Wednesday, November 24, 2004
It is an application of SEM. Here there are three blocks: Stress - independent variable (IV), Cofering - mediator variable (MV), and Epusement - indepedent variable (IV). For more information about mediator, please refer to the classic reference on this topic is Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. This model is implemented with AMOS. The difficulty I face now is how to adjust it according to modification indices.
4 Comments:
A good collection of statistics resource:
http://math.yorku.ca/SCS/StatResource.html
A collection about SEM books:
http://www.nc.gsu.edu/~mkteer/bookfaq.html
Learn to read and explain AMOS's output:
http://www2.chass.ncsu.edu/garson/pa765/semAMOS1.htm
Output of AMOS shows that modification indices exist on eSou < -- > eRes because of highest potential reduction chi-square 66.748. After I add one correlation between them, chi-square reduces a lot and related fit measurments are all much better. Based on this new version, I do adjustment in similar way by adding correlation between eStr < -- > eAC, and the result is good: RMR = 0.116 (compares with 0.711 in the first version); chi-square = 92.785 (df = 28); relative chi-square CMIN/DF = 3.314; GFI = 0.864; AGFI = 0.782.
Post a Comment
<< Home