Wiley StatsRef: Statistics Reference Online 2014
DOI: 10.1002/9781118445112.stat06415
|View full text |Cite
|
Sign up to set email alerts
|

Linear Statistical Models for Causation: A Critical Review

Abstract: We review the basis for inferring causation by means of linear statistical models. Parameters should be stable under interventions, and so should error distributions. There are also statistical conditions that must be satisfied. Stability is difficult to establish a priori , and the statistical conditions are equally problematic. Therefore, causal relationships are seldom to be inferred from a data set by running regressions, unless there is substantial prior knowledge about the mechani… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
3
0

Year Published

2015
2015
2018
2018

Publication Types

Select...
2
1

Relationship

0
3

Authors

Journals

citations
Cited by 3 publications
(3 citation statements)
references
References 69 publications
0
3
0
Order By: Relevance
“…Causation is a related but not identical con-cept as information flow. The notion of causality has many interpretations, depending on the context, from philosophical [68][69][70] , to statistical [71][72][73][74][75] , to dynamical 61,64,66,76 . Here we will avoid the philosophical direction entirely, but note that some of these do coincide with the others.…”
Section: Acknowledgmentsmentioning
confidence: 99%
“…Causation is a related but not identical con-cept as information flow. The notion of causality has many interpretations, depending on the context, from philosophical [68][69][70] , to statistical [71][72][73][74][75] , to dynamical 61,64,66,76 . Here we will avoid the philosophical direction entirely, but note that some of these do coincide with the others.…”
Section: Acknowledgmentsmentioning
confidence: 99%
“…Fourth, the study's findings are based on correlational analysis, which makes it impossible to establish causality. The sort of regression analysis used in this study to test hypotheses is often used to infer causation from association (Cox & Wermuth, 2004;Freedman, 2005). However, researchers should be cautious about trying to infer causality from correlation.…”
Section: Limitations and Directions For Future Researchmentioning
confidence: 99%
“…Neither are MC restricted to concerns over erroneous p < 0.05 significances: It is common (but extremely misguided) practice in the social and behavioral sciences to declare regressor effects as 'different' if one is significant (p < 0.05) and the other is not (p > 0.05), and this is another error that occurs with multiple inferences (Gelman and Stern, 2006;Freedman, 2005).…”
Section: Introductionmentioning
confidence: 96%