Paired t-test is a common t-test method, which refers to the analysis of samples obtained in the same whole under different conditions to evaluate the impact of conditions on them, including different storage environments, different measurement systems, etc., in addition, t-tests also include single-sample t-tests, multi-sample t-tests, etc.
<h1 class="pgc-h-arrow-right" data-track="2" > the significance of pairing the t-test</h1>
In the case of two samples correlated, the following can be done using a paired t-test:
Determine whether the difference between two paired samples, mean or target, is zero;
Calculates a range of differences that may contain the population mean.
Pairing design: Some characteristics of the test object are paired according to the principle of similarity, the purpose is to eliminate the influence of confounding factors, while observing that the objects are consistent in addition to the processing factors and research factors, and the two individuals in each pair are randomly processed.
Purpose: To determine the impact of different treatment methods.
<h1 class="pgc-h-arrow-right" data-track="8" > example of a paired t-test scenario</h1>
Classify the two kinds of objects of the same kind in different ways, as follows:
Compare and observe the same subject before and after;
For the same tester, compare the difference between two different measurement methods or two different treatment methods.
<h1 class="pgc-h-arrow-right" data-track="12" > parent word: t test</h1>
<h1 class="pgc-h-arrow-right" data-track="13" > subverse words: single-sample t-test, two-sample t-test</h1>