We present a proposal for helping students to cope with statistical word problems related to the classification of different cases of confidence intervals. The proposal promotes an environment where students can explicitly discuss the reasons underlying their classification of cases.