September 15, 2014
d linear regression data set
September 12, 2014
How to Run Anova Test and Tukey’s honest significant difference test (instead of Fisher’s) using R Stat
> x <- read.table(“h:\\rdata\\c.txt”,sep=” “,header=T)
Note: The “h:\\…” is the location of your file on your computer. For Windows computers, locate your file and right click on it, select Properties and you will find the location information. For each backslash use two backslashes.
On a Mac computer, locate your file and press apple key – i to get the location. For macs it will be something like /users/user/c.txt. Therefore for the location use // for the forward slash like this “//users//user//c.txt”.
> x
MethodA MethodB MethodC
1 32.2 35.0 34.5
2 37.4 31.2 38.9
3 35.9 29.3 32.5
4 28.1 28.4 33.3
5 41.0 34.1 45.8
6 44.5 38.1 48.9
7 35.6 32.6 41.1
8 31.2 31.5 37.6
9 34.8 27.7 45.7
> Material = c(x[[1]])
> Material
[1] 32.2 37.4 35.9 28.1 41.0 44.5 35.6 31.2 34.8
> Material = c(x[[1]],x[[2]],x[[3]])
> Material
[1] 32.2 37.4 35.9 28.1 41.0 44.5 35.6 31.2 34.8 35.0 31.2 29.3 28.4 34.1 38.1
[16] 32.6 31.5 27.7 34.5 38.9 32.5 33.3 45.8 48.9 41.1 37.6 45.7
> Method = c(rep(“MethodA”,9),rep(“MethodB”,9),rep(“MethodC”,9))
> mdata = data.frame(Method,Material)
> mdata
Method Material
1 MethodA 32.2
2 MethodA 37.4
3 MethodA 35.9
4 MethodA 28.1
5 MethodA 41.0
6 MethodA 44.5
7 MethodA 35.6
8 MethodA 31.2
9 MethodA 34.8
10 MethodB 35.0
11 MethodB 31.2
12 MethodB 29.3
13 MethodB 28.4
14 MethodB 34.1
15 MethodB 38.1
16 MethodB 32.6
17 MethodB 31.5
18 MethodB 27.7
19 MethodC 34.5
20 MethodC 38.9
21 MethodC 32.5
22 MethodC 33.3
23 MethodC 45.8
24 MethodC 48.9
25 MethodC 41.1
26 MethodC 37.6
27 MethodC 45.7
This rearranges the data set so you should have 2 columns and 27 rows of data as seen. Each row is assigned a Method. For now we will call the data “Material” and this is the independent variable. The dependent variable is Method.
> anovareport = aov(Material~Method,data=mdata)
Please note! The ~ between Materal and Method is the tilda character!
> summary(anovareport)
This will print out the results from the Anova test.
> TukeyHSD(aov(lm(Material~Method),data=mdata),conf.level=0.95)
Please note! The ~ between Materal and Method is the tilda character!
This will print out the results from Tukey. This will show the different in means between the Methods, the 95% confidence interval and the P probability the mean difference is within the confidence interval.
> summary(x)
A summary of the x table.
September 11, 2014