Statistics for everyone: ch.6 ANOVA

easier R than SPSS with Rcmdr : Contents

ch.6 ANOVA

The analysis of variance (ANOVA) can be said to be an extension of the t-test. The t-test is used to compare two populations, and ANOVA is used to compare three or more groups at once.

Select Variables. Select a dose whose Group variable has a value of at least 3.

The graph has 3 options, choose one of them.

You also have to choose whether to assume equal variance, which is what we’ve seen in the t-test as well.

There are 4 ways to compare Pairwise (also known as the post-hoc test), choose one of them.

2 pictures appear overlapping, so move one to the side to look at it. The figure on the left shows the mean and standard deviation. The figure on the right shows a 95% confidence interval of the difference value compared by two groups. If this confidence interval contains 0, it is determined that there is no difference between the two groups.

‘summary.anova’ usually shows the mean of the three groups, the standard deviation, and the p value. Since it has a p value less than 0.05, we determine that the mean of the three groups is not the same, and it show the TukeyHSD below. Since the p value of all pairs is less than 0.05, we determine that there is a difference in all groups. We used the value p adj to solve the problem of multiple tests that arise because of repeating the test several times. You can also write ‘adjusted p<0.001’ in your paper.

You first run ‘RcmdrPlugin.EZR’ and then ‘RcmdrPlugin.Kmggplot2’ or run both at the same time to see the ‘Kmggplot2 ‘ menu.

The above two graphs, created using ‘Kmggplot2’, are suitable for cases where the number of points is small or there is no normal distribution, while the lower histogram is matched when doing a normal distribution. This is because the mean or standard deviation gives more important information in the normal distribution.

On the other hand, in ‘Kmggplot2’, if you use the X variable on the S tratum variable, as on the right, i.e. using dose on the S tratum variable and supp on the X variable , you will get a graph shape like the one below.

These graphs are suitable for 2-way ANOVA, which we will look at in the next chapter. For beginners, the concept of 2-way ANOVA can be a bit daunting, so you can skip this part.

easier R than SPSS with Rcmdr : Contents

=================================================