Types of data in the healthcare domain and doubt where I require advice from PadhAI team

Curriculum Help Data Science
#1

Examples of data types in healthcare domain.
Qualitative:-
nominal:- patient sex ; state, comorbidity
ordinal: patient age group (<18; 18-40; 41-60; >60) ; severity of disease

Quantitative
continuous :- BMI, Lipid profile, height, weight
discrete :-date of first diagnosis, incidence, prevalence.

Question - will year of birth be discrete or ordinal or depends upon the usage ? Similarly can variables type change depend upon usage - e…g. year of birth may be ordinal if you use it for bucketting but year may also be discrete quantitative if you are using it for trend analyses ?

#2

As per the definition, Ordinal Data is where the variables have natural ordered categories, and distance between categories is not known.
But in case of age bands, there is a clear age gap known while setting a particular age band. Hence it cannot be termed as Ordinal Data.

#3

so it should be discrete quantitative? In our clinical trial studies we used it as a categorical data variable for demographics calculations hence asking.Also ref. article below ? Thanks so much for your help, and apologies for the follow-up questions

1 Like
#4

Believe even though the distance is known, still the distance is not consistent. Band > 60 includes far more age values (40 odd if we consider 100 the max age) than < 18 (Just 18 values) . So considering Age as Ordinal for the specific case here seems valid.

1 Like
#5

As far as I understood the the Quantitative attribute is a measurement and a Qualitative attribute is a label. Therefore the year of birth/age is a measurement while the age group is a label. Now if the label is following some natural order than it is ordinal otherwise nominal(Refer to the instructor’s explanation regarding the example of low medium high, which is an ordinal data) Therefore your year of birth will be a discrete attribute.