Quantifying the similarity of two ordered data sets

Asked Mar 16 '16 at 16:24

Active Mar 08 '21 at 01:42

Viewed 43 times

Hi I have two data sets.

Atrue = [3.5, 8.2, 4.1, 5.5, 3.1, 4.9] and Aexp = [3.2, 8.5, 3.7, 5.1, 3.3, 4.8]

Where Atrue is the theoretical data and Aexp is the experimental data.

Each index element of Atrue corresponds to the index element of Aexp. So 3.5 would be compared to 3.2 and 8.2 would be compared to 8.5...etc.

I want to quantify Aexp's closeness to Atrue - in particular I want to test to what statistical confidence they are similar.

What test can I use?

Note: The lists are actually much longer than the above but I just wanted to illustrate my point with an example.

edited Mar 08 '21 at 01:42

asked Mar 16 '16 at 16:24

piccolo

Usually, statistical tests are for *differences*. What exactly do you mean by "closeness" or "similar" here? You could test for equivalence (cf, [here](http://stats.stackexchange.com/a/85914/7290)) or agreement (cf, [here](http://stats.stackexchange.com/a/199714/7290)). – gung - Reinstate Monica Mar 16 '16 at 17:02
Statisticians often use correlation, or some other [tag:association-measure]. Maybe show us a plot? – kjetil b halvorsen May 05 '21 at 17:15

0 Answers0