Middletown III: Problems of Replication, Longitudinal Measurement, and Triangulation

This paper considers the present state of replication in the social sciences with special reference to a subtype of replication, longitudinal research as it applies to community research. Opportunities and pitfalls in replication are illustrated in the experience of the Middletown III Project, a 50-year follow-up of Robert and Helen Lynd's study of the same community in 1924-1925. Replication is defined as research undertaken with reference to a particular baseline study, and a replication may differ from the baseline study in any of its major features. A typology of replicative studies is offered based on four properties of research: time, location, subjects, and method. Longitudinal research is defined as a subtype of replication, and growing interest in and practice of longitudinal research are noted. Community research, which may involve mUltiple replications within a single local context, is contrasted with the usual longitudinal studies that trace specific individuals over time and through a variety of community contexts. Five guidelines for replications are offered: (a) clearly establish the baseline to be replicated; (b) resist the temptation to expand the replication to include topics absent in the baseline study; (c) where extensions are justified, try to include longitudinal linkages using techniques of retrospective or life-history reconstruction; (d) control the urge to update the datacollection instruments; and (e)pay close attention to the interplay between community-level variables and data on individual attitudes and experiences.