How to apply Data Reference Errors testing to the following application?

A consulting company has developed an application that converts temperature from kelvin to Fahrenheit and from Fahrenheit to kelvin.
Formulas: F= (K-273)*9/5 + 32
K= 5/9 (F-32) + 273
Min/Max values: kelvin min value 0 max value +450
Fahrenheit min value -460 max value +350
I would like to apply Data Reference Errors testing to this application and specify values that will be used for the test but I am not very familiar with white box testing techniques.
Could somebody please detail the correct approach to apply Data Reference Errors testing to this example?