I am doing R2 correlation. I generally get low values, which is expected. However, when the spread of the values is big, the R2 value goes way up. For example, for 6-7 pairs with values for one variable between 20 and 60, the R2 is 0.11 or 0.14. When I add a data pair with a value ~4000, the R2 becomes 1.00. I don't believe that the correlaton I suddenly good. Is an R2 analysis not valiud when the spread of values is wide?