Confidence Limits for the Indirect Effect: Distribution of the Product and Resampling Methods

By David P. MacKinnon, C. Lockwood, Jason J. Williams. Multivariate Behavioral Research, 2004

7,978 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/3f0e13e4300ea6043102a19da3cfbc0f7e7951bd

How researchers see this paper

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

In Statistics and Probability

Citations in the corpus

Longitudinal Evaluation of a Scale-Up Model for Teaching Mathematics With Trajectories and Technologies (2012, methodology): “The use of this procedure has been found to demonstrate both high power and low Type 1 error rates (MacKinnon, Lockwood, & Williams, 2004 ).”
Emotion Regulation Deficits as Mediators Between Trauma Exposure and Borderline Symptoms (2013, methodology): “Indirect effects were tested by bias-corrected bootstrapped confidence intervals (MacKinnon et al. 2004).”
Resampling and Distribution of the Product Methods for Testing Indirect Effects in Complex Models (2008, methodology): “An extensive simulation study of resampling approaches (MacKinnon et al., 2004) found that the percentile bootstrap outperformed the z test, but confidence limits were biased, with a greater proportion of true values falling to the right of the percentile interval.”
The mediating role of positive and negative affect in the situational motivation-performance relationship (2013, background): “In the present study, the confidence interval was bias corrected given that this correction is believed to improve power and Type 1 error rates (MacKinnon et al. 2004).”
When and Why People Engage in Different Forms of Proactive Behavior: Interactive Effects of Self-construals and Work Characteristics (2017, methodology): “Mackinnon, Lockwood, and Williams (2004) and Pituch, Whittaker, 10 In this model, we treated items for all constructs as continuous variables. and Stapleton (2005) provide evidence that the distribution-of-the-product method is more accurate than other methods in constructing confidence limits for…”
Estimating and Comparing Specific Mediation Effects in Complex Latent Variable Models (2012, methodology): “…for establishing BC bootstrap confidence intervals avoid the problems that stem from the normality assumption—an assumption that is often violated (MacKinnon et al., 2004), thus enabling them to outperform not only those approaches that rely on null hypothesis testing (e.g., the Sobel test;…”
The Mediating Role of Emotions (2015, methodology): “It is thus a nonparametric resampling procedure and an estimation strategy that improves power of a model, as it accurately measures the empirical sampling distribution of the test statistic (see MacKinnon, Lockwood, & Williams, 2004; Preacher & Hayes, 2004; Shrout & Bolger, 2002).”