Strategyproof Mean Estimation from Multiple-Choice Questions