3

My background is not in stats but I did do some in school. You're help is really appreciated.

I'm wondering what is the best way to find some correlation between a binary variable and a discrete variable with a range between 1 and 5 (inclusive).

For example, let's say the binary variable is whether an employee was working at a lemonade stand at a specific time.

The discrete variable ranges from 1 to 5 and equates to how a customer rated the lemonade at a specific time.

Data example:

Time | Rating | Bill Worked? | Susy Worked?

10:00 | 5 | 0 | 1

10:35 | 3 | 1 | 0

etc.

I'd like to see if certain employees impact the rating of the lemonade by a significant amount.

Not even sure if correlation is the best way to do this...

Thanks very much in advance!

Edit: The distribution of the "customer rating" appears to be lognormal...

1: 300 2: 70 3: 330 4: 1100 5: 3000

I ran a linear regression but I'm worried that the results will be insignificant.

Johnny876
  • 31
  • 2
  • If you use `R`, then, generally, in order to determine correlation between **heterogeneous** variables, you can use `hetcor()` function from the `polycor` R package: http://cran.r-project.org/web/packages/polycor. The values for variables of interest should be either in matrix format, or in data frame format. – Aleksandr Blekh Jan 02 '15 at 22:45

0 Answers0