Should I consider this variable continuous or discrete?

Hi everyone. I'm working on a dataset in which there are n = 24 observation on subjects, tested for physical resistance. The goal is to evaluate the burnt calories (first variable) based on the body mass in kilos (second variable) and the work level (third variable).
The column of the variable wl (work level) contains values which go from 13 to 56, all observation are discrete EXCEPT for two observations which have continuous values (34.5 and 34.5). Given so, I choose to consider this variable as continuous since not all of the 24 observations are discrete (I converted all the values like for example 27 to 27.0).
Is this procedure wrong? Would you have considered this variable in some other way?

Also, since the dataset is very small, is an istogram graph useful for a descriptive analysis of the variables or should I just go with a summary() of the three variables?

I give also the data file for better understanding. Thanks for the help.



Ambassador to the humans
Discrete vs continuous typically doesn't matter when it comes to models outside of choosing a theoretical conditional distribution for the dependent variable. Also possibly deciding if it makes sense to treat a numeric predictor as categorical but that's more in number of levels and discrete vs continuous isn't really important. You haven't really said what you're trying to do either so it's not really clear to me what you think the impact would be here.


Active Member
One could multiply all values by ten and then have integer measurements, suggesting this is an issue related to the scale of the unit of measure.

Do you foresee a need for even smaller values? Is there a point at which smaller doesn't make sense, for example exceeding the precision of measuring devices?