Biostatistics and experimental design
Amplification degree q test: higher precision > new multiple range: comparing the average error various > the LSD
Wherein, the LSD does not change with M change change, but the SSR and q-test will vary M changes vary.
The first step represents the core idea of the analysis of variance
Step F test and t test empathy
The third step is just a set of factors to know if there are differences, there are differences not know what level is required for multiple comparisons.
Playing Asterisks indicate significant
Two-factor analysis of variance:
The main effect is to test each factor independent role.
Interaction effect is not independent of each test factor, i.e., factor A and factor B to form a super element.
To ensure consistent conditions for all samples, namely SE same, although this requirement can not be achieved in practice, but solely from the point of view of the principal contradiction can be ignored.
When looking at a single factor, other factors differences ignored.
Environment need to extract the principal contradiction, the relationship between environmental factors diapause length of a function (regression analysis) and, diapause length or different environmental factors of whether there are differences (ANOVA).
After obtaining the interaction was not significant, can try to use two non-repeating factors, see SE whether it is normal. After this, the analysis result of two distinct elements, if there are not significant factors Thereafter, the factors can be classified as a secondary factor, thus establishing one-way ANOVA. For the first time the analysis is necessary to return, but the second is not necessary, because the non-repeat of one-way ANOVA and two-factor analysis of variance principle is the same, all for treatment , analysis of variance for repeated random factors .