Trial arm Number Deaths randomised Placebo 614 65 Aspirin 605 46 A standard model specification is that the number of deaths in each arm may be regarded as an independent binomially distributed random variable: Y, ~ Bin(nj, T;), where Y; is the number of deaths among nj individuals randomised to group j, with j = 1 denoting the control (placebo) group and j = 2 the active (aspirin) group. The parameters specified in this model (m, and 12) are not of direct interest themselves, but we might be interested in any of the following "effect measures": Risk difference: 0 = 12 - 11 Risk ratio: Y = 12/ T1 Odds ratio: p - 2/(1-T2) #1/(1-1 1)Now consider the parameters 6, the risk difference, and ,8 = log (y), the log of the risk ratio. (i) What are the maximum likelihood estimates of 6 and [3? Again, this doesn't require derivation from the likelihood function; you can use the invariance property of MLEs. (ii) Derive formulas for the standard errors of gm and 3m by using the formula for the variance of the sum of two independent random variables