R错误df中的1 Way Anova测试

Question

I want to run an anova test in R to find out if there's a difference in the number of weeks of each brand. My code is as follows:

weeks = c(100,96,92,96,92,76,80,75,84,82,108,100,96,98,100)
brand = c('a','a','a','a','a','b','b','b','b','b','c','c','c','c','c')
battery.exp = data.frame(brand, weeks)
test.1 = aov(battery.exp$weeks~battery.exp$brand)
summary(test.1)

I'm getting incorrect results in the summary. The df of the brand should be 2, but my output says 3. Does anyone know why I would get incorrect results in my anova test? I've also tried to replace the second to last line with the following:

test.1 = aov(battery.exp$weeks~factor(battery.exp$brand))

but that also didn't solve the problem. Any suggestions on how to fix?

Answer 1

The default is to use a model with an intercept which means you need to use a reference level for your categorical variable so you only have two free parameters. If you don't want to use an intercept, you can explicitly request an intercept free model with

summary(aov(weeks~brand-1, battery.exp))

(note the -1 in the formula)