Now we've looked at categorical data types. What about numerical data types? I want to refer you to this article again, where they looked at the combination of Ketamine and Propofol in endotracheal intubation. And if you look again at Table 1, you will see the variable crystalloid use. We see 3,180 ml, 1000 ml. We see all the values there. Now, these are numbers, pure and simple. We can really do arithmetic with them. I can add up all of those volumes. I can compare the different patients to each other. I can tell you that patient one received 3.18 times more than patient number two, because there is a set difference between those two numbers. The difference between 1,000 and 2,000 ml is exactly the same difference between 6,000 ml and 7,000 milliliters. Now there are two types of numerical data. One is called interval, and the other one's called ratio. Now we really want to make that little distinction, and the most famous example always is temperature. In the United States, you guys used to use degrees Fahrenheit. Many other places we use degrees Celsius. And both of those scales do have a zero. There is zero degrees C and there is zero degrees Fahrenheit, but that's not a true zero. If it is 50 degrees Fahrenheit outside or it's a 100 degrees, I can not say that's twice as warm because that is not a true zero. If I wanted to include a scale with a true zero, that would be the Kelvin scale. So temperature really stands apart from the others. Normally though, the data that we deal with, as far as numerical data is concerned, is of a ratio type that has a true zero. Age, for instance, has a true zero. White cell count has a true zero. So now let's get back to that column we had right in the beginning. We had the variables pain, age, gender, and temperature. And we saw all the data values that now we can put names to those data values. Remember the pain scale, those were not true numbers. There was some order to them. So those data values we can call ordinal categorical. If we looked at the age, that would be a ratio type numerical, because you can put them in order, they're true numbers, and they have a true zero. Gender, female, female, male, female, those would be nominal categorical. Remember, we can't put them in any order and we can't do any arithmetic with them. And lastly the temperature. The one that always catches us out, remember, that is interval numerical data type. Now, why do we make this distinction? Merely for one reason, one reason only, different statistical tests go with different data types. As simple as that.