ordinal, interval, or ratio data?

Question

In a study, researchers tracked the number of errors made as participants completed 50 trials of the Stroop color task. All participants made 0, 1, or 2 errors during the 50 trials. Are the results of the Stroop color task number of errors considered ordinal, interval, or ratio data? 

Answer 1

This is an interesting question.

 

It's not ordinal, because in ordinal data, the ORDER of the values matters (1 is more than 0, 2 is more than 1), but not the DIFFERENCE between them.  So for example, you might have a scale of tastes -- rank the sweetness of a food on a scale from 1 to 5.  5 is sweeter than 4, but the scale doesn't say anything about how much sweeter a 5 is than a 4 or a t is than a 3.  But in this data, 2 errors is exactly 1 more than 1 error, and 1 error is exactly 1 more than 0.

 

So is it interval or ratio?  In interval data, there is a clear meaning to the difference between two values.  That's definitely true here -- as I said above, 2 errors is exactly 1 more than 1 error, and 1 error is exactly 1 more than 0. 

 

But there's the additional feature that there is a clear meaning to the 0 value.  It means that no errors were committed.  And that'exactly what distinguishes ratio from interval.

 

So the answer is ratio.