3

I know for sure that BMI(Body Mass Index) is a quantitative variable as it is a continuous variable. But is that BMI Category derived from the BMI a qualitative variable or a quantitative variable? (Underweight, Normal Weight, Overweight. Thanks

Steven Steven
  • 39
  • 1
  • 1
    It is considered to be in this hybrid category called ordinal. Depending on what you’re doing, binning into these categories may discard useful information. What are you trying to do with BMI? – Dave Nov 10 '20 at 11:20
  • I am trying to make a contingency table with other categories, but I need to make sure that BMI Category is a quantitative variable. – Steven Steven Nov 10 '20 at 11:24
  • Why do you want to make such a contingency table ? What is your research question ? – Robert Long Nov 10 '20 at 11:39
  • I want to make a two-way contingency table with BMI Category and Favorite Ice Cream flavor to know if that a certain ice cream flavor can cause overweight. – Steven Steven Nov 10 '20 at 11:58
  • 1
    First, you can't establish causation with any kind of observational study. Second, why don't you use a regression model, with BMI as the outcome and ice cream flavour as the independent variable ? Categorising your data will lead to massive data loss. – Robert Long Nov 10 '20 at 15:52

1 Answers1

7

Usually we think of this type of data as a special form of categorical data called "ordinal", that is, ordered-categorical. This is because there is a natural ordering in the data: Underweight < Normal < Overweight.

While it is sometimes useful to create categories such as these, there is a great loss of information by doing so.

Edit: Based on comments in the question, it would be a good idea to consider a regression based model, where you do not categorise BMI but rather use it as the outcome/response. Since you seem to have a just one independent variable, which is categorical, this would be a one-way ANOVA.

Robert Long
  • 53,316
  • 10
  • 84
  • 148
  • 1
    This ANOVA ends up being just a vanilla t-test! – Dave Nov 10 '20 at 19:09
  • @Dave but we don't know how many categories the IV has. Either way, it's better than categorizing the DV imho – Robert Long Nov 10 '20 at 20:10
  • I read the comment as being between two ice cream flavors. Looking at it again, I am not sure how. Yes, if there are 3+ flavors, then it isn't a t-test. – Dave Nov 10 '20 at 20:41
  • 1
    @Dave I agree it's a bit unclear but I think it's good as pedagogical device to think about regression. – Robert Long Nov 10 '20 at 21:18