Statistics in Health Sciences PDF

B.Sc. Degree in Applied Statistics Statistics in Health Sciences 4. Classification of variables Jose Barreraab [email protected] https://sites.google.com/view/josebarrera a ISGlobal Barcelona Institute for Global Health - Campus MAR b Department of Mathematics (UAB) This work is licensed unde...

B.Sc. Degree in Applied Statistics Statistics in Health Sciences 4. Classification of variables Jose Barreraab [email protected] https://sites.google.com/view/josebarrera a ISGlobal Barcelona Institute for Global Health - Campus MAR b Department of Mathematics (UAB) This work is licensed under a Creative Commons “Attribution-NonCommercial-ShareAlike 4.0 International” license. Statistics in Health Sciences 1 Classification of variables according to the measure type 2 Classification of variables according to the role in the study 3 Types of explanatory variables Jose Barrera (ISGlobal & UAB) Statistics in Health Sciences, 2023/2024 2/5 Classification of variables according to the measure type Quantitative variables Objectively quantifiable: • Continuous: Between two possible values of the variable, we always can find other possible value. Examples: Glucose level in blood; Exposure time to UV (ultraviolet) radiation. • Discrete: Non continuous. Examples: Number of deaths; Number of hospitalizations in a given hospital. Qualitative variables Non objectively quantifiable: • Ordinal: The possible values of the variable can be naturally ordered. Examples: Seriousness of an illness as mild, moderate or grievous; Categorised BMI (body mass index) as underweight, normal weight, overweight or obese. • Nominal: Non ordinal. Examples: Profession; Marital status, and binary or dichotomous variables such as gender (in some cases). Jose Barrera (ISGlobal & UAB) Statistics in Health Sciences, 2023/2024 3/5 Classification of variables according to the role in the study Response or outcome Is the variable motivating the study. Usually, it measures certain characteristic related to health, whose description or modelling is the main aim of the study. Some studies analyze several outcomes at a time. Exposure AKA regressor or explanatory variable, depending on the context. It is the variable that could be hypothetically related to the outcome, so its variability could (partially) explain or predict the variability of the outcome. Some studies analyze several exposure at a time, even a high number of them, as is the case of the exposomea paradigm. a  Explore “exposome” concept. See, for instance, the HELIX study at https://www.projecthelix.eu/index.php/en. Confounder It is a variable that could be associated with the exposure and it also could have an effect on the outcome. A confounder is not of interest itself but it should be include in the study in order to control for confounding resulting in adjusted effects. Usually, several confounders are considered. Typical confounders are sex and age. The concept of confusion will be introduced later in this course. Jose Barrera (ISGlobal & UAB) Statistics in Health Sciences, 2023/2024 4/5 Types of explanatory variables regarding its relationship with the outcome Covariate It is any potential explanatory variable included in the study. In some fields of biostatistics, this name is applied exclusively to continuous explanatory variables. Factor It is any potential qualitative explanatory variable included in the study. If a factor has a significant negative (positive) association with the outcome, it is known as a potential risk (protection) factor. However, in epidemiology, this terminology is usually also applied to continuous explanatory variables. Jose Barrera (ISGlobal & UAB) Statistics in Health Sciences, 2023/2024 5/5