Researchers count on the law of probabilities in studies like this. Most of the people keep a diary that reflects most of their costs most of the time. There's probably someone else in the study with high costs who stopped spending money when the study started. That person would say the study doesn't reflect his or her spending at all.

Asking people to keep track for longer periods of time (up to one-year) would probably be more precise. However studies of cost diaries actually show a record for three months is a good sample of what really happens over a full year.