The mistreatment of covariate interaction terms in linear model analyses of behavioural and evolutionary ecology studies

In behavioural and evolutionary ecology, there are often large phenotypic differences between individuals in, for example, body size or large variation in abiotic conditions such as temperature, between measurements. This often inevitable source of variation may mask any effect of experimental treatment as it can have a large impact on the dependent variable of interest. In such cases, conventional statistical comparisons may have much lower power than desired. The inclusion of covariates in statistical analyses has proven a powerful method to control for such nonrandom differences between individual data points that cannot be controlled experimentally (Huitema 1980). To make correct conclusions, it is important to understand the basic assumptions underlying such a covariate analysis. In this paper I argue that this has evidently not been completely acknowledged in the scientific community. Sophisticated models relating responses to both one or more continuous covariates and one or more factors can be problematic. Factor is here used in the meaning of a categorical independent variable and its value divides individuals into discrete groups or categories, for instance experimental treatments. In the following, I use a simple one-factor ANCOVA design as an example, but the same general problem outlined here applies to all linear models with one or more covariates, including generalized linear models (GLIM) such as logistic regressions and even survival analysis. The basic design of a one-factor fixed effect ANCOVA can be written as:

[1]  Joachim G. Frommen,et al.  Inbreeding avoidance through non-random mating in sticklebacks , 2006, Biology Letters.

[2]  Bradley E. Huitema,et al.  The analysis of covariance and alternatives , 1980 .

[3]  J. Sinacore Multiple regression: Testing and interpreting interactions , 1993 .

[4]  M. Kendall Statistical Methods for Research Workers , 1937, Nature.

[5]  R. Wilcox Pairwise comparisons of J independent regression lines over a finite interval, simultaneous pairwise comparison of their parameters, and the Johnson‐Neyman procedure , 1987 .

[6]  E. Pedhazur Multiple Regression in Behavioral Research: Explanation and Prediction , 1982 .

[7]  Elazar J. Pedhazur,et al.  Measurement, Design, and Analysis: An Integrated Approach , 1994 .

[8]  W. G. Cochran Analysis of covariance: Its nature and uses. , 1957 .

[9]  Shinichi Nakagawa,et al.  Does the badge of status influence parental care and investment in house sparrows? An experimental test , 2007, Oecologia.

[10]  O. Leimar,et al.  The evolution of movements and behaviour at boundaries in different landscapes: a common arena experiment with butterflies , 2003, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[11]  Jacqueline P. Leighton,et al.  Defining Johnson-Neyman Regions of Significance in the Three-Covariate ANCOVA Using Mathematica , 1997 .

[12]  J. Neyman,et al.  Tests of certain linear hypotheses and their application to some educational problems. , 1936 .

[13]  T. Merckx,et al.  Landscape structure and phenotypic plasticity in flight morphology in the butterfly Pararge aegeria , 2006 .

[14]  W. Potts,et al.  MHC–disassortative mating preferences reversed by cross–fostering , 1998, Proceedings of the Royal Society of London. Series B: Biological Sciences.

[15]  J. Reid Experimental Design and Data Analysis for Biologists , 2003 .

[16]  Karen T D'Alonzo,et al.  The Johnson-Neyman Procedure as an Alternative to ANCOVA , 2004, Western journal of nursing research.

[17]  C. R. White,et al.  Allometric Analysis beyond Heterogeneous Regression Slopes: Use of the Johnson‐Neyman Technique in Comparative Biology , 2003, Physiological and Biochemical Zoology.