1. Since only the mean process variable from the first layer and the linear correlation analysis of variance test, and thus the same single factor analysis of variance of two factors, or the mean value of two or more factors of the process, except that the most descriptive statistics descriptive statistics is a statistical factor for each crossover group.
Conditions 2.t test: ① the sample from a normal population or approximately normal population, the population variance equal ② two samples, i.e., with homogeneity of variance (In practice, each such deviation from the above-described conditions, the results also It will not have much impact). ③ The sample should be independent of each other.
3. One-sample T-test, ie this test if there is a significant difference between the overall mean of a variable and a specified value. If it is a large-sample test sample, called U test textbook on statistics, the use of statistics U normal distribution as the test statistic. If the sample is small and normal distribution is used to obey t statistic distribution of single-sample T-test; otherwise, the nonparametric test. T test the robustness of the better, if less severe sample distribution deviates from the normal distribution, also can use T-test. 1