In unbalanced data, Two levels of SED Different, therefore LSD Also different, So there is no unified LSD.Therefore, If you want to use asreml Multiple comparisons, Need to compute sed,asreml Can calculate two levels of SED, So you can manually calculate two levels of LSD, Then you can make multiple comparisons between two levels.
tvalue The calculation of 0.05 And the calculation of degrees of freedom. Therefore, The key to multiple comparisons is SED And tvalue The calculation of. R Language Computing LSD by : # $statistics Here are the factors block The values of are : # Least significant differences of means (5% level) # Standard errors of differences of means
> mod1 = aov(yield ~ N + block, data=npk) Modeling : > # 3, modeling :yield ~ N + block Load data, View the data : > # aov Multiple comparisons of Use npk data, Modeling, Yes block Multiple comparisons. variance analysis aov Multiple comparisons of If the difference is significant, It means that at least one pair of varieties has reached significant difference, The question is which pair, Or which pairs ?Ģ.variance analysis, See if there is a significant difference between varieties.Yes 10 A variety of, Production data, Want to see if there is a significant difference between the varieties, Analysis methods : The definition of Baidu Encyclopedia above, Generally speaking : Multiple comparisons can overcome these shortcomings. Compare m A mean, It needs to be done alone (m/2)=m(m-1)/2 Time t test, It's not only a lot of work, And the error is also large. t The test can only show whether the difference between the two means is significant. Through the analysis of variance, we can explain whether the difference between the population means is significant, That is to say, the mean value is not all equal, But we can't specify which means are significantly different from each other. The multiple comparison method is the comparison method of the mean values of multiple equal variance normal populations.