Part (f-1)
1 point possible (graded)
Which one of the following is a good test statistic for the data:
"Sanity check": test statistics should be a quantity that helps you measure how likely the observed data is given your null hypothesis.
( Lambda(X)=-2 ln left(frac{max_{lambda} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda, ldots, lambda
ight)}{max _{lambda_{0}, ldots, lambda_{99}} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda_{0}, ldots, lambda_{99}
ight)}
ight) )
( Lambda(X)=-2 ln left(frac{max _{lambda} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda, ldots, lambda
ight)}{max _{lambda_{0}, ldots, lambda_{99}} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda_{0}, ldots, lambda_{99}
ight)}
ight) )
( Lambda(X)=-2 ln left(frac{max _{lambda} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda, ldots, lambda
ight)}{max _{lambda} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda, ldots, lambda
ight)}
ight) )
( Lambda(X)=-2 ln left(frac{max _{lambda_{0}, ldots, lambda_{99}} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda_{0}, ldots, lambda_{99}
ight)}{max _{lambda} fleft(G_{0}, G_{1}, ldots, G_{99} mid lambda, ldots, lambda
ight)}
ight) )
