Academic journal article Journal of Drug Issues

A Methodology for Reducing Respondent Duplication and Impersonation in Samples of Hidden Populations

Academic journal article Journal of Drug Issues

A Methodology for Reducing Respondent Duplication and Impersonation in Samples of Hidden Populations

Article excerpt

A dilemma arises for researchers who sample hidden populations, such as injection drug users (IDUs), and use financial incentives to recruit respondents. To prevent respondent duplication (a subject participates in a study multiple times by using different identities) and respondent impersonation (a subject assumes the identity of other respondents), researchers must confirm their subjects' identities. Documentation, however, introduces sampling bias against those who lack such identification, or who wish to remain anonymous. Definitive forms of identification like photography and fingerprints introduce a bias against the more distrustful members of the population, and scanner-based biometrics can be expensive. Most research projects therefore rely on staff to recognize former respondents, but staff turnover and a large number of respondents compromise accuracy. We describe and assess quantitatively the accuracy of a method for subject identification based on a statistical principle, the interchangeability of indicators, in which multiple weak indicators combine to form a stronger aggregate measure. The analysis shows that observable indicators of identity (scars, birthmarks, tattoos, eye color, ethnicity, and gender) and five biometric measures (height, forearm lengths, and wrist widths) provide the basis for a reliable and easily administered method for subject identification.

INTRODUCTION

AIDS has accentuated the problem of accessing members of hidden populations for research and intervention purposes, especially in the case of injection drug users (IDUs). Hidden populations have three characteristics: no sampling frame is available; their members are objects of hate and stigmatization within the larger society; and their members are distrustful and work hard to avoid identification.

The research literature regarding IDUs has focused primarily on improving ways to reach and recruit them into prevention services (Wiebel, 1988; Broadhead & Heckathorn, 1994; Cunningham, Cottler, & Compton, 1996; Heckathorn, Broadhead, Anthony, & Weakliem, 1999), drawing more diversified or representative samples (Biernacki & Waldorf, 1981; Watters & Biernacki, 1989; Spreen, 1992; Heckathorn, 1997), and estimating the size and composition of IDU communities within municipalities (Hser &Anglin, 1993; Frank &Snijder, 1994).

This paper focuses on a neglected methodological problem and source of sampling bias: respondent duplication and impersonation. Respondent duplication occurs when a respondent participates multiple times in a study by using different identities. Respondent impersonation occurs when a respondent participates in a study by assuming the identity of other respondents.

The difficulties that research and intervention projects face in keeping track of hundreds of IDU respondents should not be underestimated. In many municipalities, especially large urban centers where numerous drug-related research projects can be operating simultaneously-and have been for years-researchers commonly speak of "professional research subjects" who are continually cycling through many projects, or through a single project at its different intake sites. The second author's extensive field study of a large AIDS prevention project for IDUs in San Francisco (Broadhead &Fox, 1990; Broadhead &Margolis, 1993) included many IDUs who described with amusement being interviewed by the same project several times, using different names at any one of its many intake locations around town. Although the financial rewards for a single interview are generally small-$10 to $50 depending on the research project and the demands made upon respondents-they can add up if a subject gives many interviews over several days or weeks. Interview staff also speak of veteran respondents who have learned what responses will substantially shorten an interview from, for example, 1.5 hours to 20 minutes, by claiming to have no sexual partners and to use only a single type of drug. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.