I Count and Categorical Variables...

  • I
  • Thread starter Thread starter fog37
  • Start date Start date
  • Tags Tags
    Count Variables
AI Thread Summary
Count is considered a discrete non-negative numerical variable, representing how many times a specific level of a categorical variable occurs. It can be used on one axis of a scatterplot, although scatterplots typically display two numerical variables. The percentage of a population that identifies as religious can be viewed as a numeric variable derived from a count, despite being relative. Continuous variables like cost and profit can be treated as discrete when considering their smallest increments, such as cents, but binning them into intervals for a histogram creates a new categorical variable without changing the original's nature. Overall, while counts have numerical properties, they also serve as summary statistics for categorical data.
fog37
Messages
1,566
Reaction score
108
TL;DR Summary
Count and Categorical Variables...
Hello,
In the context of categorical variables, a frequency table which gives us the count (aka frequency) for each level of the categorical variable. Count is a number telling us how many times a specific level occurs. A bar-chart handles a single categorical variable (nominal or ordinal) with its levels indicated on the x-axis and count (frequency), or relative frequency, on the y-axis.

My question: is count a discrete non-negative numerical variable? If not, what is it? I don't think it is a numerical variable...
Can count ever be represented as one of the axes of a scatterplot? A scatterplot is designed to accommodate two numerical variables (both discrete, both continuous, one discrete and one continuous) on its two axes.

Here another example that confuses me: a dataset where each row represents a different country and there is a variable that reports the percentage of the country population who is religious (which is the count of people who responded YES to being religious divided by the total country population). That column contains % values and seems to represent a numeric variable even if it represents the relative count of a categorical variable. Is that correct? See attached table:

1676427604467.png


Also, cost and profit are generally considered continuous variables but I believe they are discrete numeric variables since money is a multiple of the cent, the smallest increment. When we build a histogram, we bin the continuous variable values into intervals. Does that turn the continuous variable into an ordinal categorical variable since the data now belongs in a finite number of groups?

Thank you!
 
Physics news on Phys.org
fog37 said:
My question: is count a discrete non-negative numerical variable?
It certainly could be.

fog37 said:
If not, what is it?
It is a summary statistic.

fog37 said:
Can count ever be represented as one of the axes of a scatterplot?
Sure. Just like you could put medians or standard deviations on one of the axes of a scatter plot you could also put counts on an axis of a scatter plot.

fog37 said:
Does that turn the continuous variable into an ordinal categorical variable since the data now belongs in a finite number of groups?
It could if you wanted to.
 
fog37 said:
My question: is count a discrete non-negative numerical variable? If not, what is it? I don't think it is a numerical variable...
It is a natural number (where 0 is included). They are ordered and have arithmetic properties.
fog37 said:
Can count ever be represented as one of the axes of a scatterplot? A scatterplot is designed to accommodate two numerical variables (both discrete, both continuous, one discrete and one continuous) on its two axes.
I wouldn't call it a "scatter plot", but you certainly can make charts where the categories are along one axis, in order of the count. It happens all the time when you are interested in what category occurs the most. From https://inferentialthinking.com/chapters/07/1/Visualizing_Categorical_Distributions.html:
Visualizing_Categorical_Distributions_21_0.png

fog37 said:
Here another example that confuses me: a dataset where each row represents a different country and there is a variable that reports the percentage of the country population who is religious (which is the count of people who responded YES to being religious divided by the total country population). That column contains % values and seems to represent a numeric variable even if it represents the relative count of a categorical variable. Is that correct? See attached table:

View attachment 322281

Also, cost and profit are generally considered continuous variables but I believe they are discrete numeric variables since money is a multiple of the cent, the smallest increment. When we build a histogram, we bin the continuous variable values into intervals. Does that turn the continuous variable into an ordinal categorical variable since the data now belongs in a finite number of groups?
No. The counting numbers have too many arithmetic properties to just be considered "categorical". They are ordered. You can add them. You can subtract them, although it might give a negative result. You can divide them if you allow rational numbers. The natural numbers are a subset of the integers, the rational numbers, and the real numbers and should not be considered only "categorical".
 
Last edited:
  • Like
Likes WWGD and fog37
fog37 said:
When we build a histogram, we bin the continuous variable values into intervals. Does that turn the continuous variable into an ordinal categorical variable since the data now belongs in a finite number of groups?
It doesn't "turn it into" a categorical variable, but it creates a new categorical variable. If the original continuous variable was 'income' we could call the new categorical variable 'income range'.
I occasionally work on health system patient data which, amongst hundreds of variables, has age (of patient at time of recorded health system interaction) - a floating point variable, and 'age-range' a categorical variable that classifies patients into 5-year age ranges. Although the database provides both variable, we could just read in 'age' and derive 'age range' from 'age'.
 
  • Like
Likes Dale, WWGD and fog37
I was reading documentation about the soundness and completeness of logic formal systems. Consider the following $$\vdash_S \phi$$ where ##S## is the proof-system making part the formal system and ##\phi## is a wff (well formed formula) of the formal language. Note the blank on left of the turnstile symbol ##\vdash_S##, as far as I can tell it actually represents the empty set. So what does it mean ? I guess it actually means ##\phi## is a theorem of the formal system, i.e. there is a...

Similar threads

Replies
5
Views
4K
Replies
1
Views
3K
Replies
13
Views
3K
Replies
22
Views
4K
Replies
4
Views
3K
Back
Top