3. The dataset data5. csv contains the following columns 0 y: Continuous response variable measured at a xed follow-up time for which larger values correspond to better outcomes. This variable is missing for dead subjects. 0 dead: Indicator of death before the response y could be measured (lzdead, 0=alive). o z: Randomly assigned treatment (0/1). (a) Assess the effect of treatment on mortality. (b) Assess the difference between treatment groups for the non-missing responses 37 (survivors only). Does this represent a causal effect of treatment? (c) By assigning a low score to death (lower than any observed y), assess the difference between treatment groups for the composite outcome of death and y. Does this represent a causal effect of treatment