We extend Che and Kartik (2009) to the multidimensional state space. The decision maker consults an expert before making multiple independent decisions, then differences of opinion will create more incentive for the expert to acquire information costly but burden their communication. Under the assumption of correlated information acquisition and disclosure, in the n-dimensional setting, we prove the existence of the perfect Bayesian equilibrium and find when all the experts have the same opinion with the decision maker, consulting one expert for multiple decisions is more beneficial than consulting several independent experts for different decisions. In the two-dimensional case, we show that choosing an expert with difference of opinion in at least one dimension is better than that with the same opinion in both dimensions.