0.601 0.058 -10.421 0.000 -1.453 -0.824 OCBI =~
OCBI1 1.000 1.240 0.778 OCBI2 0.839 0.072 11.703 0.000 1.041 0.721 OCBI3 0.550 0.043 12.679 0.000 0.682 0.564 OCBI4 0.969 0.084 11.583 0.000 1.202 0.769 OCBI5 0.994 0.084 11.818 0.000 1.233 0.777 OCBI6 0.966 0.080 12.025 0.000 1.198 0.768 OCBI7 0.930 0.080 11.598 0.000 1.153 0.756 OCBO =~
OCBO1 1.000 0.836 0.641 OCBO2 1.128 0.084 13.462 0.000 0.943 0.686 OCBO3 -1.074 0.073 -14.784 0.000 -0.898 -0.668 OCBO4 -0.914 0.068 -13.358 0.000 -0.764 -0.607 OCBO5 -1.154 0.083 -13.881 0.000 -0.965 -0.694 OCBO6 1.138 0.083 13.667 0.000 0.951 0.689 OCBO7 0.878 0.064 13.739 0.000 0.734 0.592

Covariances: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PR ~~
IRB 0.866 0.089 9.716 0.000 0.393 0.393 OCBI 0.630 0.053 11.930 0.000 0.557 0.557 OCBO 0.291 0.026 11.428 0.000 0.382 0.382 IRB ~~
OCBI 1.604 0.167 9.630 0.000 0.535 0.535 OCBO 1.526 0.144 10.590 0.000 0.755 0.755 OCBI ~~
OCBO 0.616 0.051 12.100 0.000 0.594 0.594

Intercepts: Estimate Std.Err z-value P(>|z|) Std.lv Std.all .PP1 0.000 0.000 0.000 .PP2 0.000 0.000 0.000 .PP3 0.000 0.000 0.000 .PP4 0.000 0.000 0.000 .PP5 0.000 0.000 0.000 .PP6 0.000 0.000 0.000 .PP7 0.000 0.000 0.000 .PP8 0.000 0.000 0.000 .PP9 0.000 0.000 0.000 .PP10 0.000 0.000 0.000 .IRB1 0.000 0.000 0.000 .IRB2 0.000 0.000 0.000 .IRB3 0.000 0.000 0.000 .IRB4 0.000 0.000 0.000 .IRB5 0.000 0.000 0.000 .IRB6 0.000 0.000 0.000 .IRB7 0.000 0.000 0.000 .OCBI1 0.000 0.000 0.000 .OCBI2 0.000 0.000 0.000 .OCBI3 0.000 0.000 0.000 .OCBI4 0.000 0.000 0.000 .OCBI5 0.000 0.000 0.000 .OCBI6 0.000 0.000 0.000 .OCBI7 0.000 0.000 0.000 .OCBO1 0.000 0.000 0.000 .OCBO2 0.000 0.000 0.000 .OCBO3 0.000 0.000 0.000 .OCBO4 0.000 0.000 0.000 .OCBO5 0.000 0.000 0.000 .OCBO6 0.000 0.000 0.000 .OCBO7 0.000 0.000 0.000 PR 0.000 0.000 0.000 IRB 0.000 0.000 0.000 OCBI 0.000 0.000 0.000 OCBO 0.000 0.000 0.000

Thresholds: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PP1|t1 -3.049 0.281 -10.864 0.000 -3.049 -2.253 PP1|t2 -0.857 0.111 -7.745 0.000 -0.857 -0.633 PP1|t3 0.020 0.101 0.194 0.846 0.020 0.015 PP1|t4 2.108 0.178 11.855 0.000 2.108 1.558 PP2|t1 -2.574 0.221 -11.627 0.000 -2.574 -1.846 PP2|t2 -1.528 0.137 -11.149 0.000 -1.528 -1.096 PP2|t3 0.913 0.116 7.897 0.000 0.913 0.655 PP3|t1 -3.337 0.334 -10.005 0.000 -3.337 -2.422 PP3|t2 -2.405 0.194 -12.404 0.000 -2.405 -1.746 PP3|t3 -1.011 0.117 -8.667 0.000 -1.011 -0.734 PP3|t4 1.485 0.133 11.157 0.000 1.485 1.078 PP4|t1 -3.868 0.418 -9.248 0.000 -3.868 -2.591 PP4|t2 -2.735 0.227 -12.042 0.000 -2.735 -1.832 PP4|t3 -1.836 0.160 -11.477 0.000 -1.836 -1.230 PP4|t4 0.986 0.126 7.840 0.000 0.986 0.660 PP5|t1 -1.568 0.144 -10.865 0.000 -1.568 -1.060 PP5|t2 -0.351 0.113 -3.091 0.002 -0.351 -0.237 PP5|t3 1.788 0.155 11.542 0.000 1.788 1.209 PP6|t1 -2.386 0.189 -12.646 0.000 -2.386 -1.797 PP6|t2 -1.426 0.126 -11.281 0.000 -1.426 -1.074 PP6|t3 0.824 0.108 7.630 0.000 0.824 0.620 PP7|t1 -2.921 0.266 -10.962 0.000 -2.921 -2.249 PP7|t2 -1.337 0.121 -11.079 0.000 -1.337 -1.030 PP7|t3 -0.510 0.100 -5.086 0.000 -0.510 -0.393 PP7|t4 1.281 0.118 10.814 0.000 1.281 0.986 PP8|t1 -3.664 0.357 -10.266 0.000 -3.664 -2.266 PP8|t2 -1.821 0.169 -10.776 0.000 -1.821 -1.126 PP8|t3 -0.423 0.125 -3.388 0.001 -0.423 -0.262 PP8|t4 1.864 0.172 10.823 0.000 1.864 1.153 PP9|t1 -2.928 0.248 -11.794 0.000 -2.928 -2.037 PP9|t2 -1.160 0.125 -9.278 0.000 -1.160 -0.807 PP9|t3 -0.052 0.108 -0.484 0.629 -0.052 -0.036 PP9|t4 1.859 0.154 12.097 0.000 1.859 1.293 PP10|t1 -1.330 0.127 -10.494 0.000 -1.330 -0.962 PP10|t2 0.008 0.104 0.076 0.940 0.008 0.006 PP10|t3 2.085 0.169 12.358 0.000 2.085 1.509 IRB1|t1 -5.550 0.679 -8.179 0.000 -5.550 -2.122 IRB1|t2 -4.065 0.408 -9.954 0.000 -4.065 -1.554 IRB1|t3 0.437 0.172 2.533 0.011 0.437 0.167 IRB2|t1 -4.730 0.490 -9.656 0.000 -4.730 -1.878 IRB2|t2 -3.328 0.320 -10.396 0.000 -3.328 -1.321 IRB2|t3 0.228 0.164 1.393 0.164 0.228 0.091 IRB3|t1 -6.353 0.821 -7.743 0.000 -6.353 -2.063 IRB3|t2 -5.154 0.612 -8.426 0.000 -5.154 -1.674 IRB3|t3 0.164 0.198 0.831 0.406 0.164 0.053 IRB4|t1 -6.224 0.833 -7.467 0.000 -6.224 -1.870 IRB4|t2 -4.515 0.565 -7.992 0.000 -4.515 -1.357 IRB4|t3 0.377 0.216 1.745 0.081 0.377 0.113 IRB5|t1 -2.065 0.163 -12.662 0.000 -2.065 -1.842 IRB5|t2 -1.319 0.108 -12.227 0.000 -1.319 -1.177 IRB5|t3 -0.711 0.086 -8.236 0.000 -0.711 -0.634 IRB5|t4 0.803 0.088 9.107 0.000 0.803 0.716 IRB6|t1 -0.142 0.142 -1.000 0.317 -0.142 -0.066 IRB6|t2 2.390 0.233 10.274 0.000 2.390 1.103 IRB6|t3 3.013 0.284 10.623 0.000 3.013 1.390 IRB7|t1 0.311 0.119 2.617 0.009 0.311 0.176 IRB7|t2 2.166 0.180 12.055 0.000 2.166 1.228 IRB7|t3 2.372 0.198 11.987 0.000 2.372 1.345 IRB7|t4 3.047 0.299 10.184 0.000 3.047 1.727 OCBI1|t1 -2.130 0.180 -11.854 0.000 -2.130 -1.337 OCBI1|t2 -1.266 0.137 -9.220 0.000 -1.266 -0.794 OCBI1|t3 1.177 0.134 8.759 0.000 1.177 0.739 OCBI2|t1 -2.119 0.169 -12.554 0.000 -2.119 -1.468 OCBI2|t2 -1.041 0.119 -8.717 0.000 -1.041 -0.721 OCBI2|t3 0.988 0.117 8.438 0.000 0.988 0.685 OCBI3|t1 -2.273 0.179 -12.673 0.000 -2.273 -1.877 OCBI3|t2 -0.907 0.099 -9.203 0.000 -0.907 -0.749 OCBI3|t3 -0.096 0.088 -1.091 0.275 -0.096 -0.079 OCBI3|t4 1.220 0.108 11.309 0.000 1.220 1.008 OCBI4|t1 -3.558 0.365 -9.746 0.000 -3.558 -2.275 OCBI4|t2 -2.401 0.200 -11.979 0.000 -2.401 -1.535 OCBI4|t3 -1.444 0.143 -10.126 0.000 -1.444 -0.923 OCBI4|t4 1.070 0.129 8.283 0.000 1.070 0.684 OCBI5|t1 -3.613 0.368 -9.808 0.000 -3.613 -2.275 OCBI5|t2 -2.156 0.178 -12.078 0.000 -2.156 -1.357 OCBI5|t3 -1.020 0.129 -7.935 0.000 -1.020 -0.642 OCBI5|t4 0.911 0.126 7.248 0.000 0.911 0.574 OCBI6|t1 -2.863 0.281 -10.206 0.000 -2.863 -1.835 OCBI6|t2 -1.730 0.153 -11.271 0.000 -1.730 -1.108 OCBI6|t3 -0.646 0.118 -5.457 0.000 -0.646 -0.414 OCBI6|t4 1.426 0.140 10.176 0.000 1.426 0.914 OCBI7|t1 -2.797 0.245 -11.400 0.000 -2.797 -1.832 OCBI7|t2 -2.035 0.171 -11.924 0.000 -2.035 -1.333 OCBI7|t3 0.802 0.119 6.720 0.000 0.802 0.525 OCBO1|t1 -2.701 0.237 -11.391 0.000 -2.701 -2.072 OCBO1|t2 -1.675 0.133 -12.554 0.000 -1.675 -1.285 OCBO1|t3 -0.830 0.100 -8.261 0.000 -0.830 -0.637 OCBO1|t4 0.494 0.095 5.208 0.000 0.494 0.379 OCBO2|t1 -2.432 0.195 -12.497 0.000 -2.432 -1.770 OCBO2|t2 -1.582 0.131 -12.048 0.000 -1.582 -1.151 OCBO2|t3 0.207 0.097 2.141 0.032 0.207 0.151 OCBO3|t1 -0.479 0.097 -4.932 0.000 -0.479 -0.356 OCBO3|t2 1.234 0.114 10.787 0.000 1.234 0.918 OCBO3|t3 1.718 0.136 12.660 0.000 1.718 1.278 OCBO3|t4 2.851 0.258 11.058 0.000 2.851 2.121 OCBO4|t1 0.261 0.090 2.911 0.004 0.261 0.208 OCBO4|t2 1.567 0.125 12.540 0.000 1.567 1.245 OCBO4|t3 1.895 0.149 12.696 0.000 1.895 1.505 OCBO5|t1 -0.511 0.101 -5.076 0.000 -0.511 -0.368 OCBO5|t2 0.921 0.108 8.500 0.000 0.921 0.663 OCBO5|t3 1.633 0.135 12.116 0.000 1.633 1.175 OCBO5|t4 2.793 0.271 10.312 0.000 2.793 2.010 OCBO6|t1 -2.241 0.176 -12.748 0.000 -2.241 -1.624 OCBO6|t2 -0.920 0.108 -8.520 0.000 -0.920 -0.667 OCBO6|t3 1.033 0.111 9.317 0.000 1.033 0.748 OCBO7|t1 -2.124 0.164 -12.943 0.000 -2.124 -1.712 OCBO7|t2 -1.038 0.102 -10.142 0.000 -1.038 -0.837 OCBO7|t3 1.139 0.105 10.823 0.000 1.139 0.919

Variances: Estimate Std.Err z-value P(>|z|) Std.lv Std.all .PP1 1.000 1.000 0.546 .PP2 1.000 1.000 0.514 .PP3 1.000 1.000 0.527 .PP4 1.000 1.000 0.449 .PP5 1.000 1.000 0.457 .PP6 1.000 1.000 0.567 .PP7 1.000 1.000 0.593 .PP8 1.000 1.000 0.382 .PP9 1.000 1.000 0.484 .PP10 1.000 1.000 0.524 .IRB1 1.000 1.000 0.146 .IRB2 1.000 1.000 0.158 .IRB3 1.000 1.000 0.105 .IRB4 1.000 1.000 0.090 .IRB5 1.000 1.000 0.796 .IRB6 1.000 1.000 0.213 .IRB7 1.000 1.000 0.321 .OCBI1 1.000 1.000 0.394 .OCBI2 1.000 1.000 0.480 .OCBI3 1.000 1.000 0.682 .OCBI4 1.000 1.000 0.409 .OCBI5 1.000 1.000 0.397 .OCBI6 1.000 1.000 0.411 .OCBI7 1.000 1.000 0.429 .OCBO1 1.000 1.000 0.589 .OCBO2 1.000 1.000 0.529 .OCBO3 1.000 1.000 0.554 .OCBO4 1.000 1.000 0.631 .OCBO5 1.000 1.000 0.518 .OCBO6 1.000 1.000 0.525 .OCBO7 1.000 1.000 0.650 PR 0.831 0.089 9.389 0.000 1.000 1.000 IRB 5.843 0.943 6.194 0.000 1.000 1.000 OCBI 1.538 0.192 8.020 0.000 1.000 1.000 OCBO 0.699 0.073 9.575 0.000 1.000 1.000

Scales y*: Estimate Std.Err z-value P(>|z|) Std.lv Std.all PP1 0.739 0.739 1.000 PP2 0.717 0.717 1.000 PP3 0.726 0.726 1.000 PP4 0.670 0.670 1.000 PP5 0.676 0.676 1.000 PP6 0.753 0.753 1.000 PP7 0.770 0.770 1.000 PP8 0.618 0.618 1.000 PP9 0.696 0.696 1.000 PP10 0.724 0.724 1.000 IRB1 0.382 0.382 1.000 IRB2 0.397 0.397 1.000 IRB3 0.325 0.325 1.000 IRB4 0.301 0.301 1.000 IRB5 0.892 0.892 1.000 IRB6 0.461 0.461 1.000 IRB7 0.567 0.567 1.000 OCBI1 0.628 0.628 1.000 OCBI2 0.693 0.693 1.000 OCBI3 0.826 0.826 1.000 OCBI4 0.640 0.640 1.000 OCBI5 0.630 0.630 1.000 OCBI6 0.641 0.641 1.000 OCBI7 0.655 0.655 1.000 OCBO1 0.767 0.767 1.000 OCBO2 0.728 0.728 1.000 OCBO3 0.744 0.744 1.000 OCBO4 0.794 0.794 1.000 OCBO5 0.720 0.720 1.000 OCBO6 0.725 0.725 1.000 OCBO7 0.806 0.806 1.000

On Jun 22, 2018, at 6:08 PM, Terrence notifications@github.com wrote:

You can request any fit measures you want using the fit.measures= argument. Yes, and you can investigate the standardized model parameters to see how different the means are. The first group's mean == 0, so any other group means are therefore already the difference from the first (reference) group. In a standardized metric (the std.all column in the summary(fit, std = TRUE) output), that is interpreted as Cohen's d. Likewise, the difference between any other groups' (e.g., groups 2 and 3) standardized means are the Cohen's d for that group comparison. You can obtain the fitted model from the list of results returned by the function. mi <- measurementInvarianceCat(...) summary(mi$fit.means, std = TRUE) — You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/simsem/semTools/issues/33#issuecomment-399606673, or mute the thread https://github.com/notifications/unsubscribe-auth/AEgWFtTxTJEDQ-xI7IB7iY66YqkI-IiDks5t_XjlgaJpZM4UtwBr.

TDJorgensen commented 6 years ago

All mean-structure parameters are reported under Intercepts. You can easily distinguish the indicator intercepts from unconditional means of latent common factors because the indicators' names are preceded by a dot, whereas exogenous factor names are not. In this output, all the means are fixed to zero in both groups. But when loadings and thresholds are constrained to equality across groups, you will be able to freely estimate the latent means in group 2. The measurementInvarianceCat() should specify this by default, if you are still using it (see mi$fit.thresholds in the output list object).