Identification and Sensitivity Analysis for Multiple Causal Mechanisms: Revisiting Evidence from Framing Experiments

By Kosuke Imai, Teppei Yamamoto. Political Analysis, 2013

396 citations recorded by the source corpus.

Social scientists are often interested in testing multiple causal mechanisms through which a treatment affects outcomes. A predominant approach has been to use linear structural equation models and examine the statistical significance of the corresponding path coefficients. However, this approach implicitly assumes that the multiple mechanisms are causally independent of one another. In this article, we consider a set of alternative assumptions that are sufficient to identify the average causal mediation effects when multiple, causally related mediators exist. We develop a new sensitivity analysis for examining the robustness of empirical findings to the potential violation of a key identification assumption. We apply the proposed methods to three political psychology experiments, which examine alternative causal pathways between media framing and public opinion. Our analysis reveals that the validity of original conclusions is highly reliant on the assumed independence of alternative causal mechanisms, highlighting the importance of proposed sensitivity analysis. All of the proposed methods can be implemented via an open source R package, mediation.

Subfields: Education, Sociology and Political Science, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/32fa962ebd7580dc5d6a08688d02b9a532403703

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Education

In Sociology and Political Science

In Statistics and Probability

Citations in the corpus

Explaining Causal Findings Without Bias: Detecting and Assessing Direct Effects (2016, background): “As Imai, Keele, and Yamamoto (2010) point out, no intermediate confounders “is an important lim- 7 Imai and Yamamoto (2013) present a method for causal mediation in the face of intermediate confounders, which they refer to as multiple causal mechanisms.”
Advances in mediation analysis: a survey and synthesis of new developments. (2015, methodology): “Imai et al. (2013) and Imai & Yamamoto (2013) introduce new experimental designs for scenarios where it is difficult to manipulate M directly but plausible to encourage participants to adopt certain values of M (i.e., imperfect manipulation).”
Mediation Analysis: A Practitioner's Guide. (2016, background): “For further discussion, the reader is it y”
Mediation Analysis with Multiple Mediators (2014, background): “With a few notable exceptions (Avin et al., 2005; Albert and Nelson, 2011; Imai and Yamamoto, 2012; Zheng and van der Laan, 2012) most of this literature has been restricted to the setting of a single mediator.”
IDENTIFYING CAUSAL MECHANISMS (PRIMARILY) BASED ON INVERSE PROBABILITY WEIGHTING (2014, methodology): “Also, OLS both with and without treatment-mediator interaction is generally biased under any b, because V is controlled for in the regression as if it was not affected by D.”
The Influence of Emotion on Trust (2016, background): “Other issues, such as what to do if different dimensions of emotional arousal are interrelated, require more sophistical statistical treatment (Imai and Yamamoto 2013), as well as better measurement tools.”
mediation: R Package for Causal Mediation Analysis (2014, methodology): “The framework of Imai and Yamamoto (2013) is based on the following varying coeﬃcient linear structural equations model, where E ( ε 2 i ) = E ( ε 3 i ) = 0 without loss of generality.”