Paired Difference Test

From GM-RKB
(Redirected from paired difference test)
Jump to navigation Jump to search

A Paired Difference Test is a location test that can be used when comparing two sets of measurements to assess whether their expected values differ.



References

2016

  • (Wikipedia, 2016) ⇒ https://en.wikipedia.org/wiki/paired_difference_test Retrieved:2016-9-14.
    • In statistics, a paired difference test is a type of location test that is used when comparing two sets of measurements to assess whether their population means differ. A paired difference test uses additional information about the sample that is not present in an ordinary unpaired testing situation, either to increase the statistical power, or to reduce the effects of confounders.

      Specific methods for carrying out paired difference tests are, for normally distributed difference t-test (where the population standard deviation of difference is not known) and the paired Z-test (where the population standard deviation of the difference is known), and for differences that may not be normally distributed the Wilcoxon signed-rank test. In addition to tests that deal with non-normality, there is also a test that is robust to the common violation of homogeneity of variance across samples (an underlying assumption of these tests): this is Welch's t-test, which makes use of unpooled variance and results in unusual degrees of freedom (e.g. df' = 4.088 rather than df = 4).

      The most familiar example of a paired difference test occurs when subjects are measured before and after a treatment. Such a "repeated measures" test compares these measurements within subjects, rather than across subjects, and will generally have greater power than an unpaired test.