File Exchange > Data Analysis >    Structural Equation Modeling

Author:
OriginLab Technical Support
Date Added:
10/20/2021
Last Update:
11/6/2024
Downloads (90 Days):
279
Total Ratings:
20
File Size:
329 KB
Average Rating:
File Name:
Structural...ng.opx
File Version:
1.30
Minimum Versions:
License:
Type:
App
Summary:

Perform structural equation modeling optimization.

Screen Shot and Video:
Description:

  • Purpose
    This app is for structural equation modeling optimization.

  • Installation
    Download the file Structural Equation Modeling.opx, and then drag-and-drop onto the Origin workspace. An icon will appear in the Apps gallery window.

    Note: If originpro package is not the latest version, please upgrade it. Click menu Connectivity: Python Packages... to open Python Packages dialog. And then select originpro, and then click Upgrade button to upgrade it.
  • Operation
    Click the app icon from the Apps gallery window. Then a dialog will pop up with multiple tabs. And then specify the data and model, and change other settings in dialog accordingly, then click OK button.

  • Dialog Settings

    • Data Tab: This tab is for specifying data, including input data and prediction data.

      • Input Data: In this box, you can choose the data for the model. The data is from column, and the column can be from Project, Current Folder (Include Subfolders), Current Workbook and Current Worksheet. For each column been selected as input data, it needs to have an identifier, which needs to be the same with the variable name defined in model. And this identifier can be in Long Name, Units, or Comments. When you filter the data by Columns in dropdown, the available columns will be listed in the left Available Columns table, and those have been selected as input data will be listed in the right Selected Columns table.

      • Prediction Data: In this box, you can change the status of Perform Prdiction checkbox to make a prediction or not. If checked, the similar options as Input Data will be shown for prediction data selection.

    • Model Tab: This tab is for model definition. The algorithm of this app is using semopy, and the model definition syntax follows the instruction in this page.

      • Specify Description By: This dropdown is used to tell how to get the model. From File is to load the model description from a file, and Typing is to write the model description in the Description text area below, and Definition will help you to define the model according to the relations, operations and constraints.

      • Description: Show the description of the model. When Specify Description By is Typing, this text area is editable for writing the description manually.

      • Definition: This is a box shown when Specify Description By is Definition, and provides the convenience to defining a model according to the concepts of model mentioned in syntax page. The concepts include relations (regression, measurement, covariance), operations and constraints (latent, ordinal, start value, bound, constraint).

      • Model Parameters: This box provides parameter settings for model, including Mimic lavaan and Baseline Model. If Mimic lavaan is checked, output variables are correlated and not conceptually identical to indicators. lavaan treats them that way, but it's less computationally effective. And if Baseline Model is checked, the model will be set to baseline model. Baseline model here is an independence model where all variables are considered to be independent with zero covariance. Only variances are estimated.

    • Setting Tab: This tab is some settings for the fitting procedure.

      • Fit: In this box, you can select objective function to minimize and optimization method.

        • Objective Function to Minimize: Option include MLW, FIML, ULS, GLS, WLS, and DWLS.

        • Optimization Method: Choose a method for optimizing the model, and available methods include Nelder-Mead, Powell, CG, BFGS, Newton-CG, L-BFGS-B, TNC, COBYLA, SLSQP, Trust-Constr, DogLeg, Trust-NCG, Trust-Exact, and Trust-Krylov.

      • Factor scores Estimation: Check to perform factor scores estimation.

      • Estimate Standardized Coefficients: Check to also output the standardized coefficients.

    • Statistics Tab: This tab provides the options for statistical calcuation of the input data. For more details, please refer to Quantities and Computation Control.

    • Output Tab: Provides options for where to output the results, in New Sheet or sheets in New Book.

    • OK and Cancel Buttons: Click OK button to perform the structural equation modeling optimization, and output the results and close dialog, and click Cancel button to close the dialog only.

Updates:

2024/1/22 v1.3 Support for estimating standardized coefficients
2023/4/25 v1.2 Fix installation issue

Reviews and Comments:
06/10/2025yukiOriginlabHi W.Midori, thank you for the additional information. If we have any updates, we will let you know. Could you please submit a ticket using the link below and provide your email address so that we can contact you as soon as possible?
https://www.originlab.com/restricted/support/newticket.aspx?c=3

06/07/2025Jiasx19790916useful and easy to operation

06/05/2025W.MidoriHello, yukiOriginLab,
Thank you very much for your kindess to report my problem to developers of Origin.
The problem that the SEM app cannot produce model verification indicators including AIC and RMSEA have been accidently solved recently. When I run the SEM model, if the model is built successfully, a dialogue box containing a few lines of program language will jump into the middle of the Origin interface, which goes as follows(the values of indicators shown in the form of scientific notation are generated by a few lines of random data as an example):
Index(['DoF', 'DoF Baseline', 'chi2', 'chi2 p-value', 'chi2 Baseline', 'CFI',
'GFI', 'AGFI', 'NFI', 'TLI', 'RMSEA', 'AIC', 'BIC', 'LogLik'],
dtype='object')
[[ 3.00000000e+00 5.00000000e+00 2.04304964e-08 1.00000000e+00
3.23908753e+00 -7.03662183e-01 9.99999994e-01 9.99999989e-01
9.99999994e-01 -1.83943697e+00 0.00000000e+00 6.00000000e+00
7.19368581e+00 1.85731786e-09]]
And this dialogue box has given me all the indicators I need to verify the cridibility of the model. I had just neglected the information presented in it, that was why I thought Origin had not produce these indicators in SEM app.
Just as a suggestion, I think that it would be much better if you can add these indicators to the result interface of the newest version of SEM app so that users can easily find them. This SEM app is a very convenient tool, and I sincerely hope that you can continuously perfect it:-)

05/29/2025yukiOriginlabHi W.Midori, thank you for the suggestion. We've added this requirement to our feature database, and our developer will work on implementing it. I’ll let you know if there are any updates.

05/28/2025W.MidoriHello, I have used this app to successfully build a model, and optimization terminated successfully. I need to further verify the cridibility of this model so indicators including AIC(Akaike information criterion), RMSEA(root mean square error of approximation) are required. However, it seemed that no such indicators would be directly presented in the SEM result page. So is there any way to calculate these two indicators mentiond above within Origin? Thank you:-)

05/28/2025W.MidoriHello, I have used this app to successfully build a model, and optimization terminated successfully. I need to further verify the cridibility of this model so indicators including AIC(Akaike information criterion), RMSEA(root mean square error of approximation) are required. However, it seemed that no such indicators would be directly presented in the SEM result page. So is there any way to calculate these two indicators mentiond above within Origin? Thank you:-)

11/28/2024abdullahamEXELLENT

11/12/2024 Not working

05/15/2024cornelius.chisambi.42sThis is just too brief. I do not understand. Please put a video and explain into details as well as how to use the results.

05/15/2024cornelius.chisambi.42sThis is just too brief. I do not understand. Please put a video and explain into details as well as how to use the results.

12