Hey guys,

first, if you allow some clarifications on terms: What you mean is not the direct effect but the total effect. Second, if I may disagree with Matt, you don't have to show a total effect with leaving out the mediator. Your idea is basically the old Baron-Kenny-approach which is nonsense and outdated. That is, there doesn't have to be an overall relationship (=total effect) in order to mediation being present. The reason is that sometimes indirect and direct effects have opposing signs; omitting the mediator leads to both effects cancel each other out. That will leads to the apparently paradoxical situation that the overall/total effect is zero while there is still an indirect effect.

Briefly, set up your model with the mediator in it and test for the indirect effect. Much more important would be, whether you can incorporate variables that serve as instruments for the mediator, the IV or both. These are variables that have only an effect of their respective target and not on the variable further down the stream. This helps to reduce the possibility of confounding bias. If you do not have instruments, the quality of the model stands and falls with control variables that represent potential confounders of the X-M or M-Y link.

HTH

Holger

Hayes, A. F. (2009). Beyond baron and kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76(4), 408-420.

Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: Examples from the child-clinical and pediatric psychology literatures. Journal of Consulting and Clinical Psychology, 65(4), 599-610.

James, L. R., Mulaik, S. A., & Brett, J. M. (2006). A tale of two methods. Organizational Research Methods, 9(2), 233-244.

Kline, R. B. (2015). The mediation myth. Basic and Applied Social Psychology, 37(4), 202-213. doi:10.1080/01973533.2015.1049349

MacKinnon, D. P., & Pirlott, A. G. (2015). Statistical approaches for enhancing causal interpretation of the m to y relation in mediation analysis. Personality and Social Psychology Review, 19(1), 30-43. doi:10.1177/1088868314542878

Preacher, K. J. (2015). Advances in mediation analysis : A survey and synthesis of new developments. Annual Review of Psychology, 825-852. doi:10.1146/annurev-psych-010814-015258

Thoemmes, F. (2015). Reversing arrows in mediation models does not distinguish plausible models. Basic and Applied Social Psychology, 37(4), 226-234. doi:10.1080/01973533.2015.1049351

Murayama, K., & Elliot, A. J. (2012). The competition-performance relation: A meta-analytic review and test of the opposing processes model of competition and performance. Psychological Bulletin, 138(6), 1035-1070. doi:10.1037/a0028324

I'd say yes, assuming you have no additional controls in direct effect.

Hope this helps,

Matt

Hey guys,

first, if you allow some clarifications on terms: What you mean is not the direct effect but the total effect. Second, if I may disagree with Matt, you don't have to show a total effect with leaving out the mediator. Your idea is basically the old Baron-Kenny-approach which is nonsense and outdated. That is, there doesn't have to be an overall relationship (=total effect) in order to mediation being present. The reason is that sometimes indirect and direct effects have opposing signs; omitting the mediator leads to both effects cancel each other out. That will leads to the apparently paradoxical situation that the overall/total effect is zero while there is still an indirect effect.

Briefly, set up your model with the mediator in it and test for the indirect effect. Much more important would be, whether you can incorporate variables that serve as instruments for the mediator, the IV or both. These are variables that have only an effect of their respective target and not on the variable further down the stream. This helps to reduce the possibility of confounding bias. If you do not have instruments, the quality of the model stands and falls with control variables that represent potential confounders of the X-M or M-Y link.

HTH

Holger

Hayes, A. F. (2009). Beyond baron and kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76(4), 408-420.

Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity in the study of mediators and moderators: Examples from the child-clinical and pediatric psychology literatures. Journal of Consulting and Clinical Psychology, 65(4), 599-610.

James, L. R., Mulaik, S. A., & Brett, J. M. (2006). A tale of two methods. Organizational Research Methods, 9(2), 233-244.

Kline, R. B. (2015). The mediation myth. Basic and Applied Social Psychology, 37(4), 202-213. doi:10.1080/01973533.2015.1049349

MacKinnon, D. P., & Pirlott, A. G. (2015). Statistical approaches for enhancing causal interpretation of the m to y relation in mediation analysis. Personality and Social Psychology Review, 19(1), 30-43. doi:10.1177/1088868314542878

Preacher, K. J. (2015). Advances in mediation analysis : A survey and synthesis of new developments. Annual Review of Psychology, 825-852. doi:10.1146/annurev-psych-010814-015258

Thoemmes, F. (2015). Reversing arrows in mediation models does not distinguish plausible models. Basic and Applied Social Psychology, 37(4), 226-234. doi:10.1080/01973533.2015.1049351

Murayama, K., & Elliot, A. J. (2012). The competition-performance relation: A meta-analytic review and test of the opposing processes model of competition and performance. Psychological Bulletin, 138(6), 1035-1070. doi:10.1037/a0028324