metaSEM
http://openmx.ssri.psu.edu/taxonomy/term/44/0
enThis forum is about the metaSEM package for meta-analysis
http://openmx.ssri.psu.edu/thread/2946
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p><a href="http://courses.nus.edu.sg/course/psycwlm/internet">Mike Cheung's</a> metaSEM package is introduced <a href="http://courses.nus.edu.sg/course/psycwlm/Internet/metaSEM">here</a></p>
<p>Post questions to this forum</p>
</div>
</div></div></div>Sat, 26 Apr 2014 21:26:16 +0000tbates2946 at http://openmx.ssri.psu.eduTSSEM near 0 heterogeneity variances and Hessian matrix error
http://openmx.ssri.psu.edu/node/4244
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Dear Mike and other users,</p>
<p>I am doing random effects TSSEM and had run into problems with openmx status=6 with tssem1(). In earlier communication with Mike, he noted that some of the heterogeneity variances (I2) were near 0 (e.g., .0000, .0001) and said that tssem1() does not handle them well. Thus he suggested fixing the near 0 heterogeneity variances to 0, and letting the larger heterogeneity variance vary with starting value 0.01, then running tssem1() with this user-defined structure:</p>
<h2>Try user-defined structure for the random effects</h2>
<h2>Fix the elements in 1, 2, 3, 4, and 6</h2>
<p>RE <- Diag(c(0,0,0,0, "0.01<em>Tau2_5", 0, "0.01</em>Tau2_7", "0.01<em>Tau2_8", "0.01</em>Tau2_9", "0.01*Tau2_10"))<br />
RE<br />
random1 <- tssem1(vector, n, method="REM", RE.type="User",<br />
RE.constraints = RE)</p>
<p>Openmx status became 0 after doing this. I will need to conduct similar analyses with several other datasets and thus have some follow-up questions about fixing heterogeneity variances in further analyses.</p>
<ol>
<li>
<p>Mike suggested previously that I should fix a heterogeneity variance only if there are "NA in the estimated heterogeneity or the OpenMx status is not 0 or 1," and only if the variance is as small as 1e-8 or 1e-10. But how do I check that the heterogeneity variance for any intercept is <1e-8, since the tssem1() output shows only up to 4 decimal places?</p>
</li>
<li>
<p>For some of my analyses, openmx status is 0 after running/rerunning tssem1() and running tssem2(), but I get the following error message after running tssem2():</p>
</li>
</ol>
<p>hIn .solve(x = objectmx.fit@outputmx.fit@outputcalculatedHessian, parameters = my.name) :<br />
Error in solving the Hessian matrix. Generalized inverse is used. The standard errors may not be trustworthy.</p>
<p>Thus I fixed to 0 the heterogeneity variances between .0000 and .0003 in the tssem1() output, and then reran tssem1() with user-defined structure, which made the error message go away. I can't tell how small the .0000 really is, but the .0003 is >1e-8. So if openmx status=0, is it better to set heterogeneity variances to 0 to avoid the error, or better not to set to 0 if the variances are >1e-8 but get the error?</p>
<p>Does the error really matter for interpretation of findings, since the Std Error is NA in my tssem2() output anyway, and I use the lbound and ubound to interpret whether the coefficients, indirect effect, and direct effect are significant?</p>
<p>Any suggestions or advice would be much appreciated!<br />
Mei Yi</p>
</div>
</div></div></div>Wed, 22 Mar 2017 21:58:19 +0000myng4244 at http://openmx.ssri.psu.eduNA in Indirect Effects for 95% likelihood based CI’s metaSEM
http://openmx.ssri.psu.edu/node/4237
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Plain text icon" title="text/plain" src="/modules/file/icons/text-plain.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/R%20Code.txt" type="text/plain; length=9677" title="R Code.txt">R Code</a></span></td><td>9.45 KB</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="Image icon" title="image/png" src="/modules/file/icons/image-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/Model.png" type="image/png; length=23150">Model.png</a></span></td><td>22.61 KB</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Dear Mike and Others,</p>
<p>I am trying to estimate a random effects tssem for my dissertation.I have read your book and related papers. I am following the wonderful resources provided by you and your team. My goal is to perform some moderator analyses using categorical variables, after I successfully run the tssem model.</p>
<p>I am attaching my R script and the structural model image. In this data and model, I found 2 issues, and have 2 clarifications.</p>
<ol>
<li>
<p>Some of the 95% likelihood based CI’s are shown as “NA”. This happens mainly for the indirect effects – for example for my main tssem2 model – the first one in my R code. This issue is more pronounced when I run moderator analyses and estimate two tssem2 models (split based on the categorical moderator )after I perform the moderator analysis. In these cases, the lbound and ubound values of even the direct effects are showing as “NA”. Can you please let me know if I have set up anything wrong with respect to my model specification or data. Please let me know if and how I have to use starting values from the prior estimation?</p>
</li>
<li>
<p>I get the following warning message when I run some of the tssem2 models. For example, in my first moderator analysis with variable “tc”.</p>
</li>
</ol>
<p>Warning message:<br />
In .solve(x = object$mx.fit@output$calculatedHessian, parameters = my.name) :<br />
Error in solving the Hessian matrix. Generalized inverse is used. The standard errors may not be trustworthy.</p>
<p>I assume I can ignore this warning given that I am primarily using 95% likelihood based CI’s, and provided R can estimate these 95% likelihood based CI’s for all my parameters.</p>
<ol start="3">
<li>I tried to not use the intervals="LB" option and see if I atleast get standard errors. Though I was successful in getting the standard errors and the CI for the direct effects, I could not get them for the indirect effects. Moreover, due to the following warning message, I was not sure if I can report them for review to a top journal.</li>
</ol>
<p>Warning message:<br />
In vcov.wls(object, R = R) :<br />
Parametric bootstrap with 50 replications was used to approximate the sampling covariance matrix of the parameter estimates. A better approach is to use likelihood-based confidence interval by including the intervals.type="LB" argument in the analysis.</p>
<p>My question is , can I report these std errors? Or can I increase the replications? How do I obtain the std errors for indirect effects when we do not specify the intervals="LB" option?</p>
<ol start="4">
<li>This is a clarification regarding my setup of the S matrix. In my model (please see the figure attached), since I am not explicitly modeling the link between T to J or vice versa, I wanted to correlate them. Can you please verify if my S matrix makes sense in this regard? Is it okay if I do not correlate them?</li>
</ol>
<p>Your response will greatly help me in completing my manuscript. Thanks in advance for your help.</p>
<p>Regards,<br />
Srikanth Parameswaran</p>
</div>
</div></div></div>Wed, 08 Mar 2017 04:08:49 +0000Srikanth4237 at http://openmx.ssri.psu.eduCode 6 Error in the second stage
http://openmx.ssri.psu.edu/node/4225
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Binary Data" title="application/octet-stream" src="/modules/file/icons/application-octet-stream.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/code_0.R" type="application/octet-stream; length=859">code.R</a></span></td><td>859 bytes</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="File" title="text/csv" src="/modules/file/icons/text-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/Data.csv" type="text/csv; length=1621">Data.csv</a></span></td><td>1.58 KB</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hi all, Hi Mike,</p>
<p>I really appreciate Mike's guidance in TSSEM.</p>
<p>I recently encountered the code 6 error (http://openmx.ssri.psu.edu/wiki/errors) in the last step of my TSSEM analysis.</p>
<p>I think it might be because there are too many missing values in the data: after I deleted rows with two NAs, it worked.</p>
<p>I am just wondering if there are any alternative solutions for code 6 error in TSSEM? I noticed that mxTryHard can be used to get a better result but I am not so sure about it....</p>
<p>Hope to know more about how to deal with Code 6 Error occurred in TSSEM.</p>
<p>Btw,<br />
My data and code are also attached.</p>
</div>
</div></div></div>Tue, 24 Jan 2017 23:03:47 +0000Arant4225 at http://openmx.ssri.psu.edu"Cov" is not positive definite.
http://openmx.ssri.psu.edu/node/4218
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Binary Data" title="application/octet-stream" src="/modules/file/icons/application-octet-stream.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/R%20Script_0.R" type="application/octet-stream; length=1520">R Script.R</a></span></td><td>1.48 KB</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="Plain text icon" title="text/plain" src="/modules/file/icons/text-plain.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/data_1.txt" type="text/plain; length=5246">data.txt</a></span></td><td>5.12 KB</td> </tr>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Image icon" title="image/png" src="/modules/file/icons/image-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/Model_0.png" type="image/png; length=103532">Model.png</a></span></td><td>101.11 KB</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Dear Mike and Others,</p>
<p>I am trying to estimate a random effects tssem for my paper.I have read your book and related papers. I am following the wonderful resources provided by you and your team. My goal is to perform some moderator analyses using categorical variables, after I successfully run the tssem model.</p>
<p>I am attaching my data, R script and the structural model. In this data and model, the tssem1 did not execute due to the positive definite problem. So, I followed your suggestions in Cheung and Hafdahl (2016), and hence tried adding the options acov = "weighted" and acov = "unweighted". This helped in successfully running the tssem1 model. But the tssem2 model produced the following error:</p>
<p>Error in wls(Cov = pooledS, asyCov = asyCov, n = tssem1.obj$total.n, Amatrix = Amatrix, :<br />
"Cov" is not positive definite.</p>
<p>Indeed, the matrix was not positive definte when I inspected. How do I work around this error? How can I find a subset of studies in my dataset that is positive definite (without loosing the data points)? Interestingly, the tssem1 worked when I ran the with the first 40 studies in this dataset (even without the acov option).</p>
<p>I find the multivariate metasem to be fascinating. I want to strengthen my paper with this rigorous approach to metasem. So, any help in solving my issue would be great.</p>
<p>Thanks in advance for your help.<br />
Regards,<br />
Srikanth Parameswaran</p>
</div>
</div></div></div>Thu, 29 Dec 2016 22:46:50 +0000Srikanth4218 at http://openmx.ssri.psu.eduA question about latent variables in MASEM
http://openmx.ssri.psu.edu/node/4214
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hi Mike</p>
<p>As mentioned before, I am trying to exam a structural model with three mediators by using the lavaan syntax. After review a couple of articles, I have a question which makes me feel confused. In MASEM, the variables included in correlation matrices should be observable variables and we could explore latent variables by using the lavaan syntax in the structural model. However, some authors reported latent variables correlation matrices in some articles. So, if there are correlation matrices of observable variables in some articles and correlation matrices of latent variables in other articles, how could I deal with this situation?</p>
<p>Many thanks,</p>
<p>Yu Xie</p>
</div>
</div></div></div>Thu, 15 Dec 2016 03:27:31 +0000Xie4214 at http://openmx.ssri.psu.eduUse MASEM to explore several potential mediators
http://openmx.ssri.psu.edu/node/4211
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hi Mike,</p>
<p>I have reviewed several MASEM articles as well as the R code. I am excited that it could examine mediation model. However, I found the R code of meta-analytic structural equation modeling which examined only one mediator. Now I want to exam a structural model with three mediators by using MASEM. So, I am just wondering is that possible to use MASEM to explore several potential mediators. If so, please let me how to find the R code.</p>
<p>Thanks a lot,</p>
<p>Yu Xie</p>
</div>
</div></div></div>Fri, 25 Nov 2016 01:34:50 +0000Xie4211 at http://openmx.ssri.psu.eduhow to improve (poor) fit indices
http://openmx.ssri.psu.edu/node/4210
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="PDF icon" title="application/pdf" src="/modules/file/icons/application-pdf.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/diagram%20model%201.pdf" type="application/pdf; length=5358" title="diagram model 1.pdf">diagram model 1</a></span></td><td>5.23 KB</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="File" title="text/csv" src="/modules/file/icons/text-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/data_model1.csv" type="text/csv; length=11048" title="data_model1.csv">data model 1</a></span></td><td>10.79 KB</td> </tr>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Plain text icon" title="text/plain" src="/modules/file/icons/text-plain.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/script%20model%201.txt" type="text/plain; length=2851" title="script model 1.txt">script (tssem) model 1</a></span></td><td>2.78 KB</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hi all, Hi Mike,</p>
<p>I want to test five rival conceptualizations of a construct (service quality) through path analysis with the two stage approach meta-analytic structural equation modelling (with the metaSEM in R). I am using a path analysis instead of a factor analysis since the dimensions' construct are discussed to be formative rather than reflective. In all five models, I have three or four exogeneous variables and one endogeneous variable (the endogeneous variable is the same for all models, that is, behavioral intention).</p>
<p>I was able to run all five models. However, the fit indices of the models was very poor. For example, the following fit indices pertain to the model 1 (four exogeneous variables and one endogeneous variable):</p>
<p>Model 1 - Fitting structural equation models (Stage 2)<br />
Sample size 59832<br />
Chi-squared of target model 1196.1572<br />
DF of target model 6<br />
p value of target model 0.000<br />
Root Mean Square Error of Approximation (RMSEA) 0.0576<br />
Standardised Root Mean Square Residual (SRMR) 0.4498<br />
TLI -0.0343<br />
CFI 0.3794<br />
AIC 1184.1572<br />
BIC 1130.1614</p>
<p>I suspect that it happened because I did not let the exogenous variables to covary in the models. However, when I let them to covary, the models became saturated and I do not have the fit indices anymore (and I need them to compare the five models). Attached, it is the diagram from the model 1 (with exogeneous variables not covarying), the data from model 1, and the script.</p>
<p>Do you have any suggestion what I could do to obtain acceptable fit indices? And should I not present results like those from model 1, right?!</p>
<p>Thank you very much for your time.</p>
<p>Best, Rafael.</p>
</div>
</div></div></div>Thu, 24 Nov 2016 17:16:35 +0000rafael.lionello4210 at http://openmx.ssri.psu.eduNA values in base model output and OpenMx status1: 5
http://openmx.ssri.psu.edu/node/4209
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>I am currently in the process of running a meta analysis using the metaSEM package in OpenMx. As you can see in the first pass at the base model I am getting NA values for Std.Error, lbound, ubound, z value and Pr(>|z|) for both Tau2_2 and Tau2_3. In addition, the OpenMx status1: 5, which I found meant "5: means that the Hessian at the solution is not convex." However I am unclear how to resolve this as a problem. In addition when I run models with both level-2 and level-3 constraints I no longer see this issue, but I don't think it would be appropriate to do model comparisons to a base model that has so many NA values. Any help on how to resolve this issue would be great.</p>
<p>Here is the first few lines of data:</p>
<p>AUTHOR YEAR EXP yFINAL vFINAL typeFINAL<br />
Greenwood 2009 I -1.346327273 0.228655603 G<br />
Greenwood 2009 I 0.220196364 0.224990678 G<br />
Siette 2014 J 0.8974913 0.2625858 MT<br />
Siette 2014 J 1.2197971 0.2732485 MT<br />
Siette 2014 J 0.1487526 0.2503457 MT</p>
<p>> t(aggregate(yFINAL~EXP, data=META_B, FUN=length))<br />
[,1] [,2] [,3] [,4] [,5] [,6] [,7]<br />
EXP "I" "J" "K" "L" "M" "N" "R"<br />
yFINAL " 2" " 3" " 2" " 1" "12" " 2" " 1"</p>
<p>> ## ## Model 0: Random-effects model<br />
> summary( Model0 <- meta3(y=yFINAL, v=vFINAL, cluster=EXP,<br />
+ data=META_B, model.name="3 level") )</p>
<p>Call:<br />
meta3(y = yFINAL, v = vFINAL, cluster = EXP, data = META_B, model.name = "3 level")</p>
<p>95% confidence intervals: z statistic approximation<br />
Coefficients:<br />
Estimate Std.Error lbound ubound z value<br />
Intercept 2.8690e-01 2.2100e-01 -1.4626e-01 7.2005e-01 1.2982<br />
Tau2_2 1.0000e-10 NA NA NA NA<br />
Tau2_3 2.6152e-01 NA NA NA NA<br />
Pr(>|z|)<br />
Intercept 0.1942<br />
Tau2_2 NA<br />
Tau2_3 NA</p>
<p>Q statistic on the homogeneity of effect sizes: 38.8583<br />
Degrees of freedom of the Q statistic: 22<br />
P value of the Q statistic: 0.01464862</p>
<p>Heterogeneity indices (based on the estimated Tau2):<br />
Estimate<br />
I2_2 (Typical v: Q statistic) 0.0000<br />
I2_3 (Typical v: Q statistic) 0.5675</p>
<p>Number of studies (or clusters): 7<br />
Number of observed statistics: 23<br />
Number of estimated parameters: 3<br />
Degrees of freedom: 20<br />
-2 log likelihood: 40.06405<br />
OpenMx status1: 5 ("0" or "1": The optimization is considered fine.<br />
Other values may indicate problems.)</p>
<p>> ## ## Model 1: Testing tau^2_3 = 0<br />
> Model1 <- meta3(y=yFINAL, v=vFINAL, cluster=EXP,<br />
+ data=META_B,<br />
+ RE3.constraints=0, model.name="2 level")<br />
Call:<br />
meta3(y = yFINAL, v = vFINAL, cluster = EXP, data = META_B, RE3.constraints = 0,<br />
model.name = "2 level")</p>
<p>95% confidence intervals: z statistic approximation<br />
Coefficients:<br />
Estimate Std.Error lbound ubound z value Pr(>|z|)<br />
Intercept 0.055543 0.119228 -0.178139 0.289224 0.4659 0.6413<br />
Tau2_2 0.102482 0.113505 -0.119984 0.324948 0.9029 0.3666</p>
<p>Q statistic on the homogeneity of effect sizes: 38.8583<br />
Degrees of freedom of the Q statistic: 22<br />
P value of the Q statistic: 0.01464862</p>
<p>Heterogeneity indices (based on the estimated Tau2):<br />
Estimate<br />
I2_2 (Typical v: Q statistic) 0.3396<br />
I2_3 (Typical v: Q statistic) 0.0000</p>
<p>Number of studies (or clusters): 7<br />
Number of observed statistics: 23<br />
Number of estimated parameters: 2<br />
Degrees of freedom: 21<br />
-2 log likelihood: 45.00194<br />
OpenMx status1: 0 ("0" or "1": The optimization is considered fine.<br />
Other values may indicate problems.)</p>
<p>> Model2 <- meta3(y=yFINAL, v=vFINAL, cluster=EXP,<br />
+ data=META_B,<br />
+ RE2.constraints=0, model.name="tau2_2 EQ 0")<br />
> summary(Model2)<br />
Call:<br />
meta3(y = yFINAL, v = vFINAL, cluster = EXP, data = META_B, RE2.constraints = 0,<br />
model.name = "tau2_2 EQ 0")</p>
<p>95% confidence intervals: z statistic approximation<br />
Coefficients:<br />
Estimate Std.Error lbound ubound z value Pr(>|z|)<br />
Intercept 0.28690 0.25058 -0.20422 0.77801 1.1449 0.2522<br />
Tau2_3 0.26152 0.24783 -0.22423 0.74726 1.0552 0.2913</p>
<p>Q statistic on the homogeneity of effect sizes: 38.8583<br />
Degrees of freedom of the Q statistic: 22<br />
P value of the Q statistic: 0.01464862</p>
<p>Heterogeneity indices (based on the estimated Tau2):<br />
Estimate<br />
I2_2 (Typical v: Q statistic) 0.0000<br />
I2_3 (Typical v: Q statistic) 0.5675</p>
<p>Number of studies (or clusters): 7<br />
Number of observed statistics: 23<br />
Number of estimated parameters: 2<br />
Degrees of freedom: 21<br />
-2 log likelihood: 40.06405<br />
OpenMx status1: 0 ("0" or "1": The optimization is considered fine.<br />
Other values may indicate problems.)</p>
</div>
</div></div></div>Sun, 20 Nov 2016 20:22:20 +0000Rroquet4209 at http://openmx.ssri.psu.eduHeterogeneity indices and constraints on the between-studies variance-covariance matrix
http://openmx.ssri.psu.edu/node/4208
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><p>Hi all, hi Mike,</p>
<p>I am asking a question in reference to this project (<a href="https://osf.io/awz2p/">https://osf.io/awz2p/</a>). I am revising a paper based on the project in response to reviewer comments.</p>
<p>For most analyses in this project, we fit multivariate meta-analytic models, estimating 11 quantities across all studies. Of course, not every study contributes an estimate of each quantity, so the between-studies variance-covariance matrix is constrained such that all variances are equal and all covariances are equal. This ensures that the model is identifiable.</p>
<p>In the first submission, we reported I2 for each meta-analytic quantity as reported by summary.meta(). One of our reviewers asked us to report tau2 as well as an absolute measure of heterogeneity. This seems reasonable.</p>
<p>However, over the course of thinking about heterogeneity, it occurs to me that it might not make sense to report separate heterogeneity indices for each of the 11 estimated meta-analytic quantities -- after all, all the between-studies variances are constrained to be equal.</p>
<p>Does it make sense to report separate heterogeneity indices given that the between-studies variances are constrained to be equal? If not, how would I obtain a single heterogeneity estimate? From Cheung (2008; <a href="https://www.statmodel.com/download/MCheung.pdf">https://www.statmodel.com/download/MCheung.pdf</a>), I'm guessing that I can use these formulas:</p>
<p>H2 = Q/Qdf<br />
I2 = (H2-1)/H2</p>
<p>Assuming that mod is a model fit using meta() in R, I believe this would be:</p>
<p>H2 <- summary(mod)$Q.stat$Q/summary(mod)$Q.stat$Q.df<br />
I2 <- (H2-1)/H2</p>
<p>(P.S.: Mike, your advice on this and other projects over the years has been exceptionally helpful! You'll see that you are acknowledged in the author notes of this project and your work is referenced extensively throughout the paper)</p>
</div></div></div>Mon, 14 Nov 2016 23:42:14 +0000forscher4208 at http://openmx.ssri.psu.eduThe minimum number of studies to examine the mediator using metaSEM
http://openmx.ssri.psu.edu/thread/4190
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Dear Mike,</p>
<p>I'm Yu Xie. Now I am conducting a meta-analysis using metaSEM, examining a mediator between two variables. I only identify 3 studies in this meta-analysis. I am just wondering that is there any minimun number of studies to examine the mediator using metaSEM?</p>
<p>Best wishes,</p>
<p>Yu Xie</p>
</div>
</div></div></div>Tue, 11 Oct 2016 03:12:13 +0000Xie4190 at http://openmx.ssri.psu.eduLooking for a global factor for the latent factors
http://openmx.ssri.psu.edu/thread/4185
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Image icon" title="image/png" src="/modules/file/icons/image-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/Rplot_X.png" type="image/png; length=9150" title="Rplot_X.png">Rplot_X.png</a></span></td><td>8.94 KB</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="Binary Data" title="application/octet-stream" src="/modules/file/icons/application-octet-stream.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/Script_5.R" type="application/octet-stream; length=802" title="Script.R">Script.R</a></span></td><td>802 bytes</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Dear Mike,</p>
<p>thank you very much for your metaSEM-package, I really enjoy working with it!</p>
<p>I tested a model with 8 variables and 3 latent factors and now I would like to check, if there is a common factor for these three latent factors at the superordinate level - is it possible with metaSEM? I was looking for specifications like Gamma matrix in LISREL, but I unfortunately couldn't find anything.</p>
<p>Thank you!</p>
<p>Best wishes,<br />
Nastja</p>
</div>
</div></div></div>Fri, 16 Sep 2016 16:38:59 +0000nastjuscha4185 at http://openmx.ssri.psu.eduPossible bug and question: Variable name interference with dummy variable name & Average effect size
http://openmx.ssri.psu.edu/thread/4182
<div class="field field-name-upload field-type-file field-label-hidden"><div class="field-items"><div class="field-item even"><table class="sticky-enabled">
<thead><tr><th>Attachment</th><th>Size</th> </tr></thead>
<tbody>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="File" title="text/csv" src="/modules/file/icons/text-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/metaSEMCAP.csv" type="text/csv; length=2455" title="metaSEMCAP.csv">metaSEMCAP.csv</a></span></td><td>2.4 KB</td> </tr>
<tr class="even"><td><span class="file"><img class="file-icon" alt="File" title="text/csv" src="/modules/file/icons/text-x-generic.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/metaSEMnocap.csv" type="text/csv; length=2455" title="metaSEMnocap.csv">metaSEMnocap.csv</a></span></td><td>2.4 KB</td> </tr>
<tr class="odd"><td><span class="file"><img class="file-icon" alt="Binary Data" title="application/octet-stream" src="/modules/file/icons/application-octet-stream.png" /> <a href="http://openmx.ssri.psu.edu/sites/default/files/metaSEMnames.R" type="application/octet-stream; length=1329" title="metaSEMnames.R">metaSEMnames.R</a></span></td><td>1.3 KB</td> </tr>
</tbody>
</table>
</div></div></div><div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hi Mike,</p>
<p>I have two questions, both are related to finding the average effect size of categories in covariates with metaSEM. I've been trying to find the correct way to find the average effect sizes and have found a variety of results for different approaches. One of the reason for these differences seem to be related to my first question/issue, and may be a bug. The second question is simply about the correct way to find average effect sizes.</p>
<ol>
<li>When running one of my 3-level mixed-effects models I use the same name for a dummy variable to indicate categories as for a non-related variable in the data set ("Parent"). This seems to interfere with the results of the 3-level mixed-effects model.</li>
</ol>
<p>When using the exact same name for the dummy variable ("Parent") as for the non-related variable in the data set, the intercept effect size is -.55. When giving the dummy variable name low caps ("parent") the intercept effect size of the same analysis is -.46 (same when giving a completely different name). The same difference in effect size appears when changing the name of the non-related variable in the data set. I attached an Rscript and two csv files for you to test this. Both csv data sets are identical, apart from the non-related variable name for which one uses a capital ("Parent") and the other does not ("parent").</p>
<ol start="2">
<li>I want to find the average effect sizes with metaSEM for each category for every covariate in my meta-analysis. I understand the following 3 approaches should give the same average effect sizes (as example I'll use the covariate "Parent", consisting of groups "Mother" and "Father"; note, I use a 3-level mixed-effects model in my analysis):</li>
</ol>
<p>-Run a 3-level mixed-effects model, and use Mother as a reference group, the intercept will give me the average effect size for Mother. I could get the effect size of Father by adding slope1 estimate of those results to the effect size of Mother, or simply run the same model with Father as reference group.<br />
-Run the 3-level mixed effects model while constraining the intercept to 0 and get the effect sizes of Mother and Father both at once.<br />
-Simply run a 3-level mixed-effects model with a data set that ONLY has effect sizes related to Mother, and do the same for Father separately.</p>
<p>The first and second approach mostly give the same or similar effect sizes. The third approach gives different, but more plausible average effect sizes. Which of these methods am I supposed to use to find the average effect sizes of each group with metaSEM?</p>
<p>Thank you kindly in advance Mike.</p>
<p>Best,<br />
Jasper</p>
</div>
</div></div></div>Mon, 12 Sep 2016 12:07:49 +0000jasperd4182 at http://openmx.ssri.psu.edumissing data
http://openmx.ssri.psu.edu/thread/4178
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>How to deal with the missing data using MetaSEM package? (Almost all the variables have missing data and these data are missing at completely random)</p>
</div>
</div></div></div>Wed, 24 Aug 2016 14:54:06 +0000crystalzsq4178 at http://openmx.ssri.psu.eduOverlap of input matrices (missing variables)
http://openmx.ssri.psu.edu/thread/4177
<div class="field field-name-taxonomy-forums field-type-taxonomy-term-reference field-label-above"><div class="field-label">Forums: </div><div class="field-items"><div class="field-item even"><a href="/forums/third-party-software/metasem">metaSEM</a></div></div></div><div class="field field-name-body field-type-text-with-summary field-label-hidden"><div class="field-items"><div class="field-item even"><div class="tex2jax">
<p>Hello,</p>
<p>I am Valerie, 25 years old and absolutely fascinated by applying Metasem respectively TSSEM by using metaSEM. When I was doing my first analysis, I came across a problem concerning the input data respectively the input correlation matrices and did not find an solution so far:</p>
<p>Let's say, I have two studies, which are dedicated to the same topic, but slightly differ in their observed variables, what is also reflected in their correlation matrices.<br />
1) Study1 shows a correlation matrix with the variables x1,x2,x3<br />
2) Study2 shows a corrleation matrix with the variables x1,x2,x4<br />
This data is different when compared to the datasets such as Hunter83 provided in the book (Cheung 2015), because in this case, there is no correlation matrix, which is "complete", as variables (and as a result also the correlations) are missing in the studies.<br />
Then input matrix would look as follows (taking random values for correlation coefficients):<br />
1<br />
0.2 1<br />
0.3 0.5 1<br />
NA NA NA NA<br />
1<br />
0.7 1<br />
NA NA NA<br />
0.4 0.2 NA 1</p>
<p>When I imported the data using the readLowMat() function, it worked out. However, when applying tssem1(), there is an error message in case of method="FEM" (Error in if (!all(isPD)) warning(paste("Group ", (1:no.groups)[!isPD], :<br />
missing value where TRUE/FALSE needed) and in case of method="REM" (Error in if (all.equal(covMatrix, t(covMatrix))) { :<br />
argument is not interpretable as logical). My guess was that the kind of input data is not suitable for TSSEM as there is not "complete matrix" available, when it comes to the equality constraints. But I do not the correct procedure to apply now. I have read about pairwise or listwise deletion, but isn't this an issue, TSSEM is able to avoid? Or do I mix up the cases of missing variables and missign correlations?</p>
<p>I really would appreciate any help and good advice. I send best wishes from Germany!</p>
</div>
</div></div></div>Mon, 22 Aug 2016 16:33:22 +0000grafvale4177 at http://openmx.ssri.psu.edu