Multiple comparison tests are one of statistical tests that often used in amount of psychological studies. But there are too many kinds of multiple comparison tests to make proper use of the tests. In this article, we redefined multiple comparison tests for interval scale data as "tightening reduplicative t-tests". There are three lights for tightening t-tests; 1) level of significance, 2) distribution, and 3) statistic. Based on the each view many kinds of multiple comparison tests are developed. By setting information on each view for tightening t-tests before in this article, it will become possible to make proper use of multiple comparison tests.