Dear Paula,

first, let’s get straight that I understood you correctly:

I refer to MacKinnon, Fairchild & Fritz (2007) when talking about inconsistent mediation. Here, the indirect effect (ab) has opposite signs to the direct/total effect (c /c’).

Inconsistent mediation could be present in your example when the mediator variable (upward social comparison) acts, e.g. as a suppressor variable of the total effect.

Since TE = DE + IE = c’ + ab

A straightforward test is to inspect if TEM, M->Y, X->Y and absence of any M->Y confounders affected by X (Pearl, 2014).

You can check the presence of unobserved confounding either through rho, which is the correlation of error-terms between the mediator and outcome model (Imai et al., 2010), or by including a latent variable capturing the correlation.

2) Second, after considering 1) maybe also consider checking for XM-interactions, recent studies suggest integrating XM-interaction tests as a part of sensitivity analysis for mediation (e.g., Mackinnon, Valente & Gonzalez, 2020).

3) Third, ensure that your analysis does not suffer from technical flaws, such as the variables' wrong coding (e.g., direction), common method effects, etc.

Likely there are other aspects that I am currently unaware of. Generally, I also found the paper by Fairchild & McDaniel (2017) intuitive.

Best regards

Ibrahim

• Fairchild, A. J., & McDaniel, H. L. (2017). Best (but oft-forgotten) practices: mediation analysis. The American journal of clinical nutrition, 105(6), 1259–1271. https://doi.org/10.3945/ajcn.117.152546
• Imai, K., Keele, L., & Tingley, D. (2010). A general approach to causal mediation analysis. Psychological Methods, 15(4), 309–334.
• https://doi.org/10.1037/a0020761
• MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation analysis. Annual review of psychology, 58, 593–614. https://doi.org/10.1146/annurev.psych.58.110405.085542
• MacKinnon, D. P., Valente, M. J., & Gonzalez, O. (2020). The Correspondence Between Causal and Traditional Mediation Analysis: the Link Is the Mediator by Treatment Interaction. Prevention science : the official journal of the Society for Prevention Research, 21(2), 147–157. https://doi.org/10.1007/s11121-019-01076-4
• Pearl, J. (2014). Interpretation and identification of causal mediation. Psychological Methods, 19(4), 459–481. https://doi.org/10.1037/a0036434