Why are variables that have a `0` value considered to be on a Ratio Scale since the ratio might not be defined for those values?

Mission Step - https://app.dataquest.io/m/284/variables-in-statistics/7/the-difference-between-ratio-and-interval-scales

There are some variables which have 0 as a value. For those, we can’t use ratio as a means to measure the difference between two values (0/0 situation or maybe even x/0 situation)

So, why are they considered to be on the Ratio Scale? The ratio would either be not defined or it would be infinite and both don’t make much sense in this context to me.

Or is it just meant to be an either-or situation for Ratio Scales - you can measure the difference in terms of ratios or in terms of distance?


hi @the_doctor

Interesting question. I would like to know the answer myself. I guess the either-or situation is a general & applicable principle for ratio scales.

We will most likely compare, some quantity with some other quantity and/or no (0) quantity, but we can’t compare no quantity with no quantity.
In practical terms, we can compare ratio of medals an athlete won as compared to another athlete, when both have participated in multiple events and have won in some events. But there is no way for us to compare medal ratio or difference in count, when both these athletes are participating for the first time (their first ever event), but the tally will still show No. of medals as 0. And we can’t also conclude, that they have the same amount of medals - when they have none.

Please do share your thoughts.

1 Like

Hi Rucha

I have the same question, in Variable Statistics part 7 - the answer includes 3PT%, FT%, and FG% as ratio. However, a 0 in those variables can mean that the athlete shot 0 which is not an absence of quantity. i.e. if they make 0 out of 5 FT their FT% would be 0%.