Interpretation of interaction: A review

By Amy Berrington de González, D. R. Cox. 2007

131 citations recorded by the source corpus.

Several different types of statistical interaction are defined and distinguished, primarily on the basis of the nature of the factors defining the interaction. Illustrative examples, mostly epidemiological, are given. The emphasis is primarily on interpretation rather than on methods for detecting interactions.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/1c7edfb84c76f169360430558036001d1861dc33

How researchers see this paper

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

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

The Connection Between Varying Treatment Effects and the Crisis of Unreplicable Research (2015, background): “We can say it in a number of different ways: •• A statistical hypothesis (for example, θ = 0 or θ1 = θ2 = . . . = θk) is much more specific than a scientific hypothesis (for example, that a certain comparison averages to zero in the population, or that any net effects are too small to be detected).”
Decomposing Treatment Effect Variation (2016, methodology): “Departing from much of the previous literature (e.g., Cox 1984; Berrington de González and Cox 2007; Crump et al. 2008), we study the properties of the least-square estimator under complete randomization, without assuming that model (8) is correctly specified.”
Causal Interaction in Factorial Experiments: Application to Conjoint Analysis (2018, unknown): “= Our article builds on the causal inference and experimental design literatures that are concerned about interaction e ff ects (see, e.g., Cox 1984; Jaccard and Turrisi 2003; de González and Cox 2007; VanderWeele and Knol 2014).”
A re‐evaluation of fixed effect(s) meta‐analysis (2018, background): “First, β is an average effect, averaging over the whole population contributing to the analysis, and therefore may not capture the association in any subgroups (Berrington de González and Cox, 2007).”