One hundred children take part in a clinical trial of the minor surgical procedure to alleviate "glue ear". On referral the children are divided into two equal sized groups according to their degree of hearing loss. These two groups are further, randomly, divided into a group that are operated on and a group that are not. Two years later all the children are contacted and given a quality-of-life questionnaire to fill in. This yields a normal distribution of scores. The hypothesis to be tested, suggested by the hospital managers, is that only the high hearing loss group will benefit significantly from the operation.