Course – Multivariate analysis in biomedical studies: Logistic regression models (jamovi)

This course is designed to provide solid, applied training in logistic regression, covering its main variants: binary, ordinal, and multinomial. Throughout the course, you will learn to use these models to analyze categorical response variables, a common situation in biomedical research, where outcomes are often expressed as the presence or absence of disease, severity levels, risk categories, or types of clinical response. Logistic regression offers a flexible, interpretable, and statistically robust model.

Logistic regression allows us to model probabilities and study how they change depending on one or more explanatory factors. This course will cover how to interpret the coefficients (and odds ratios), how to build and select models, how to evaluate their assumptions, how to incorporate confounding variables, and how to explore relationships between predictors to better describe the behavior of the data. 

The course will combine effect estimation and inference with probability prediction, paying attention to both statistical interpretation and the clinical relevance of the results. Finally, a significant portion will be dedicated to learning how to correctly interpret and report findings, so that analyses can be communicated clearly, rigorously, and usefully in scientific and healthcare settings. 

In this course, we use Jamovi software (jamovi.org). It's a software that offers a graphical user interface (GUI) using menus and dialog boxes. This type of software is ideal for people who only analyze data occasionally, as it only requires them to recognize what they need from the menus, rather than having to remember programming statements.

(*) The course is aimed at healthcare personnel working in any of the specialized or primary healthcare centers of the SALUD and the Department, as well as members of research groups of the IIS Aragón.