[Lecture Notes]Advanced Data Analysis: Structural Equation Modeling

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

Handouts and Overheads

• Overhead: Brief History of SEM
• Summary of Matrix Commands in SPSS
• Summary of Matrix Commands in R
• Matrix Operation Examples in SPSS and R
• Wright's Rules of Tracing and Decomposition of Correlation Matrices
• Data Set Preparation and Reading Raw Data with Mplus
• Path Analysis Example
• Latent Variables
• One-Factor CFA Examples: Mplus and lavaan, Amos Example (.amw file), Amos Output (.AmosOutput file)
• LISREL Notation, Notation Explanation
• Basic Concepts of Fit with Mediational Path Models
• Maximum Likelihood Estimation
• Some Clarifications and Recommendations on Fit Indices
• Nested Models, Model Modifications, and Correlated Errors
• Two-factor Confirmatory Factor Analysis
• Modification Indices Examples: Mplus and lavaan, Amos
• Improper Solutions in SEM
• Missing Data and Missing Data Estimation
• Missing Data Examples: Mplus and lavaan
• Full Structural Equation Model Examples: Mplus and lavaan, Amos (.amw file)
• Alternative Estimation Methods
• Practical Approaches to Dealing with Nonnormal and Categorical Variables
• Examples of Estimates for Nonnormal Data, lavaan
• Invariance Tests in Multigroup SEM
• Mplus Multigroup SEM Example
• Latent Variable Cross-lagged Panel Models
• Mplus Cross-lagged Panel Model Example with Measured Variables
• Mplus Cross-lagged Panel Model Example with Latent Variables
• Mediation Analysis with Two Waves
• Mean Structures
• Overheads: Growth Curves
• Latent Growth Curve Example: Mplus and lavaan
• Second-Order Latent Growth Curve Model
• Second-Order Growth Curve Example Syntax: Mplus and lavaan
• Minimum Sample Size Recommendations
• Advanced Topics and Further Reading
  

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏6 回帖

关键词：Structural Advanced Analysis equation Modeling available presented Matrix

1. title: Multigroup structural model example--All parameters free ;
   
2. data: file=C:\Jason\mplus\semclass\stack1.dat; format=11f1.0;
   
3. listwise=on;
   
4. variable: names = widow panas1 panas2 panas3 panas4 panas5
   
5. panas6 panas7 panas8 panas9 panas10 ;
   
6. grouping is widow (0=notwidow,1=widow);
   
7. missing = blank;
   
8. analysis: type=general; iterations = 200;
   
9. model=nomeanstructure; information=expected;
   
10. model:
    
11. posaff by panas1-panas5*;
    
12. posaff@1;
    
13. negaff by panas6-panas10* ;
    
14. negaff@1;
    
15. ! Note: by default in Mplus, measurement errors and factor correlations are not
    
16. ! constrained to be equal across groups;
    
17. Model notwidow:
    
18. posaff by panas1-panas5*;
    
19. posaff@1;
    
20. negaff by panas6-panas10* ;
    
21. negaff@1;
    
22. Model widow:
    
23. posaff by panas1-panas5*;
    
24. posaff@1;
    
25. negaff by panas6-panas10* ;
    
26. negaff@1;

复制代码

谢楼主分享

Cross-lagged Panel Model of Positive and Negative Social Exchanges

1. title: Cross-lag panel model of positive and negative exchanges;
   
2. data: file=c:\jason\mplus\semclass\long1.dat; format=4f9.6;
   
3. listwise=on;
   
4. variable: names = pos posf neg negf;
   
5. missing=blank;
   
6. analysis: type=general;
   
7. estimator=mlm;
   
8. model: negf on neg pos;
   
9. posf on neg pos;
   
10. neg with pos;
    
11. negf with posf;
    
12. output: stdyx;

复制代码

Latent Variable Cross-lagged Panel Model of Positive and Negative Social Exchanges

1. Without Correlated Errors Over Time
   
2. title: Latent Cross-lag panel model of positive and negative exchanges;
   
3. data: file=c:\jason\mplus\semclass\long2.dat; format=28f9.6;
   
4. listwise=on;
   
5. variable: names = trust emo info tang social comm se
   
6. trustf emof infof tangf socialf commf sef
   
7. ntrust nemo ninfo ntang nsocial ncomm nse
   
8. ntrustf nemof ninfof ntangf nsocialf ncommf nsef;
   
9. missing=blank;
   
10. analysis: type=general;
    
11. estimator=mlm;
    
12. !model=nomeanstructure; information=expected;
    
13. model: pos by trust (1)
    
14. emo (2)
    
15. info (3)
    
16. tang (4)
    
17. social (5)
    
18. comm (6)
    
19. se (7);
    
20. posf by trustf (1)
    
21. emof (2)
    
22. infof (3)
    
23. tangf (4)
    
24. socialf (5)
    
25. commf (6)
    
26. sef (7);
    
27. neg by ntrust (8)
    
28. nemo (9)
    
29. ninfo (10)
    
30. ntang (11)
    
31. nsocial (12)
    
32. ncomm (13)
    
33. nse (14);
    
34. negf by ntrustf (8)
    
35. nemof (9)
    
36. ninfof (10)
    
37. ntangf (11)
    
38. nsocialf (12)
    
39. ncommf (13)
    
40. nsef (14);
    
41. negf on neg pos;
    
42. posf on neg pos;
    
43. neg with pos;
    
44. negf with posf;
    
45. output: stdyx;

复制代码

Latent Growth Curve Example

1. title: Latent Growth Curve Model Example 1;
   
2. data: file=c:\jason\mplus\semclass\growth1.dat; format=3f10.6;
   
3. listwise=on;
   
4. ! I'm using listwise deletion here, but one could argue that this
   
5. ! may lead to biases due to attrition and therefore use FIML for
   
6. ! missing data estimation, perhaps using auxillary varibles to account
   
7. ! for non-ignorable missingness;
   
8. variable: names = emo1 emo2 emo3 ;
   
9. missing=blank;
   
10. analysis: ! by default mean structures are estimated;
    
11. ! Below is the syntax for a typical latent growth model,
    
12. ! where square brackets estimate or set means and intercepts.
    
13. ! It is typical to set individual indicator intercepts to zero
    
14. ! for identification purposes.
    
15. !model: intercep by emo1@1 emo2@1 emo3@1;
    
16. ! slope by emo1@0 emo2@1 emo3@2;
    
17. ! [emo1-emo3@0];
    
18. ! [intercep slope];
    
19. ! Mplus has shortcut syntax for growth models, the following
    
20. ! statements produce the same results as the above statements;
    
21. model: i s | emo1@0 emo2@1 emo3@2;
    
22. output: stdyx ;

复制代码