slipforce=read.table("slipforce.txt",header=T)
attach(slipforce)
#plot main effects
par(mfrow=c(2,4))
for(i in 1:8) boxplot(Force~slipforce[,i],main=paste("X",i))
par(mfrow=c(1,1))
#find effects - need to include the right number of terms (2^4 - 1)
#     note the interactions are highly confounded, see Table B.1-8, p.765
2*coef(lm(Force~X1+X2+X3+X4+X5+X6+X7+X8+X1:X2+X1:X3+X1:X4+X1:X5+X1:X6+X1:X7+
   X1:X8))
#how can we tell which are significant?  
#assume most of the effects are zero and check the normal probability plot
# (without the intercept)
qqnorm(2*coef(lm(Force~X1+X2+X3+X4+X5+X6+X7+X8+X1:X2+X1:X3+X1:X4+X1:X5+X1:X6+
   X1:X7+X1:X8))[2:16])
#this default line might not be as helpful in this case:
qqline(2*coef(lm(Force~X1+X2+X3+X4+X5+X6+X7+X8+X1:X2+X1:X3+X1:X4+X1:X5+X1:X6+
   X1:X7+X1:X8))[2:16])
sort(2*coef(lm(Force~X1+X2+X3+X4+X5+X6+X7+X8+X1:X2+X1:X3+X1:X4+X1:X5+X1:X6+
   X1:X7+X1:X8))[2:16])
#here's a slightly better line
qqnorm(2*coef(lm(Force~X1+X2+X3+X4+X5+X6+X7+X8+X1:X2+X1:X3+X1:X4+X1:X5+X1:X6+
   X1:X7+X1:X8))[2:16])
abline(a=4.8,b=14.3)