A “Link” System for Assuring Confidentiality of Research Data in Longitudinal Studies1

Behavioral scientists have long recognized the importance of longitudinal research data in studies of human growth and development. A major logistical problem in such studies, however, is that the research subjects must be identified in some manner so that they may be resurveyed periodically. Although most researchers are aware that possessing identifying information imposes on them certain obligations to protect the anonymity of their subjects (Privacy and Behavioral Research, 1967), few attempts have been made to develop improved techniques for insuring data security and respondent anonymity. That such efforts are sorely needed is evident from massive anecdotal evidence (e.g., Westin, 1967), and from empirical and systemic studies (e.g., Nugent, 1969; Boruch, 1969).