Table 4. A brief overview of the types of target variables and regression models in MXM.

Target variable type	Regression model or test
Continuous and percentages without zeros	Linear, MM and quantile regression, Pearson and Spearman correlation coefficients
Multivariate continuous	Multivariate linear regression
Compositional data	Multivariate linear regression
(Strictly) positive valued	Gaussian and Gamma regression with a log-link
Percentages with or without zeros	Beta regression and quasi binomial regression
Counts	Poisson, quasi Poisson, negative binomial and zero inflated Poisson regression
Binary	Logistic regression, quasi binomial regression and G 2 test of independence
Nominal	Multinomial regression and G 2 test of independence
Ordinal	Ordinal regression
Number of successes out of trials	Binomial regression
Time-to-event	Cox, Weibull and exponential regression
Matched case-control	Conditional logistic regression
Left censored	Tobit regression
Repeated/clustered, longitudinal	Generalised linear mixed models (GLMM) and Generalised estimating equations (GEE)