For A two sample KS test data should be independent. Which of the following has to be true to use the KS test reliably(i.e. what does independent mean)

  1. Data should be Independent within each sample and independent between the two samples. E.g. take 200 people and weigh them and split into two samples of 100 and measure KS distance.
    1. Data should be independent within each sample and correlation/dependence between the samples is allowable. E.g. take 100 people and weigh them on the same day using two different weighing machines. Then measure KS distance between the two samples of 100 numbers.
    2. Data has to be independent between the two samples and data within samples can be dependent. E.g. take two people and weigh them each month for 100 consecutive months to create two samples of 100 numbers and then measure KS distance between the two samples of 100 numbers
  • $\begingroup$ How does this question arise? Is this for some subject? $\endgroup$ – Glen_b Sep 4 '17 at 23:19
  • $\begingroup$ Example potential use. Am weighing people with two different measuring scales and then have two sets of observations and am thinking about using the KS distance as a way of measuring if the two scales are aligned (one scale is known to be accurate). $\endgroup$ – James65 Jan 27 '18 at 10:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.