Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 1 down vote favorite

I have read pretty much everything I can find on performing a Sobel test on the internet however the huge discrepancy in my unstandardized coefficients makes me feel like I am doing something completely wrong (a = 0.000306, SEa = 0.000067, b = 430037.87. SEb = 258407.87). Did I miss a step somewhere? Thank you for any insights.

http://i.imgur.com/7ktbT.png

share|improve this question
1  
What do you expect us to be able to say without seeing the data? – Michael Chernick Sep 8 '12 at 15:36
I was wondering if simply by looking at that huge discrepancy (a being practically zero and b being in the very high digits) someone could immediately identify that this was incorrect. By your response I sense that this is not the case? There seem to be very few examples of Sobel tests on the internet. Edit: Unfortunately I do not have the RAW data at this moment and this is all I have at the moment. I may be able to provide more data later. Thank you! – BillyHx Sep 8 '12 at 16:15

1 Answer

up vote 4 down vote accepted

If you want to do a Sobel test for assessing the significance of a mediation, don't do it this procedure is considered old fashioned.

From the website of Andrew Hayes (who wrote the SPSS macro for Sobel tests):

"If you intend to use this macro merely to implement the "Baron and Kenny" steps to mediation analysis or the Sobel test, I advise you against this, for bootstrapping has become one of the more highly recommended approaches for inference about indirect effects."

Instead use the bootstrap approach, either implemented for SPSS by Hayes, in R or SPSS by Kristopher Preacher (Preacher & Hayes have developed most together), or use the R mediation packahe by Kosuke Imai and colleagues.

share|improve this answer
Wow, this helped. Thanks! – BillyHx Sep 8 '12 at 17:44

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.