PARTICIPANT OBSERVATION IN A MULTIPLE·METHODS STUDY OF A RETIREMENT COMMUNITY: A RESEARCH NARRATIVE

In social science research, choice of methodology is con strained by the real or perceived biases of funding agencies' conception of what will persuade audiences who may or may not use the results of the study for good or evil, and the convic tions of investigators as to what constitutes valid data. For most audiences in sociology and social gerontology, and I suspect for most funding agencies, precisely defined and measured variables and numerically tested hypothesis constitute the language of persuasion. As Herbert Blumer (1969:37) argues: