# What kinda of Variable is this?

1. Feb 16, 2012

### nitsuj

The following values are tonnes of a specific animal feed for a specific consumer.

Jan - 13.97
Feb - 12
Mar - 11
Apr - 5.87
May - 10
Jun - 13.95
Jul - 15.96

Are these variables best described as "quantitative ratio variables" and why?

If not what would be the most accurate class for these variables and why?

I pretty green when it comes to stats , hopefully this isn't too easy to be worth a reply.

2. Feb 16, 2012

### Stephen Tashi

My personal opinion is that such data isn't obviously a "ratio" of any kind, but you should list all the variable types we have to pick from.

Such classifications of variable types are not standard terminology across all statistics texts. Perhaps you are studying statistics in a specialized field.

3. Feb 16, 2012

### SW VandeCarr

Mathematically, there are discrete and continuous variables. In applications, variables are sometimes broken down into ordered categorical variables, unordered categorical variables and indicator variables. Since virtually all data are discrete, the question becomes: "What kind of distribution do the data represent?" In your case, I would say that the observations come from a continuous distribution of measure, limited by the precision of measurement and the convenience of rounding. By the this I mean that, in principle, uncountably many possible values lie between the values you actually observed.

EDIT: Your particular data set has some problems in that you have numbers like 10 and others like 13.97. At the very least a value like "10" should be written as 10.00. However, you really can't properly analyze data with data points having such mixed levels of precision.

Last edited: Feb 16, 2012
4. Feb 16, 2012

### nitsuj

Thanks for the replies guys!

SW Vandecarr,