CODEWITHSUNDEEP
statisticsprobabilityp-value
HD Ly
HD Ly

Reputation: 45

How to calculate significant different between two frequency distribution

I had a dataset with 100 rows and 2 columns. Each row stands for the bin of histogram and two columns are two measurements displayed in fraction. If we plot two columns in two curves, we will easily to see that two curves are different but I want to find the significant difference statistically for example P value etc. Anyone have idea how to do that ?

Upvotes: 2

Views: 4389

Answers (1)

michele de meo
michele de meo

Reputation: 61

I think you should specify if your data are repeated measurements. If so, each row represents the same sample unit, while columns indicate the results observed before (col 1) and after (col 2) the treatment. In this case, you should consider the Wilcoxon Signed-Rank Test or, more in general, a Paired difference test. Here you can find some useful information in R.

If your data are not repeated measurements (as I imagine), you have to test if the two samples come from the same distribution, without specifying this distribution. The simplest test for "binned" distribution is the chi-square two sample test.

Upvotes: 1

Related Questions