Before empirically based theories of loneliness can be formulated, it may be necessary to probe people's feelings, explanations, and reactions in a relatively unstructured way. Toward this end, three multiple response items relating to these issues were included in an 84-item questionnaire, published in the Spring of 1978 in several northeastern newspapers. Each of the three questions, dealing with feelings, reasons, and responses, was followed by a list of answer alternatives derived from open-ended questions in a pilot study. Approximately 24,000 respondents filled out the questionnaire. Factor analyses performed on each of the three questions, separately for each sample, revealed highly stable factor structures. Of the three factor analyses, feelings, reasons, and reactions to loneliness, the one for feelings produced the most highly intercorrelated oblique factors. (Author)