Let X1, X2, X3 be i.i.d. Bernoulli(p) where 0 < p < 1 is the unknown parameter. Evaluate Fisher information IX e (p), I X (p) and IT (p), where X e = (X1, X2, X3), X is the sample mean and T = X1 X2. Can X be a sufficient statistic for p? Can T be a sufficient statistic for p