I need to compare three different data series. The subject of the data is the same, but the collection method is different.

My initial approach was to only compare two of the series at a time, using Kendall's correlation coefficient. I chose this because it seemed like the best method to compare the "likeness" of the data at any given point.

What would be the best way to compare these sets of data? Is correlation the best way?

Any help would be appreciated.

EDIT: Please allow em to correct myself. I did not mean to compare "likeness" of the data at any given point, rather to compare the entirety of the data. This is why I selected Kendall's coeff over Pearson's.

My initial approach was to only compare two of the series at a time, using Kendall's correlation coefficient. I chose this because it seemed like the best method to compare the "likeness" of the data at any given point.

What would be the best way to compare these sets of data? Is correlation the best way?

Any help would be appreciated.

EDIT: Please allow em to correct myself. I did not mean to compare "likeness" of the data at any given point, rather to compare the entirety of the data. This is why I selected Kendall's coeff over Pearson's.

Last edited: