Checking assumptions: Advancing the analysis of sex and gender in human health and psychological sciences

Sex and gender are dissociable, multi-component variables. Focusing on the analytic problems associated with dichotomising continuous variables, we synthesize a new approach to collecting and analysing sex and gender data in health research, in contrast to the conventional use of dichotomous tickboxes to code sex/gender.Methods. Using a literature review and data simulations in R, we examined the magnitude of the statistical and methodological problems associated with the use of a single dichotomised sex/gender variable, including construct validity, predictive validity, measurement error, residual confounding, misclassification and bias due to cut points, power, and representative sampling.Results. Using the dichotomous sex/gender predictor rather than a continuous sex/gender predictor increased residual confounding up to 80% and misclassification of individual participants up to 50%. Further, there was substantial bias in model parameters when continuous sex/gender variables were dichotomised. Finally, we found that using the dichotomous sex/gender predictor decreased power, in some cases by more than 50%.Conclusions. Using a dichotomous sex/gender predictor in place of a continuous sex/gender predictor has profound impacts on the statistical model and the validity of inferences drawn from such a model. We describe measurement and analytic approaches to reduce the statistical problems related to a dichotomised sex/gender analysis.