Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits
Abstract Many biological traits covary with body size, resulting in an allometric relationship. Identifying the evolutionary drivers of these traits is complicated by possible relationships between a candidate selective agent and body size itself, motivating the widespread use of multiple regression...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Oxford University Press
2022-06-01
|
| Series: | Evolution Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/evl3.258 |
| _version_ | 1827880205728874496 |
|---|---|
| author | Sam F. Walmsley Michael B. Morrissey |
| author_facet | Sam F. Walmsley Michael B. Morrissey |
| author_sort | Sam F. Walmsley |
| collection | DOAJ |
| description | Abstract Many biological traits covary with body size, resulting in an allometric relationship. Identifying the evolutionary drivers of these traits is complicated by possible relationships between a candidate selective agent and body size itself, motivating the widespread use of multiple regression analysis. However, the possibility that multiple regression may generate misleading estimates when predictor variables are correlated has recently received much attention. Here, we argue that a primary source of such bias is the failure to account for the complex causal structures underlying brains, bodies, and agents. When brains and bodies are expected to evolve in a correlated manner over and above the effects of specific agents of selection, neither simple nor multiple regression will identify the true causal effect of an agent on brain size. This problem results from the inclusion of a predictor variable in a regression analysis that is (in part) a consequence of the response variable. We demonstrate these biases with examples and derive estimators to identify causal relationships when traits evolve as a function of an existing allometry. Model mis‐specification relative to plausible causal structures, not collinearity, requires further consideration as an important source of bias in comparative analyses. |
| first_indexed | 2024-03-12T18:17:30Z |
| format | Article |
| id | doaj.art-b9b9b2056f9643db95e443f8de4360dc |
| institution | Directory Open Access Journal |
| issn | 2056-3744 |
| language | English |
| last_indexed | 2024-03-12T18:17:30Z |
| publishDate | 2022-06-01 |
| publisher | Oxford University Press |
| record_format | Article |
| series | Evolution Letters |
| spelling | doaj.art-b9b9b2056f9643db95e443f8de4360dc2023-08-02T09:03:01ZengOxford University PressEvolution Letters2056-37442022-06-016323424410.1002/evl3.258Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traitsSam F. Walmsley0Michael B. Morrissey1Scottish Oceans Institute, School of Biology, University of St. Andrews East Sands St. Andrews United KingdomDyers Brae House, School of Biology, University of St. Andrews Greenside Pl St. Andrews United KingdomAbstract Many biological traits covary with body size, resulting in an allometric relationship. Identifying the evolutionary drivers of these traits is complicated by possible relationships between a candidate selective agent and body size itself, motivating the widespread use of multiple regression analysis. However, the possibility that multiple regression may generate misleading estimates when predictor variables are correlated has recently received much attention. Here, we argue that a primary source of such bias is the failure to account for the complex causal structures underlying brains, bodies, and agents. When brains and bodies are expected to evolve in a correlated manner over and above the effects of specific agents of selection, neither simple nor multiple regression will identify the true causal effect of an agent on brain size. This problem results from the inclusion of a predictor variable in a regression analysis that is (in part) a consequence of the response variable. We demonstrate these biases with examples and derive estimators to identify causal relationships when traits evolve as a function of an existing allometry. Model mis‐specification relative to plausible causal structures, not collinearity, requires further consideration as an important source of bias in comparative analyses.https://doi.org/10.1002/evl3.258Allometrybrain sizecausal inferencecoevolutioncomparative methodscorrelated response to selection |
| spellingShingle | Sam F. Walmsley Michael B. Morrissey Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits Evolution Letters Allometry brain size causal inference coevolution comparative methods correlated response to selection |
| title | Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| title_full | Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| title_fullStr | Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| title_full_unstemmed | Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| title_short | Causation, not collinearity: Identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| title_sort | causation not collinearity identifying sources of bias when modelling the evolution of brain size and other allometric traits |
| topic | Allometry brain size causal inference coevolution comparative methods correlated response to selection |
| url | https://doi.org/10.1002/evl3.258 |
| work_keys_str_mv | AT samfwalmsley causationnotcollinearityidentifyingsourcesofbiaswhenmodellingtheevolutionofbrainsizeandotherallometrictraits AT michaelbmorrissey causationnotcollinearityidentifyingsourcesofbiaswhenmodellingtheevolutionofbrainsizeandotherallometrictraits |