Self-Report Studies

History and Findings

In 1947, a sample of adults was mailed a self-report survey, and 99 percent of the respondents admitted to having committed at least one offense. In another such study, admitted delinquency among Texas college students was found to be as high as officially processed delinquents. However, it was a 1957 study of Nashville students by James F. Short Jr. and F. Ivan Nye that set the tone for most of the self-report studies to come. These students committed more crimes than were reported to the police, and most students reported having committed crimes. Such results were taken as an indication of a higher rate of deviance and crime among a general population of “noncriminals.” Further, the picture painted by official statistics (the Uniform Crime Reports, or UCR, primarily), which suggested that lower-class juveniles were far more delinquent than juveniles from other, higher classes, was shown to be false. While lower-class youth committed slightly more serious, and more frequent, delinquencies, middle- and upper-class youth also reported a substantial amount of delinquent acts. The only official-statistic picture that remained the same was the gender difference, with males reporting about nine offenses for every one committed by females.

Some methods that have been employed in self-report surveys have included totally anonymous questionnaires, confidential questionnaires in which the respondent is identifiable and can be later checked by interviews and police records, and signed questionnaires that can be validated against official records. Other variations include anonymous questionnaires with an identifying number that can be later validated by follow-up interviews and threat of polygraph (lie detector test), interviews, and interviews that can be checked against official records. In all of these approaches, “what people claim” is checked against independent criteria or “what people do.”

Modern Self-Report Surveys

Self-report surveys have been used in a variety of settings and populations. Inmate reports of past conduct have been used to construct parole prediction (risk) scales. Such crime admissions by inmates have been found to complement official data. Surveys of youth drug use have commonly used self-report data. The best known of these is the National Youth Survey. This survey examines 11- to 17-year-olds and features a national probability sample and a multicohort design. The National Youth Survey was found to yield valid (accurate) and reliable (consistent) results and indicated delinquency/crime levels higher than the National Criminal Victimization Survey or the UCR.

A particularly innovative example of the use of self-report data was the Drug Use Forecasting Program, which is now called Arrestee Drug Abuse Monitoring (ADAM). In this annual program sponsored by the National Institute of Justice (NIJ), volunteer arrestees in various cities are asked details about their current and past drug usage at the time of their arrest. Such subjects are guaranteed anonymity and assured that none of the information that is provided will be used against them. Obviously, arrestees would not normally be viewed as the most honest in answering such questions. The program asked those surveyed to provide urine specimens to be tested for drugs. Begun in 1987 with 23 cities, plans are to expand the program to 75 cities. Also in 1998, the NIJ launched International ADAM, involving a partnership with criminal justice agencies in many countries and providing a framework for the global assessment of drug use and for strengthening drug control policies and their coordination. In 2000, the NIJ began plans for publishing an annual report based on ADAM that includes trends in drug use, gun markets, gang migration, the suburbanization of crime, and drug treatment availability.

...