Hypothesis-generating interviews: These are preliminary interviews conducted with a limited number of cases (perhaps 8-10) in order to identify possible sources of exposure to the infectious agent. The cases should be typical cases drawn from a representative sample of the population (e.g. not all one race, gender, or age group), and the interviews should ideally be conducted by a single investigator. Before conducting interviews, the investigator should review the features of the disease and consider likely sources based on the biology of the disease (e.g. known modes of transmission) and information from past outbreaks. The interview should explore exposures during a specific time period, based on the onset of disease and the know incubation period; this focuses the interview on the period during which exposure was likely to have occurred. The information gathered usually includes basic demographic information (age, gender, residence, occupation), clinical details (date of onset, duration of disease), and many questions regarding possible exposures (contact with other ill persons, food history, water exposures, exposure to animals, travel, occupational exposures, hobbies, etc). The conversation should be open-ended and wide-ranging and not strictly confined to well-established sources of disease or modes of transmission. In difficult investigations with few clues investigators have sometimes gathered several cases together to discuss possible common exposures. In other cases investigators have found it useful to visit the homes of the cases and look through their refrigerators and food shelves for clues.