Consider moving beyond the Sobel test and the Baron/Kenny approach with a direct test of the mediation effect. Hayes and his colleagues have written extensively on how to do this (e.g., http://www.tandfonline.com/doi/abs/10.1080/03637750903310360). Though Hayes's mediation text (http://afhayes.com/introduction-to-mediation-moderation-and-conditional-process-analysis.html) is primarily based on path analysis, the principles can be applied to SEM.

In the attached article, entitled: "For fathersraisingchildrenwithautism,docopingstrategies

mediate ormoderatetherelationshipbetweenparenting

stress andqualityoflife?"

I have used Sobel test in mediation and moderation.

Hope it help

Consider moving beyond the Sobel test and the Baron/Kenny approach with a direct test of the mediation effect. Hayes and his colleagues have written extensively on how to do this (e.g., http://www.tandfonline.com/doi/abs/10.1080/03637750903310360). Though Hayes's mediation text (http://afhayes.com/introduction-to-mediation-moderation-and-conditional-process-analysis.html) is primarily based on path analysis, the principles can be applied to SEM.

In my view, research is a very open-minded activity and any way you get providing support your thesis is helpful and, hence, welcome. As methodological researchers are continuously improving statistics, methods as Sobel's test are a little bit updated. They are still valid, but increasingly unreliable to add support to your research predictions.

thank you  Mr. Solomon Kurz for providing the link

Regards

AMOS provides several methods for testing the significance of indirect effects. The most "preferred" of these is probably the bootstrapping option, which you can select in the Analysis Properties window. I would say that reporting the bootstrapped significance test is (probably) a better approach than the Sobel's Test. 

Kindly find the attached guide on mediation analysis

Article Making Sense of Mediating Analysis: A Marketing Perspective