Are complex causal models less likely to be true than simple ones? A critical comment on Trafimow (2017)
- PDF / 176,765 Bytes
- 4 Pages / 595.276 x 790.866 pts Page_size
- 32 Downloads / 159 Views
Are complex causal models less likely to be true than simple ones? A critical comment on Trafimow (2017) Marco Del Giudice 1 Accepted: 2 September 2020 # The Psychonomic Society, Inc. 2020
Abstract Trafimow (2017) used probabilistic reasoning to argue that more complex causal models are less likely to be true than simpler ones, and that researchers should be skeptical of causal models involving more than a handful of variables (or even a single correlation coefficient) [Trafimow, D. (2017). The probability of simple versus complex causal models in causal analyses. Behavior Research Methods, 49, 739–746]. In this comment, I point out that Trafimow’s argument is misleading, and reduces to the observation that more informative models (that make definite statements about certain causal relations) are less likely to be true than less informative models (that remain silent about those relations, by omitting some variables from consideration). This correct but trivial statement does not deliver the epistemological leverage promised in the paper. When complexity is evaluated with reasonable criteria (such as the number of nonzero effects in alternative models involving the same variables), more complex models can be more, less, or equally likely to be true compared with simpler ones. I also discuss Trafimow’s claim that, if a model is unlikely to be true a priori, researchers will seldom be able to gather evidence of sufficient quality to support it; in practice, even low-probability models can receive strong support without the need for extraordinary evidence. Researchers should evaluate the plausibility of causal models on a case-by-case basis, and be skeptical of overblown claims about the dangers of complex theories. Keywords Causal models . Epistemology . Likelihood ratio . Occam’s razor . Probability
In “The probability of simple versus complex causal models in causal analyses”, Trafimow (2017) used probabilistic reasoning to argue that “a simple causal model based on a single correlation coefficient is more likely to be true than a complex causal model based on several correlation coefficients” (p. 743). The author wondered: “given that a simple causal model is much more likely to be true than is a complex one, why do journal editors and reviewers favor complexity?” (p. 743). After suggesting that the answer may lie in the cognitive limitations and biases of scientists (e.g., the conjunction fallacy), he concluded that, to be rational, “researchers should be more open to simple causal models based on a single correlation coefficient or they should be less open to complex causal models based on many correlation coefficients” (p. 745; emphasis in the original).
* Marco Del Giudice [email protected] 1
Department of Psychology, University of New Mexico, Logan Hall, 2001 Redondo Dr. NE, Albuquerque, NM 87131, USA
Even if a given model is unlikely to be true a priori, the evidence in its favor can be so strong that it leads to a high posterior probability. Trafimow (2017) considered this Bayesian argument, but noted th
Data Loading...
