The study was an international collaborative research enterprise with a cross-national description and explanation of juvenile delinquency as its main objective. The major purpose of the ISRD-2 study was to examine
correlates of juvenile delinquency and victimization
derived from criminological theories. Specifically, there were eleven objectives for this study including the following:
- The data described the prevalence and incidence of offending and victimization among youths between the ages 12 and 15 (corresponded to grades seven to nine or the first, second, and third class in secondary
schools in most participating countries).
- The data obtained measures of the relative rank ordering of
prevalence of different types of youthful misbehavior
- The data examined cross-national variability in patterns of
correlates of self-reported delinquent behavior.
- The data described cross-national differences in the
importance of minority status with respect to selfreported
offending and victimization patterns in
this age group.
- The data learned more about correlates of criminal behavior
in this age group and tested different explanations
of crime, such as social control, self control, social
disorganization and life style theory.
- The data examined the importance of the school and
neighborhood context of this age group's
- The data described the aspects of delinquent trajectories
among this age group in countries that participated,
such as age of onset, frequency and versatility.
- The data described the reactions of official authorities and those of other agents, such as parents, teachers or
shopkeepers, to juvenile delinquency in this age
- The data studied the importance of micro-level (individual),
meso-level (school and neighborhood), and
macro-level (city and country) variables for self-reported
delinquency in this age group in countries that participated.
- The data advanced knowledge of the methodological
issues involved in conducting cross-national survey
- The data contributed to the development of repeat studies
to measure trends in youth delinquency over time
in a number of (primarily) European and North
American cities and countries.
The ISRD-2 study was conducted in a total of 31 countries. School classes were the primary sampling units and the aim was to have about 2,100 youths per participating country. Children between the seventh and ninth grade were given ISRD-2 questionnaire. Most of the surveys were conducted in a classroom setting, and self-administered (pencil-and-paper) by the students (generally, with supervision by researchers; in some cases, with supervision by teachers). In a few countries, the administration of the questionnaires was computerized (Switzerland, Denmark and
The ISRD-2 study was conducted in 15 western
European countries, 12 of which were European Union (EU) member
states: Austria, Belgium, Denmark, Finland, France,
Germany, Ireland, Italy, The Netherlands, Portugal,
Spain, Sweden, Iceland, Norway and Switzerland. In
addition, ten countries in the eastern part of Europe did
participate, of which six new EU member states were
funded by the European commission (one EU member
state joined the study after the application was introduced),
and three non-EU members were funded by
the Swiss National Science Foundation: Cyprus, the
Czech Republic, Estonia, Hungary, Lithuania, Poland,
Slovenia, Armenia, Bosnia-Herzegovina, and Russia.
Furthermore, Canada and the United States, represented
by four states (Illinois, Massachusetts, New Hampshire
and Texas), were part of the study, and countries outside Europe and North
America participated as well; Aruba, Netherlands Antilles, Suriname, and Venezuela.
The ISRD-2 was a school-based study with school classes as primary sampling units; the aim was to have about 2,100 youths per participating country. The sampling process involved two stages: (1) selection of cities/ towns; and (2) achievement of a random sample of classrooms from the seventh, eighth and ninth grades (i.e. of classes of 12 and 13 to 14 and 15 year old students) in the cities and towns. The city-based sampling design was based on a minimum of five cities or towns per country. The main selection criteria was size, degree of urbanization, and demographic and economic variables. The aim was to obtain three subsamples, including a metropolitan area (defined as one of the main economic centers of a country with a population between 500,000 and one million inhabitants), a medium sized city (of size 100,000 plus or minus 20 percent inhabitants), and three small rural towns (10,000-75,000 inhabitants). The design allowed for optional additional samples for those who wished to enlarge the scope of their sample, for example, adding specific, significant cities, in terms of geographic or economic criteria and differential crime rates. The three subsample groups were equally represented in the final sample: a metropolitan subsample with 700 students, a mid-size city subsample with 700 students, and a small town cluster subsample with 700 students (combined from three small towns). Each country attempted to select cities, which were considered typical for the country. The selected cities were comparable to other cities/towns of the same size. Although not selected randomly and limited in numbers (and in the potential to generalize), the cities that were used provided a reasonable representation of countries that participated.
The second stage of the sample selection was random. The sampling
plan asked for a random selection of seventh, eighth and ninth
grade classrooms in the selected cities (represented
700 students each, 2,100 total). All samples were stratified to grade level (seventh, eighth and ninth grade),
some additional to school type (academic, technical or
vocational). The minimum core sample was randomly
selected from among the seventh, eighth and ninth grade classrooms at the schools in the selected cities/towns or nations. A stratified multi-stage sampling procedure was used. First, a listing of all secondary educational schools of the selected cities was created. It included public and
private schools, vocational, technical and academic
schools. Then, a listing of all seventh, eighth and ninth grade
classrooms that were in the institutions was constructed. The number
of students drawn was proportional to the proportion
of students in each school type. The achieved sample size of the merged ISRD-2 data set was 71,400 cases.
Students who were between the ages of 12 and 15 years old or grades seven through nine in the 31 countries that took part in the study.
administrative records data
The data contain a total of 695 variables covering the following topics:
- Social demographics include variables about age, gender, family composition, socio-economic status and education level.
- Delinquent acts include variables about lifetime prevalence, current prevalence, frequency, the age of onset, the circumstances of the act, and social reactions to the offense of a number of different delinquent acts.
- Theoretical variables include variables about relationships with parents, parental supervision, attachment to school, commitment to school, truancy and information on peers.
- Victimization variables include variables about whether the respondent has ever been the victim of extortion, physical violence, theft, bullying and whether they reported offenses to police.
- Lifestyle variables include variables about leisure occupations, friends of different religion or ethnic groups and number of delinquent friends.
- Attitude toward violence variables include whether the respondent thought a bit of violence was fun, whether one needed to make use of force to be respected, whether they would attack someone if attacked first and whether they thought it is was normal that boys wanted to prove themselves in physical fights with others.
- Grasmick self-control scale variables include items on impulsivity, risk seeking, self-centeredness and temper.
- School context variables include what the respondents' school had to offer, what did school mean to respondent, and whether stealing, fighting, vandalism and drug use happened in school.
- Life event variables include death or serious illness of parent/family member, parental conflicts and separation/divorce of parents.
- Information on neighborhood variables include attachment, cohesion and disorganization of neighborhood.
It was not possible to calculate very precise estimates of the overall school cooperation rate for the entire sample, but a rough estimate of the overall positive response rates of schools that participated in the total sample was about 74 percent. That
combined with the noted overall very low parent or student nonresponse or refusal rates at less than five percent for the total sample (with the exceptions of the Czech Republic, Poland, Canada and the US), and it can be seen that the total ISRD-2 sample had a response
rate of somewhere between 65 and 70 percent.
The Grasmick self-control scale: measured impulsivity, risk seeking, self-centeredness and temper.
Several Likert type scales were used.