National Academies Press: OpenBook

Measurement Problems in Criminal Justice Research: Workshop Summary (2003)

Chapter: 3. Comparison of Self-Report and Official Data for Measuring Crime

« Previous: 2. Measuring Crime and Crime Victimization: Methodological Issues
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

3
Comparison of Self-Report and Official Data for Measuring Crime

Terence P. Thornberry and Marvin D. Krohn

There are three basic ways to measure criminal behavior on a large scale. The oldest method is to rely on official data collected by criminal justice agencies, such as data on arrests or convictions. The other two rely on social surveys. In one case, individuals are asked if they have been victims of crime; in the other, they are asked to self-report their own criminal activity. This paper reviews the history of the third method—self-report surveys—assesses its validity and reliability, and compares results based on this approach to those based on official data. The role of the self-report method in the longitudinal study of criminal careers is also examined.

HISTORICAL OVERVIEW

The development and widespread use of the self-report method of collecting data on delinquent and criminal behavior together were one of the most important innovations in criminology research in the twentieth century. This method of data collection is used extensively both in the United States and abroad (Klein, 1989). Because of its common use, we often lose sight of the important impact that self-report studies have had on the study of the distribution and patterns of crime and delinquency, the etiological

This study was supported by the National Consortium on Violence Research.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

study of criminality, and the study of the juvenile justice and criminal justice systems.

Sellin made the simple but critically important observation that “the value of a crime rate for index purposes decreases as the distance from the crime itself in terms of procedure increases” (1931:337). Thus, prison data are less useful than court or police data as a measure of actual delinquent or criminal behavior. Moreover, the reactions of the juvenile and criminal justice systems often rely on information from victims or witnesses of crime. It does not take an expert on crime to recognize that a substantial amount of crime is not reported and, if reported, is not officially recorded. Thus, reliance on official sources introduces a number of layers of potential bias between the actual behavior and the data. Yet, through the first half of the twentieth century, our understanding of the behavior of criminals and those who reacted to crime was based almost entirely on official data.

While researchers were aware of many of these limitations, the dilemma they faced was how to obtain valid information on crime that was closer to the source of the behavior. Observing the behavior taking place would be one method of doing so, but given the illegal nature of the behavior and the potential consequences if caught committing the behavior, participants in crime are reluctant to have their behavior observed. Even when observational studies have been conducted—for example, gang studies (e.g., Thrasher, 1927)—researchers could observe only a very small portion of the crimes that took place. Hence, observational studies had limited utility in describing the distribution and patterns of criminal behavior.

If one could not observe the behavior taking place, self-reports of delinquent and criminal behavior would be the data source nearest to the actual behavior. There was great skepticism, however, about whether respondents would be willing to tell researchers about their participation in illegal behaviors. Early studies (Porterfield, 1943; Wallerstein and Wylie, 1947) found that not only were respondents willing to self-report their delinquency and criminal behavior, they did so in surprising numbers.

Since those very early studies, the self-report methodology has become much more sophisticated in design, making it more reliable and valid and extending its applicability to myriad issues. Much work has been done to improve the reliability and validity of self-reports, including the introduction of specialized techniques intended to enhance the quality of self-report data. These developments have made self-report studies an integral part of the way crime and delinquency are studied.

Although the self-report method began with the contributions of

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Porterfield (1943, 1946) and Wallerstein and Wylie (1947), the work of Short and Nye (1957, 1958) “revolutionized ideas about the feasibility of using survey procedures with a hitherto taboo topic” and changed how the discipline thought about delinquent behavior itself (Hindelang et al., 1981:23). Short and Nye’s research is distinguished from previous self-report measures in their attention to methodological issues, such as scale construction, reliability and validity, and sampling and their explicit focus on the substantive relationship between social class and delinquent behavior. A 21-item list of criminal and antisocial behaviors was used to measure delinquency, although in most of their analyses a scale comprised of a subset of only seven items was employed. Focusing on the relationship between delinquent behavior and the socioeconomic status of the adolescents’ parents, Nye et al. (1958) found that relatively few of the differences in delinquent behavior among the different socioeconomic status groups were statistically significant.

Short and Nye’s work stimulated much interest in both use of the self-report methodology and the relationship between some measure of social status (socioeconomic status, ethnicity, race) and delinquent behavior. The failure to find a relationship between social status and delinquency served at once to question extant theories built on the assumption that an inverse relationship did in fact exist and to suggest that the juvenile justice system may be using extra-legal factors in making decisions concerning juveniles who misbehave. A number of studies in the late 1950s and early 1960s used self-reports to examine the relationship between social status and delinquent behavior (Akers, 1964; Clark and Wenninger, 1962; Dentler and Monroe, 1961; Empey and Erickson, 1966; Erickson and Empey, 1963; Gold, 1966; Reiss and Rhodes, 1959; Slocum and Stone, 1963; Vaz, 1966; Voss, 1966). These studies advanced the use of the self-report method by applying it to different, more ethnically diverse populations (Clark and Wenninger, 1962; Gold, 1966; Voss, 1966), attending to issues concerning validity and reliability (Clark and Tifft, 1966; Dentler and Monroe, 1961; Gold, 1966), and constructing measures of delinquency that specifically addressed issues regarding offense seriousness and frequency (Gold, 1966). These studies found that, while most juveniles engaged in some delinquency, relatively few committed serious delinquency repetitively. With few exceptions, these studies supported the general conclusion that, if there were any statistically significant relationship between measures of social status and self-reported delinquent behavior, it was weak and clearly did not mirror the findings of studies using official data sources.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

During this period of time researchers began to recognize the true potential of the self-report methodology. By including questions concerning other aspects of an adolescent’s life as well as a delinquency scale on the same questionnaire, researchers could explore a host of etiological issues. Theoretically interesting issues concerning the family (Dentler and Monroe, 1961; Gold, 1970; Nye et al., 1958; Stanfield, 1966; Voss, 1964), peers (Erickson and Empey, 1963; Gold, 1970; Matthews, 1968; Reiss and Rhodes, 1964; Short, 1957; Voss, 1964), and school (Elliott, 1966; Gold, 1970; Kelly, 1974; Polk, 1969; Reiss and Rhodes, 1963) emerged as the central focus of self-report studies. The potential of the self-report methodology in examining etiological theories of delinquency was perhaps best displayed in Hirschi’s (1969) Causes of Delinquency.

The use of self-report studies to examine theoretical issues continued throughout the 1970s. In addition to several partial replications of Hirschi’s arguments (Conger, 1976; Hepburn, 1976; Hindelang, 1973; Jensen and Eve, 1976), other theoretical perspectives such as social learning theory (Akers et al., 1979), self-concept theory (Jensen, 1973; Kaplan, 1972), strain theory (Elliott and Voss, 1974; Johnson, 1979), and deterrence theory (Anderson et al., 1977; Jensen et al., 1978; Silberman, 1976; Waldo and Chiricos, 1972) were evaluated using data from self-report surveys.

Another development during this period was the introduction of national surveys on delinquency and drug use. Williams and Gold (1972) conducted the first nationwide survey, with a probability sample of 847 boys and girls 13 to 16 years old. Monitoring the Future (Johnston et al., 1996) is a national survey on drug use that has been conducted annually since 1975. It began as an in-school survey of a nationally representative sample of high school seniors and was expanded to include eighth- and tenth-grade students.

One of the larger undertakings on a national level is the National Youth Survey (NYS), conducted by Elliott and colleagues (1985). The NYS began in 1976 by surveying a national probability sample of 1,725 youth ages 11 through 17. The survey design was sensitive to a number of methodological deficiencies of prior self-report studies and has been greatly instrumental in improving the self-report method. The NYS is also noteworthy because it is a panel design, having followed the original respondents into their thirties.

Despite the expanding applications of this methodology, questions remained about what self-report instruments measure. The discrepancy in findings regarding the relationship between social status and delinquency

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

based on self-report data versus official (and victim) data continued to perplex scholars. Early on, self-reports came under heavy criticism on a number of counts, including the selection of respondents and the selection of delinquency items. Nettler (1978:98) stated that “an evaluation of these unofficial ways of counting crime does not fulfill the promise that they would provide a better enumeration of offensive activity.” Gibbons (1979:84) was even more critical in his summary evaluation, stating:

The burst of energy devoted to self-report studies of delinquency has apparently been exhausted. This work constituted a criminological fad that has waned, probably because such studies have not fulfilled their early promise.

Two studies were particularly instrumental at that time in pointing to flaws in self-report measures. Hindelang and colleagues (1979) illustrated the problems encountered when comparing the results from studies using self-reports and those using official data or victimization data by comparing characteristics of offenders across the three data sources. They observed more similarity in those characteristics between victimization and Uniform Crime Reports data than between self-report data and the other two sources. They argued that self-report instruments did not include the more serious crimes for which people are arrested and that are included in victimization surveys. Thus, self-reports tap a different, less serious domain of behaviors than either of the other two sources, and discrepancies in observed relationships when using self-reports should not be surprising. The differential domain of crime tapped by early self-report measures could also explain the discrepancy in findings regarding the association between social status and delinquency.

Elliott and Ageton (1980) also explored the methodological shortcomings of self-reports. They observed that a relatively small number of youth commit a disproportionate number of serious offenses. However, most early self-report instruments failed to include serious offenses in the inventory and truncated the response categories for the frequency of offenses. In addition, many of the samples did not include enough high-rate offenders to clearly distinguish them from other delinquents. By allowing respondents to report the number of delinquent acts they committed rather than specifying an upper limit (e.g., 10 or more) and by focusing on high-rate offenders, Elliott and Ageton found relationships between engaging in serious delinquent behavior and race and social class that are more consistent with results from studies using official data.

Hindelang and colleagues (1979) and Elliott and Ageton (1980) sug

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

gested designing self-report studies so that they would acquire sufficient data from those high-rate, serious offenders who would be most likely to come to the attention of the authorities. They also suggested a number of changes in the way in which self-report data are measured, so that the data reflect the fact that some offenders contribute disproportionately to the rate of serious and violent delinquent acts.

The development of instruments to better measure serious offenses and the suggestion to acquire data from high-rate offenders coincided with a substantive change in the 1980s in the focus of much criminology work on the etiology of offenders. The identification of a relatively small group of offenders who commit a disproportionate amount of crime and delinquency led for a call to focus research efforts on “chronic” or “career” criminals (Blumstein et al., 1986; Wolfgang et al., 1972, 1987). Blumstein et al.’s observation that we need to study the careers of criminals, including early precursors of delinquency, maintenance through the adolescent years, and later consequences during the adult years, was particularly important in recognizing the need for examining the life-course development of high-rate offenders with self-report methodology.

The self-report methodology continues to advance in terms of both its application to new substantive areas and the improvement of its design. Gibbons’s (1979) suggestion that self-reports were just a fad, likely to disappear, is clearly wrong. Rather, with improvements in question design, administration technique, reliability and validity, and sample selection, this technique is being used in the most innovative research on crime and delinquency. The sections that follow describe the key methodological developments that have made such applications possible.

DEVELOPMENT OF THE SELF-REPORT METHOD

Self-report measures of delinquent behavior have advanced remarkably in the 30-odd years since their introduction by Short and Nye (1957). Considerable attention has been paid to the development and improvement of their psychometric properties. The most sophisticated and influential work was done by Elliott and colleagues (Elliott and Ageton, 1980; Elliott et al., 1985; Huizinga and Elliott, 1986) and by Hindelang, Hirschi, and Weis (1979, 1981). From their work a set of characteristics for acceptable (i.e., reasonably valid and reliable) self-report scales has emerged. Five of the most salient of these characteristics are the inclusion of (1) a wide array of offenses, including serious offenses; (2) frequency response sets; (3)

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

screening for trivial behaviors; (4) application to a wider age range; and (5) the use of longitudinal designs. Each is discussed below.

Inclusion of a wide array of delinquency items. The domain of crime covers a wide range of behaviors, from petty theft to aggravated assault and homicide. If the general domain of delinquent and criminal behavior is to be represented in a self-report scale, it is necessary for the scale to cover that same wide array of human activity. Simply asking about a handful of these behaviors does not accurately represent the theoretical construct of crime. In addition, empirical evidence suggests that crime does not have a clear unidimensional structure that would facilitate the sampling of a small number of items from a theoretically large pool to adequately represent the entire domain.

These considerations suggest that an adequate self-report scale for delinquency will be relatively lengthy. Many individual items are required to represent the entire domain of delinquent behavior, to represent each of its subdomains, and to ensure that each subdomain (e.g., violence, drug use) is itself adequately represented.

In particular, it is essential that a general self-reported delinquency scale tap serious as well as less serious behaviors. Early self-report scales tended to ignore serious criminal and delinquent events and concentrated almost exclusively on minor forms of delinquency. Failure to include serious offenses misrepresents the domain of delinquency and contaminates comparisons with other data sources. In addition, it misrepresents the dependent variable of many delinquency theories (e.g., Elliott et al., 1985; Thornberry, 1987) that attempt to explain serious, repetitive delinquency.

Inclusion of frequency response sets. Many early self-report studies relied on response sets with a relatively small number of categories, thus tending to censor high-frequency responses. For example, Short and Nye (1957) used a four-point response with the highest category being “often.” Aggregated over many items, these limited response sets had the consequence of lumping together occasional and high-rate delinquents, rather than discriminating between these behaviorally different groups.

Screening for trivial behaviors. Self-report questions have a tendency to elicit reports of trivial acts that are very unlikely to elicit official reactions and even acts that are not violations of the law. This occurs more frequently with the less serious offenses but also plagues responses to serious

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

offenses. For example, respondents have included as thefts such pranks as hiding a classmate’s books in the respondent’s locker between classes, or as serious assault events that are really roughhousing between siblings.

Some effort must be made to adjust or censor the data to remove these events if the delinquency of the subjects is to be reflected properly and if the rank order of subjects with respect to delinquency is to be portrayed properly. Two strategies are generally available. First, one can ask a series of follow-up questions designed to elicit more information about an event, such as the value of stolen property, the extent of injury to the victim, and the like. Second, one can use an open-ended question asking the respondent to describe the event and then probe to obtain the information necessary to classify the act. Both strategies have been used with some success.

Application to a wider age range. With increasing emphasis on the study of crime across the entire life course, self-report surveys have had to be developed to take into account both the deviant behavior of very young children and the criminal behavior of older adults. The behavioral manifestations of illegal behaviors or the precursors of such behavior can change depending on the stage in the life course at which the assessment takes place. For the very young child, measures have been developed that are administered to parents to assess antisocial behavior such as noncompliance, disobedience, and aggression (Achenbach, 1992). For the school-age child, Loeber and colleagues (1993) have developed a checklist that expands the range of antisocial behaviors to include such behaviors as stubbornness, lying, bullying, and other externalizing problems.

There has been less development of instruments targeted at adults. Weitekamp (1989) has criticized self-report studies for being primarily concerned with the adolescent years and simply using the same items for adults. This is particularly crucial given the concern over the small but very significant problem of chronic violent offenders.

Use of longitudinal designs. Perhaps the most significant development in the application of the self-report methodology is its use in following the same subjects over time in order to account for changes in their criminal behavior. This has enabled researchers to examine the effect of age of onset, to track the careers of offenders, to study desistance, and to apply developmental theories to study both the causes and consequences of criminal behavior over the life course.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

While broadening the range of issues that can be examined, application of the self-report technique within longitudinal panel designs introduces potential threats to the reliability and validity of the data. In addition to concern over construct continuity in applying the technique to different-aged respondents, researchers need to consider the possibility of panel or testing effects.

All of these newer procedures are likely to improve the validity, and to some extent the reliability, of self-report scales since they improve our ability to identify delinquents and to discriminate among different types of delinquents. These are clearly desirable qualities.

To gain these desirable qualities, however, requires a considerable expansion of the self-report schedule. This can be illustrated by describing the major components of the index currently being used in the Rochester Youth Development Study (Thornberry et al., in press) as well as the other two projects of the Program of Research on the Causes and Correlates of Delinquency (see Browning et al., 1999). The inventory includes 32 items that tap general delinquency and 12 that tap drug use, for a total of 44 items. For each item the subjects are asked if they committed the act since the previous interview. For all items to which they respond in the affirmative, a series of follow-up questions are asked, such as whether they had been arrested. In addition, for the most serious instance of each type of delinquency reported in the past six months, subjects are asked to describe the event by responding to the question: “Could you tell me what you did?” If that open-ended question does not elicit the information needed to describe the event adequately, a series of questions, which vary from 2 to 14 probes depending on the offense, are asked.

Although most of these specific questions are skipped for most subjects since delinquency remains a rare event, this approach to measuring self-reported delinquency is a far cry from the initial days of the method, when subjects used a few categories to respond to a small number of trivial delinquencies with no follow-up items. Below we evaluate the adequacy of this approach for measuring delinquent and criminal behavior.

RELIABILITY AND VALIDITY

For any measure to be scientifically worthwhile it must possess both reliability and validity. Reliability is the extent to which a measuring procedure yields the same result on repeated trials. No measure is absolutely, perfectly reliable. Repeated use of a measuring instrument will always pro

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

duce some variation from one application to another. That variation can be very slight or quite large. So the central question in assessing the reliability of a measure is not whether it is reliable but how reliable it is; reliability is always a matter of degree.

Validity is a more abstract notion. A measure is valid to the extent to which it measures the concept you set out to measure and nothing else. While reliability focuses on a particular property of the measure—namely, its stability over repeated uses—validity concerns the crucial relationship between the theoretical concept one is attempting to measure and what one actually measures. As is true with the case of reliability, the assessment of validity is not an either/or proposition. There are no perfectly valid measures, but some measures are more valid than others. We now turn to an assessment of whether self-reported measures of delinquency are psychometrically acceptable.

Assessing Reliability

There are two classic ways to assess the reliability of social science measures: test-retest reliability and internal consistency. Huizinga and Elliott (1986) make a convincing case that the test-retest approach is fundamentally more appropriate for assessing self-reported measures of delinquency.

Internal consistency means that multiple items measuring the same underlying concept should be highly intercorrelated. Although a reasonable expectation for attitudinal measures, this expectation is less reasonable for behavioral inventories such as self-report measures of delinquency. Current self-report measures typically include 30 to 40 items measuring a wide array of delinquent acts. Just because someone was truant is no reason to expect that they would be involved in theft or vandalism. Similarly, if someone reports that they have been involved in assaultive behavior, there is no reason to assume they have been involved in drug sales or loitering. Indeed, given the relative rarity of involvement in delinquent acts, it is very likely that most people will respond in the negative to most items and in the affirmative to only a few items. This is especially the case if we are asking about short reference periods (e.g., the past year or past six months). There is no strong underlying expectation that the responses will be highly intercorrelated, and therefore an internal consistency approach to assessing reliability may not be particularly appropriate. (See Huizinga and Elliott, 1986, for a more formal discussion of this point.)

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Some theories of crime (e.g., Gottfredson and Hirschi, 1990; Jessor et al., 1991) assume there is an underlying construct, such as self-control, that generates versatility in offending. If so, there should be high internal consistency among self-reported delinquency items. While this result may be supportive of the theoretical assumption, it is not necessarily a good indicator of the reliability of the measures. If internal consistency were low, it may not have any implication for reliability but may simply mean that this particular theoretical assumption was incorrect. Nevertheless, we do note that studies that have examined the internal consistency of self-report measures generally find acceptable alpha coefficients. For example, Hindelang and colleagues report alphas between 0.76 and 0.93 for various self-report measures (1981:80).

We will focus our attention on the test-retest method of assessing reliability. This approach is quite straightforward. A sample of respondents is administered a self-reported delinquency inventory (the test) and then, after a short interval, the same inventory is readministered (the retest). In doing this the same questions and the same reference period should be used at both times.

The time lag between the test and the retest is also important. If it is too short, it is likely the answers provided on the retest will be a function of memory. If so, estimates of reliability would be inflated. If the time period between the test and the retest is too great, it is likely the responses given on the retest would be less accurate than those given on the test because of memory decay. In this case the reliability of the scale would be underestimated. There is no hard-and-fast rule for assessing the appropriateness of this lag, but somewhere in the range of one to four weeks appears to be optimal.

A number of studies have assessed the test-retest reliability of self-reported delinquency measures. In general, the results indicate that these measures are acceptably reliable. The reliability coefficients vary somewhat depending on the number and types of delinquent acts included in the index and the scoring procedures used (e.g., simple frequencies or ever-variety scores), but scores well above 0.80 are common. In summarizing much of the previous literature in this area, Huizinga and Elliott (1986:300) state:

Test-retest reliabilities in the 0.85 - 0.99 range were reported by several studies employing various scoring schemes and numbers of items and using test-retest intervals of from less than one hour to over two months (Kulik et al., 1968; Belson, 1968; Hindelang et al., 1981; Braukmann et al., 1979;

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Patterson and Loeber, 1982; Skolnick et al., 1981; Clark and Tifft, 1966; Broder and Zimmerman, 1978).

Perhaps the most comprehensive assessment of the psychometric properties of the self-report method was conducted by Hindelang and his colleagues (1981). Their self-report inventory was quite extensive, consisting of 69 items divided into the following major subindices: official contact index, serious crime index, delinquency index, drug index, and school and family offenses index. While mindful of the limitations of internal consistency approaches, Hindelang and colleagues (1981) reported Cronbach’s alpha coefficients for a variety of demographic subgroups and for the ever-variety, last-year variety, and last-year frequency scores. The coefficients range from 0.76 to 0.93. Most of the coefficients are above 0.8, and 8 of the 18 coefficients are above 0.9.

Hindelang and colleagues (1981) also estimated test-retest reliabilities for these three self-report measures for each of the demographic subgroups. Unfortunately, only 45 minutes elapsed between the test and the retest, so it is quite possible the retest responses were strongly influenced by memory effects. Nevertheless, most of the test-retest correlations are above 0.9.

Hindelang et al. point out that reliability scores of this magnitude are higher than those typically associated with many attitudinal measures and conclude that “the overall implication is that in many of the relations examined by researchers, the delinquency dimension is more reliably measured than are many of the attitudinal dimensions studied in the research” (p. 82).

The other major assessment of the psychometric properties of the self-report method was conducted by Huizinga and Elliott (1986) using data from the NYS. At the fifth NYS interview, 177 respondents were randomly selected and reinterviewed approximately four weeks after their initial assessment. Based on these data, Huizinga and Elliott estimated test-retest reliability scores for the general delinquency index and a number of subindices. They also estimated reliability coefficients for frequency scores and variety scores.

The general delinquency index appears to have an acceptable level of reliability. The test-retest correlation for the frequency score is 0.75 and for the variety score, 0.84. For the various subindices—ranging from public disorder offenses to the much more serious index offenses—the reliabilities vary from a low of 0.52 (for the frequency measure of felony theft) to a high of 0.93 (for the frequency measure of illegal services). In total,

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Huizinga and Elliott report 22 estimates of test-retest reliability (across indices and across frequency and variety scores), and the mean reliability coefficient is 0.74.

Another way Huizinga and Elliott assessed the level of test-retest reliability is by estimating the percentage of the sample that changed frequency responses by two or less. If the measure is highly reliable, few changes would be expected over time. For most subindices there appears to be acceptable reliability based on this measure. For example, for index offenses 97 percent of respondents changed their answers by two delinquent acts or less. Huizinga and Elliott (1986:303) summarized these results as follows:

Scales representing more serious, less frequently occurring offenses (index offenses, felony assault, felony theft, robbery) have the highest precision, with 96 to 100 percent agreement, followed by the less serious offenses (minor assault, minor theft, property damage), with 80 to 95 percent agreement. The public disorder and status scales have lower reliabilities (in the 40 to 70 percent agreement range), followed finally by the general SRD [self-reported delinquency] scale, which, being a composite of the other scales, not surprisingly has the lowest test-retest agreement.

Huizinga and Elliott did not find any consistent differences across sex, race, class, place of residence, or delinquency level in terms of test-retest reliabilities (see also Huizinga and Elliott, 1983).

Assessing Validity

There are several ways to assess validity. We concentrate on three: content validity, construct validity, and criterion validity.

Content Validity

Content validity is a subjective or logical assessment of the extent to which a measure adequately reflects the full domain, or full content, that is contained in the concept being measured. To argue that a measure has content validity, the following three criteria must be met. First, the domain of the concept must be defined clearly and fully. Second, questions or items must be created that cover the whole range of the concept under investigation. Third, items or questions must be sampled from that range so that the ones that appear on the test are representative of the underlying concept.

A reasonable definition of delinquency and crime is the commission of

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

behaviors that violate criminal law and that place the individual at some risk of arrest if the behavior were known to the police. Can a logical case be made that self-report measures of delinquency are valid in this respect?

As noted above, the earlier self-report inventories contained relatively few items to measure the full range of delinquent behaviors. For example, Short and Nye’s (1957) inventory contains only 21 items, and most of their analysis was conducted with a 7-item index. Similarly, Hirschi’s self-report measure (1969) is based on only 6 items. More importantly, the items included in these scales are clearly biased toward the minor or trivial end of the continuum.

The more recent self-report measures appear to be much better in this regard. For example, the Hindelang and colleagues (1981) index includes 69 items that range from status offenses, such as skipping class, to violent crimes, like serious assault and armed robbery. The NYS index (Elliott et al., 1985) has 47 items designed to measure all but one (homicide) of the Uniform Crime Reports Part I offenses, 60 percent of the Part II offenses, and offenses that juveniles are particularly likely to commit. The self-report inventory used by the three projects of the Program of Research on the Causes and Correlates of Delinquency has 32 items that measure delinquent behavior and 12 that measure substance use.

These more recent measures, while not perfect, tap into a much broader range of delinquent and criminal behaviors. As a result, they appear to have reasonable content validity.

Construct Validity

Construct validity refers to the extent to which the measure being validated is related in theoretically expected ways to other concepts or constructs. In our case the key question is: Are measures of delinquency based on the self-report method correlated in expected ways with variables expected to be risk factors for delinquency?

In general, self-report measures of delinquency and crime, especially the more recent longer inventories, appear to have a high degree of construct validity. They are generally related in theoretically expected ways to basic demographic characteristics and to a host of theoretical variables drawn from various domains such as individual attributes, family structure and processes, school performance, peer relationships, neighborhood characteristics, and so forth. Hindelang and colleagues (1981) offer one of the clearer assessments of construct validity. They correlate a number of etio

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

logical variables with different self-report measures collected under different conditions. With a few nonsystematic exceptions, the correlations are in the expected direction and of the expected magnitude.

Overall, construct validity may offer the strongest evidence for the validity of self-report measures of delinquency and crime. Indeed, if one examines the general literature on delinquent and criminal behavior, it is surprising how few theoretically expected relationships are not observed for self-reported measures of delinquency and crime. It is unfortunate that this approach is not used to assess validity more formally and more systematically.

Criterion Validity for Delinquency and Crime

Criterion validity “refers to the relationship between test scores and some known external criterion that adequately indicates the quantity being measured” (Huizinga and Elliott, 1986:308). There is a fundamental difficulty in assessing the criterion validity of self-reported measures of delinquency and crime and for that matter all measures of delinquency and crime. Namely, there is no gold standard by which to judge the self-report measure. That is, there is no fully accurate assessment that can be used as a benchmark. In contrast, to test the validity of self-reports of weight, people could be asked to self-report their weight and each respondent could then be weighed on an accurate scale—the external criterion. Given the secretive nature of criminal behavior, however, there is nothing comparable to a scale in the world of crime. As a result, the best that can be done is to compare different flawed measures of criminal involvement to see if there are similar responses and results. If so, the similarity across different measurement strategies heightens the probability that the various measures are tapping into the underlying concept of interest. While not ideal, this is the best that can be done in this area of inquiry.

There are several ways to assess criterion validity. One of the simplest is called known group validity. In this approach one compares scores for groups of people who are likely to differ in terms of their underlying involvement in delinquency. For example, the delinquency scores of seminarians would be expected to be lower than the delinquency scores of street gang members.

Over the years a variety of group comparisons have been made to assess the validity of self-report measures. They include comparisons between individuals with and without official arrest records, between individuals

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

convicted and not convicted of criminal offenses, and between institutionalized adolescents and high school students. In all cases these comparisons indicate that the group officially involved with the juvenile justice system self-reports substantially more delinquent acts than the other group. (See, for example, the work by Erickson and Empey, 1963; Farrington, 1973; Hardt and Petersen-Hardt, 1977; Hindelang et al., 1981; Hirschi, 1969; Kulik et al., 1968; Short and Nye, 1957; and Voss, 1963.)

While comparisons across known groups are helpful, they offer a minimal test of criterion validity. The real issue is not whether groups differ but the extent to which individuals have similar scores on the self-report measure and on other measures of criminal behavior. A variety of external criteria have been used (see the discussion in Hindelang et al., 1981). The two most common approaches are to compare self-reported delinquency scores with official arrest records and self-reports of arrest records with official arrest records.

We can begin by examining the correlation between self-reported official contacts and official measures of delinquency. These correlations are quite high in the Hindelang et al. study, ranging from 0.70 to 0.83. Correlations of this magnitude are reasonably large for this type of data.1

The most recent investigation of this topic is by Maxfield, Weiler, and Widom (2000), using Widom’s (1989) sample of child maltreatment victims and their matched controls. Unlike most studies in this area, the respondents were adults (mean age = 28). They were interviewed only once, so all of the self-reported arrest data are retrospective, with relatively long recall periods. Nevertheless, the concordance between having an official arrest and a self-report of being arrested is high. Of those arrested, 73 percent reported an arrest. Maxfield et al. noted lower levels of reporting for females than males and for blacks than whites. The gender differences were quite persistent, but the race differences were more pronounced for less frequent offenders and diminished considerably for more frequent offenders.

Maxfield et al. also studied “positive bias,” the self-reporting of arrests that are not found in official records. They found that 21 percent of respondents with no arrest history self-reported being arrested. Positive bias

1  

This is particularly the case given the level of reliability that self-reported data have (see previous section). By adding random error to the picture, poor reliability attenuates or reduces the size of the observed correlation coefficients.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

was higher for males than females, but there were no race differences. It is not clear whether this is a problem with the self-reports (i.e., positive bias) or with the official records such as sealed records, sloppy record keeping, use of aliases, and so forth. This is an understudied topic that needs greater investigation.

The generally high level of concordance between self-reports of being arrested or having a police contact and having an official record has been observed in other studies as well (Hardt and Petersen-Hardt, 1977; Hathaway et al., 1960; Rojek, 1983). When convictions are examined, even higher concordance rates are reported (Blackmore, 1974; and Farrington, 1973).

It appears that survey respondents are quite willing to self-report their involvement with the juvenile justice and criminal justice systems. Are they also willing to self-report their involvement in undetected delinquent behavior? This is the central question. The best way to examine this is to compare self-reported delinquent behavior and official measures of delinquency. If these measures are valid, a reasonably large positive correlation between them would be expected.

Hindelang and colleagues (1981) presented correlations using a number of different techniques for scoring the self-report measures, but here we focus on the average correlation across these different measures and on the correlation based on the ever-variety scores, as presented in their Figure 3. Overall, these correlations are reasonably high, somewhere around 0.60 for all subjects. The most important data though are presented for race-bygender groups. For white and African American females and for white males, the correlations range from 0.58 to 0.65 when the ever-variety score is used; for the correlations that are averaged across the different self-report measures, the magnitudes range from 0.50 to 0.60. For African American males, however, the correlation is at best moderate. For the ever-variety self-reported delinquency score, the correlation is 0.35, and the average across the other self-reported measures is 0.30.

Huizinga and Elliott (1986), using data from the NYS, also examined the correspondence between self-reports of delinquent behavior and official criminal histories. They recognized that there can be considerable slippage between these two sources of data even when the same event is actually contained in both data sets. For example, while an adolescent can self-report a gang fight, it may be recorded in the arrest file as disturbing the peace, or an arrest for armed robbery can be self-categorized as a mugging or theft by the individual. Because of this, Huizinga and Elliott pro

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

vided two levels of matching, “tight matches” and “broad matches.” The analysis provides information on both the percentage of people who provide tight and broad matches to their arrest records and the percentage of arrests that are matched by self-reported behavior.

For the tight matches, almost half of the respondents (48 percent) concealed or forgot at least some of their offensive behavior, and about a third (32 percent) of all the offenses were not reported. When the broad matches are used, the percentage of respondents concealing or forgetting some of their offenses dropped to 36 percent and the percentage of offenses not self-reported to 22 percent. While the rates of underreporting are substantial, it should be noted that the majority of individuals who have been arrested self-report their delinquent behavior, and the majority of offenses they commit also are reported.

The reporting rates for gender, race, and social class groupings are quite comparable to the overall rates, with one exception. As was the case with the Seattle data, African American males substantially underreported their involvement in delinquency.

Farrington and colleagues (1996), using data from the middle and oldest cohorts of the Pittsburgh Youth Study, also examined this issue. The Pittsburgh study, as one of three projects in the Program of Research on the Causes and Correlates of Delinquency, uses the same self-reported delinquency index as described earlier for the Rochester Youth Development Study. Farrington et al. classified each of the boys in the Pittsburgh study into one of four categories based on the seriousness of their self-reported delinquency: no delinquency, minor delinquency only, moderate delinquency only, and serious delinquency. They then used juvenile court petitions as an external criterion to assess the validity of the self-reported responses. Both concurrent validity and predictive validity were assessed.

Overall, this analysis suggests that there is a substantial degree of criterion validity for the self-report inventory used in the Program of Research on the Causes and Correlates of Delinquency. Respondents who are in the most serious category based on their self-report responses are significantly more likely to have juvenile court petitions, both concurrently and predictively. For example, the odds ratio of having a court petition for delinquency is about 3:0 for the respondents in the most serious self-reported delinquency category versus the other three.

African American males are no more or less likely to self-report delinquent behavior than white males. With few exceptions, the odds ratios comparing self-reported measures and official court petitions are signifi

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

cant for both African Americans and whites; in some cases the odds ratios are higher for whites, and in some cases they are higher for African Americans.

These researchers also compared the extent to which boys with official court petitions self-reported being apprehended by the police. Overall, about two-thirds of the boys with court petitions answered in the affirmative. Moreover, there was no evidence of differential validity. Indeed, the African American respondents were more likely to admit being apprehended by the police than were the white respondents. Farrington and his colleagues (1996:509) concluded that “concurrent validity for admitting offenses was higher for Caucasians but concurrent validity for admitting arrests was higher for African Americans. There were no consistent ethnic differences in predictive validity.”

Finally, Farrington and colleagues (2000) used data from the Seattle Social Development Project to assess the concurrent and predictive validity of self-report data. They compared self-report responses for a variety of indices and offense types to the odds of having a court referral. For the general delinquency index the concurrent odds ratio was 2:8 and the predictive odds ratio was 2:2. Validity was highest for self-reports of drug involvement and lowest for property offenses, with violent offenses falling in the middle.

Putting all this together leads to a somewhat mixed assessment of the validity of self-report measures. On the one hand, it seems that the overall validity of self-report data is in the moderate-to-strong range, especially for self-reports of being arrested. For the link between self-reported delinquent behavior and official measures of delinquency, the only link based on independent sources of data, the overall correlations are somewhat smaller but still quite acceptable. On the other hand, looking at the issue of differential validity, there are some disturbing differences by race. It is hard to determine whether this is a problem with the self-report measures, the official measures, or both. We will return to a discussion of this issue after additional data are presented.

Criterion Validity for Substance Use

The previous studies focused on delinquent or criminal behavior where, as mentioned earlier, there is no true external criterion for evaluating validity. There is an external criterion for one class of deviant behavior—substance use. Physiological data (e.g., from saliva or urine) can be

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

used to independently assess recent use of various substances. The physiological data can then be compared to self-reports of substance use to assess the validity of the self-report instruments. A few examples of this approach can be offered.

We begin with a study of a minor form of deviant behavior—adolescent tobacco use. Akers and colleagues (1983) examined tobacco use among a sample of junior and senior high school students in Muscatine, Iowa. The respondents provided saliva samples that were used to detect nicotine use by the level of salivary thiocyanate. The students also self-reported whether they smoked and how often. The self-report data had very low levels of either underreporting of tobacco use or overreporting. Overall, the study estimated that 95 to 96 percent of the self-reported responses were accurate and valid.

The Drug Use Forecasting (DUF) project (1990), sponsored by the National Institute of Justice, is an ongoing assessment of the extensiveness of drug use for samples of arrestees in cities throughout the country. Individuals who have been arrested and brought to central booking stations are interviewed and asked to provide urine specimens. Both the urine samples and the interviews are provided voluntarily, and there is an 80 percent cooperation rate for the urine samples and a 90 percent cooperation rate for the interviews. The urine specimens are tested for 10 different drugs, and in some of the interviews there is a self-reported drug use inventory. Assuming the urine samples provide a reasonably accurate estimate of actual drug use, they can be used to validate self-reported information.

The results vary considerably by type of drug. There is generally a fairly high concordance for marijuana use. For example, in 1990 in New York City 28 percent of the arrestees self-reported marijuana use and 30 percent tested positive for marijuana use. Similarly, in Philadelphia 28 percent self-reported marijuana use and 32 percent tested positive. The worst comparison in this particular examination of the Drug Use Forecasting data was from Houston, where 15 percent of arrestees self-reported marijuana use and 43 percent tested positive.

For more serious drugs, the level of underreporting is much more severe. For example, 47 percent of the New York City arrestees self-reported cocaine use and 74 percent tested positive. Very similar numbers were generated in Philadelphia, where 41 percent self-reported cocaine use but 72 percent tested positive. Similar levels of underreporting were observed for other hard drugs such as heroin and in other cities.

The data collected in the Drug Use Forecasting project are obviously

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

quite different from those collected in typical self-report surveys. The sample is limited to people who have just been arrested, and they are asked to provide self-incriminating evidence to a research team while in a central booking station. It is not entirely clear how this setting affects the results. On the one hand, individuals may be reluctant to provide additional self-incriminating evidence after having just been arrested. On the other hand, if one has just been arrested for a serious crime like robbery or burglary, admitting to recent drug use may not be considered a big deal. In any event, caution is needed in using these data to generalize to the validity of typical self-report inventories.

SUMMARY

We have examined three different approaches to assessing the validity of self-reported measures of delinquency and crime: content, construct, and criterion validity. Several conclusions, especially for the more recent self-report inventories, appear warranted.

The self-report method for measuring this rather sensitive topic—undetected criminal behavior—appears to be reasonably valid. The content validity of the recent inventories is acceptable, the construct validity is quite high, and the criterion validity appears to be in the moderate-to-strong range. Putting this all together, it could be concluded that for most analytical purposes, self-reported measures are acceptably accurate and valid.

Despite this general conclusion, there are still several substantial issues concerning the validity of self-report measures. First, the validity of the earlier self-report scales, and the results based on them, are at best questionable. Second, based on the results of the tests of criterion validity, there appears to be a substantial degree of either concealing or forgetting past criminal behavior. While the majority of individual respondents report their offenses and the majority of all offenses are reported, there is still a good deal of underreporting.

Third, there is an unresolved issue of differential validity. As compared to other race-gender groups, some studies have found that the responses provided by African American males appear to have lower levels of validity (Hindelang et al., 1981; Huizinga and Elliott, 1986). More recently, however, Farrington et al. (1996) and Maxfield et al. (2000) found no evidence of differential validity by race. Maxfield and colleagues (2000) did find lower reporting for females than males. The level of differential validity is

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

one of the most important methodological issues confronting the self-report method and should be a high priority for future research efforts.

Fourth, based on studies of self-reported substance use, there is some evidence that validity may be less for more serious types of offenses. In the substance use studies, the concordance between the self-report and physiological measures was strongest for adolescent tobacco use, then for marijuana use, and it was weakest for hard drugs such as cocaine and heroin. A similar pattern is seen for several studies of self-reported delinquency and crime (e.g., Elliott and Voss, 1974; Huizinga and Elliott, 1986).

What then can be said about the psychometric properties of self-reported measures of delinquency and crime? With respect to reliability, this approach to measuring involvement in delinquency and crime appears to be acceptable. Most estimates of reliability are quite high, and there is no evidence of differential reliability. With respect to validity, the conclusion is a little murkier. There is a considerable amount of underreporting, and there is also the potential problem of differential validity. Nevertheless, content validity and construct validity appear to be quite high, and an overall estimate of criterion validity would be in the moderate-to-strong range. Perhaps the conclusion reached by Hindelang and colleagues (1981:114) is still the most reasonable:

The self-report method appears to behave reasonably well when judged by standard criteria available to social scientists. By these criteria, the difficulties in self-report instruments currently in use would appear to be surmountable; the method of self-reports does not appear from these studies to be fundamentally flawed. Reliability measures are impressive and the majority of studies produce validity coefficients in the moderate to strong range.

SPECIALIZED RESPONSE TECHNIQUES

Because of the sensitive nature of this area—asking people to report previously undetected criminal behavior—there has always been concern about how best to ask such questions to maximize the accuracy of the responses. Some early self-report researchers favored self-administered questionnaires while others favored more personal face-to-face interviews. Similarly, some argued that anonymous responses were inherently better than nonanonymous responses. In their Seattle study, Hindelang and his colleagues (1981) directly tested these concerns by randomly assigning respondents to one of four conditions: nonanonymous questionnaire, anonymous questionnaire, nonanonymous interview, and anonymous interview.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Their results indicate that there is no strong method effect in producing self-report responses, and that no one approach is consistently better than the other approaches. Similar results were reported by Krohn and his colleagues (1974). Some research, especially in the alcohol and drug use area, has found a method effect. For example, Aquilino (1994) found that admission of alcohol and drug use is lowest in telephone interviews, somewhat higher in face-to-face interviews, and highest in self-administered questionnaires (see also Aquilino and LoSciuto, 1990; Turner et al., 1992). While evident, the effect size is typically not very large.

Although basic method effects do not appear to be very strong, there is still concern that in all of these approaches to the collection of survey data, respondents will feel vulnerable about reporting sensitive information. Because of that, a variety of more specialized techniques have been developed to protect the individual respondent’s confidentiality, hopefully increasing the level of reporting.

Randomized Response Technique

The randomized response technique assumes that the basic problem with the validity of self-reported responses is that respondents are trying to conceal sensitive information; that is, they are unwilling to report undetected criminal behavior as long as there is any chance of others, including the researchers, linking the behavior to them. Randomized response techniques allow respondents to conceal what they really did while at the same time providing useful data to researchers. There are various ways to accomplish this, and how the basic process works can be illustrated with a simple example of measuring the prevalence of marijuana use. The basic question is: “Have you ever smoked marijuana?”

Imagine an interview setting in which there is a screen between the interviewer and respondent so that the interviewer cannot see what the respondent is doing. The interviewer asks a sensitive question (e.g., “Have you ever smoked marijuana?”) with the following special instruction: Before answering, please flip a coin. If the coin lands on heads, please answer “yes” regardless of whether or not you smoked marijuana. If the coin lands on tails, please tell me the truth. It is impossible for the interviewer to know whether a “yes” response is produced by the coin or by the fact that the respondent actually smoked marijuana. In this way the respondent can admit to sensitive behavior but other people, including the interviewer, do not know if the admission is truthful or not.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

From the resulting data the prevalence of marijuana use can be estimated quite easily. Say we receive 70 “yes” responses in a sample of 100 respondents. Fifty of those would be produced by the coin landing on heads, and these 50 respondents can simply be ignored. Of the remaining 50 respondents though, 20 said “yes” because they smoked marijuana, so the prevalence of marijuana use is 20 out of 50, or 40 percent.

This technique is not limited to “yes” or “no” questions or to flipping coins. Any random process can be used as long as the probability distribution of bogus versus truthful responses is known. From these data, prevalence, variety, and frequency scores and means and variances can be estimated, and the information can be correlated with other variables, just as is done with regular self-report data.

Weis and Van Alstyne (1979) tested a randomized response procedure in the Seattle study. They concluded that the randomized response approach is no more efficient in eliciting positive responses to sensitive items than are traditional methods of data collection. This finding is consistent with the overall conclusion in the Seattle study that the method of administration is relatively unimportant.

The other major assessment of the randomized response technique was conducted by Tracy and Fox (1981). They sampled people who had been arrested in Philadelphia and then went to their homes to interview them. Respondents were asked if they had been arrested and, if so, how many times. There were two methods of data collection: a randomized response procedure and a regular self-report interview.

The results indicated that the randomized response approach does make a difference. For all respondents there was about 10 percent less error in the randomized response technique. For respondents who had been arrested only once, the randomized response approach actually increased the level of error. But for recidivists the randomized response technique reduced the level of error by about 74 percent.

Also, the randomized response technique generated random errors; that is, the errors were not correlated with other important variables. The regular self-reported interview, however, generated systematic error or bias. In this approach, underreporting was related to females, African American females, respondents with a high need for approval, lower-income respondents, and persons with a larger number of arrests.

Overall, it is not clear to what extent a randomized response approach actually generates more complete and accurate reporting. The two major studies of this topic produced different results: Weis and Van Alstyne (1979)

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

reported no effect, and Tracy and Fox (1981) reported sizable positive effects.

Computer-Assisted Interviewing

Advances in both computer hardware and software have made the introduction of computers in the actual data collection process not only a possibility but, according to Tourangeau and Smith (1996:276), “perhaps the most commonly used method of face-to-face data collection today.” The use of computers in the data collection process began in the 1970s with computer-assisted telephone surveys (Saris, 1991). The technology was soon adapted to the personal interview setting with either the interviewer administering the schedule, the computer-assisted personal interview, or the respondent self-administering the schedule by reading the questions on the computer screen and entering the responses—the computer-assisted self-administered interview (CASI). It is also possible to have an audio version in which the questions are recorded and the respondent listens to them on headphones rather than having them read aloud by the interviewer. This is called an audio computer-assisted self-administered interview (ACASI).

One reason for the use of computer-assisted data collection that is particularly relevant for this paper is its potential for collecting sensitive information in a manner that increases the confidentiality of responses. Another advantage is that it allows for the incorporation of complex branching patterns (Beebe et al., 1998; Saris, 1991; Tourangeau and Smith, 1996; Wright et al., 1998). Computer software can be programmed to include skip patterns and increase the probability that the respondent will answer all appropriate questions. An added advantage of computer-assisted presentation is that the respondent does not see the implication of answering in the affirmative to questions with multiple follow-ups.

ACASI has two additional advantages. First, it circumvents the potential problem of literacy; the respondent does not have to read the questions. Second, in situations where other people might be nearby, the questions and responses are not heard by anyone but the respondent. Hence, the respondent can be more assured that answers to sensitive questions will remain private.

While computer-assisted administration of sensitive questions provides obvious advantages in terms of efficiency of presentation and data collection, the key question is the difference in the responses that are elicited

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

when such technology is used. Tourangeau and Smith (1996) reviewed 18 studies that compared different modes of data collection. The types of behavior examined included health problems (e.g., gastrointestinal problems), sexual practices, abortion, and alcohol and drug use. Tourangeau and Smith indicate that techniques that are self-administered generally elicit higher rates of problematic behaviors than those administered by an interviewer. Moreover, CASIs elicit higher rates than either self-administered questionnaires or paper-and-pencil interviews administered by an interviewer. Also, ACASIs elicit higher rates than CASIs.

Estimates of prevalence rates of illegal and embarrassing behaviors appear to be higher when computer-assisted techniques, particularly those involving self-administration, are used. The higher prevalence rates need to be externally validated. The added benefits of providing for schedule complexity and consistency in responses make these techniques attractive, and it is clear that they will continue to be used with increasing frequency.

SELF-REPORT MEASURES ACROSS THE LIFE COURSE

One of the most exciting developments in criminology over the past 15 years has been the emergence of a life-course or developmental focus (Farrington, 1986; Jessor, 1998; Thornberry, 1997; Thornberry and Krohn, 2001; Weitekamp, 1989). Theoretical work has expanded from a narrow focus on the adolescent years to encompass the entire criminal careers of individuals, from the precursors of delinquency that are manifest in early childhood (Moffitt, 1997; Tremblay et al., 1999) through the high-delinquency years of middle and late adolescence, on into adulthood when most, but not all, offenders decrease their participation in illegal behaviors (Loeber et al., 1998; Moffitt, 1997; Sampson and Laub, 1990; Thornberry and Krohn, 2001). Research on criminal careers (Blumstein et al., 1986) has documented the importance of examining such issues as the age of onset (Krohn et al., 2001) and the duration of criminal activity (Wolfgang et al., 1987).

In addition, a growing body of research has demonstrated that antisocial behavior is rather stable from childhood to adulthood (Farrington, 1989a; Huesmann et al., 1984; Moffitt, 1993; Olweus, 1979). Much of this work has relied on official data. However, criminological research increasingly relies on longitudinal panel designs using self-report measures of antisocial behavior to understand the dynamics of criminal careers. Nevertheless, relatively little attention has been paid to the use of self-report tech

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

niques in longitudinal studies over the life course, even though this introduces a number of interesting measurement issues. Several of these issues are discussed in this section. Some of them involve the construction of valid measures at different developmental stages; others involve the consequences of repeated measures.

Construct Continuity

While many underlying theoretical constructs such as involvement in crime remain constant over time, their behavioral manifestations can change as subjects age. Failure to adapt measures to account for these changes may lead to age-inappropriate measures with reduced validity and reliability. To avoid this, measures need to adapt to the respondent’s developmental stage to reflect accurately the theoretical constructs of interest (Campbell, 1990; LeBlanc, 1989; Patterson, 1993; Weitekamp, 1989). In some cases this may mean defining the concept at a level to accommodate the changing contexts in which people at different ages act. In other cases it may mean recognizing that different behaviors at different ages imply consistency in behavioral style (Campbell, 1990).

Construct continuity creates a difficult design dilemma. If the measure does not change to reflect the developmental stage, the accuracy of the measure is likely to deteriorate and the study of change is compromised. Changing the measure over time, however, creates its own set of problems, especially for the study of change. If change is observed, is it a function of changes in the person’s behavior or of changes in the measure?

Relatively little attention has been paid to this issue in the study of criminal careers and, in particular, the study of self-report measures. At a more practical level, several studies have adapted self-report measures to both childhood and adulthood.

Self-Report Measures for Children

Antisocial behavior has been likened to a chimera (Patterson, 1993) with manifestations that change and accumulate with age. At very young ages (2 to 5 years) behavioral characteristics such as impulsivity, noncompliance, disobedience, and aggression are seen as early analogs of delinquent behavior. At these young ages, self-report instruments are not practical. Rather, researchers have measured these key indicators through either parental reports or observational ratings. Many studies of youngsters at

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

these ages have used Achenbach’s (1992) Child Behavior Checklist (CBCL), which is a parent-completed inventory and has versions for children as young as 2 to assess “externalizing” problem behaviors. Studies using either the CBCL, some other parental or teacher report of problem behaviors, or observational ratings have demonstrated that there is a relationship between these early manifestations of problem behavior and antisocial behavior in school-age children (Belsky et al., 1996; Campbell, 1987; Richman et al., 1982; Shaw and Bell, 1993).2

Starting at school age, the range of antisocial behaviors expands to include stubbornness, lying, bullying, and other externalizing problems (Loeber et al., 1993). School-age children, even those as young as first grade, begin to participate in delinquent behaviors. However, self-report instruments of delinquent behavior have rarely been administered to preteenage children (Loeber et al., 1989). Some studies have administered self-report instruments to 10 or 11 year olds, slightly modifying the standard delinquency items (Elliott et al., 1985).

Loeber et al. (1989) provide one of the few attempts not only to gather self-report information from children under 10 but also to examine the reliability of those reports. They surveyed a sample of 849 first-grade and 868 fourth-grade boys using a 33-item self-reported antisocial behavior scale. This is a younger-age version of the self-reported delinquency index used by the three projects of the Program of Research on the Causes and Correlates of Delinquency. Items that were age appropriate were selected, and some behaviors were placed in a number of different contexts in order to make them less abstract for the younger children. A special effort was made to ensure that each child understood the question by preceding each behavior with a series of questions to ascertain whether the respondent knew the meaning of the behavior. If the child did not understand the question, the interviewer gave an example and then asked the child to do the same. If the child still did not understand the question, the item was skipped. The parents and teachers of these children also were surveyed using a combination of the appropriate CBCL and delinquency items.

Loeber and colleagues reported that the great majority of boys understood most of the items. First-grade boys did have problems understanding the items regarding marijuana use and sniffing glue, and fourth-grade boys had difficulty understanding the question regarding sniffing glue.

2  

The CBCL also assesses internalizing problems.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

To assess the validity of self-reported delinquent behavior among elementary school children, Loeber and his colleagues compared the children’s self-reports to parental reports about similar behaviors. They found a surprisingly high degree of concordance between children’s and parents’ reports about the prevalence of delinquent behavior. This is especially true for behaviors that are likely to come to the attention of parents, such as aggressive behaviors and school suspension. Concordance was higher for first graders than fourth graders, which Loeber et al. suggest would be expected since parents would be more likely to know about any misbehavior that takes place at younger ages. These findings are encouraging and suggest that self-report instruments, if administered with concern for the age of the respondents, can be used for very young children.

Self-Report Measures for Adults

Interest in assessing antisocial behavior across the life span has also led to an increasing number of longitudinal surveys that have followed respondents from their adolescent years into early adulthood (e.g., Elliott, 1994; Farrington, 1989b; Hawkins et al., 1992; Huizinga et al., 1998; LeBlanc, 1989; Loeber et al., 1998; Krohn et al., 1997). The concern in constructing self-report instruments for adults is to include items that take into account the different contexts in which crime occurs at these ages (e.g., work instead of school), the opportunities for different types of offenses (e.g., domestic violence, fraud), the inappropriateness or inapplicability of offenses that appear on adolescent self-report instruments (e.g., status offenses), and the potential for very serious criminal behaviors, at least among a small subset of chronic violent offenders.

Weitekamp (1989) has criticized self-report studies for not only being predominantly concerned with the adolescent years but also, when covering the adult years, for using the same items used for juveniles. He argues that even such studies as the NYS (Elliott, 1994) do not include many items that are more serious, and therefore appropriate for adults, than the items included in the original Short and Nye study (1957). Weitekamp asserts that different instruments need to be used during different life stages. Doing so, however, raises questions about construct continuity. If researchers want to document the change in the propensity to engage in antisocial behavior throughout the life course, it must be assumed that different items used to measure antisocial behavior at different ages do indeed measure the same underlying construct. LeBlanc (1989) suggests that a strategy of in

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

cluding different but overlapping items on instruments covering different ages across the life span is the best compromise.

There have been relatively few assessments of the validity of self-report data collected from adults. The data that are available suggest that the validity of adult self-report data is not fundamentally different from that of adolescent self-report data, however. For example, the validity data from Maxfield and his colleagues (2000) and from the DUF project presented above are from adult samples. Their estimates of validity are in the same range as those of most adolescent surveys. Elliott (1994) has presented information from the NYS that suggests adult self-report data are more congruent with adult arrests than juvenile self-report data are with juvenile arrests.

Panel or Testing Effects

Developments in self-report methods have improved the quality of data collected and have expanded their applicability to the study of antisocial behavior throughout the life course. While these advances are significant, they have increased the potential for the data to be contaminated by testing or panel effects. Testing effects are any alterations of a respondent’s response to an item or scale that is caused by the prior administration of the same item or scale (Thornberry, 1989).

Improvements in self-report instruments have led to the inclusion of a longer list of items in order to tap more serious offenses, and often a number of follow-up questions are asked. The more acts that a respondent admits to, the longer the overall interview will take. The concern is that this approach will make respondents increasingly unwilling to admit to delinquent acts because those responses will increase the overall length of the interview. This effect likely would be unequally distributed across respondents because those who had the most extensive involvement in delinquency would have the most time to lose by answering affirmatively to the delinquency items.

It is also possible that the simple fact that a respondent is reinterviewed may create a generalized fatigue and lead to decreased willingness by the respondent to respond to self-report items. Research using the National Crime Victimization Survey found that the reduction in reporting was due more to the number of prior interviews than to the number of victimizations reported in prior interviews (Lehnen and Reiss, 1978).

Three studies have examined testing effects in the use of self-report

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

studies; all are based on data from the NYS (Elliott et al., 1985). They were conducted by Thornberry (1989), Menard and Elliott (1993), and Lauritsen (1998). The NYS surveyed a nationally representative sample of 1,725 youth ages 11 to 17 in 1976. The same subjects were reinterviewed annually through 1981. These data allow researchers to examine age-specific prevalence rates by the number of times a respondent was interviewed. For example, some respondents were 14 at the time of their first interview; some were 14 at their second interview (the original 13-year-old cohort); some were 14 at their third interview (the original 12-year-old cohort); and so forth. Because of this, a 14-year-old prevalence rate can be calculated from data collected when the respondents were interviewed for only the first time, from data collected when they were interviewed a second time, etc. If a testing or panel effect plays a role in response rates, the more frequently respondents are interviewed the lower the age-specific rates should be.

Thornberry (1989) analyzed these rates for 17 NYS self-report items representing the major domains of delinquency and, for the most part, the most frequently occurring items. The overall trend suggests a panel effect. For all offenses except marijuana use, comparisons between adjacent waves indicated that age-specific prevalence rates decreased more often than they increased. For example, comparing the rate of gang fights from wave to wave, Thornberry found that for 67 percent of the comparisons there was a decrease in age-specific prevalence rates, whereas there was an increase in only 20 percent of the comparisons and in 13 percent there was no change. The magnitude of the changes was substantial in many cases. For example, for stealing something worth $5 to $50, the rate for 15-year-olds dropped by 50 percent for 15-year-olds from wave 1 to wave 4.

The NYS did not introduce detailed follow-up questions to the delinquency items until the fourth wave of data collection. The data analyzed by Thornberry show that the decline in reporting occurred across all waves. Hence, it appears that the panel design itself, rather than the design of the specific questions, had the effect of decreasing prevalence rates. The observed decline in age-specific rates could be due to an underlying secular drop in offenses during these years (1976-1981). Cross-sectional trend data from the Monitoring the Future (MTF) study, which cannot be influenced by a testing effect, do not indicate any such secular decline (see Thornberry, 1989).

Menard and Elliott (1993) reexamined this issue using both NYS and MTF data. They rightfully pointed out that comparisons between these

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

studies need to be undertaken cautiously because of differences in samples, design features, item wording, and similar concerns. Menard and Elliott’s analysis also showed that at the item level, declining trends are more evident in the NYS data than the MTF data. Most of these year-to-year changes are not statistically significant, however. Menard and Elliott then used a modified Cox-Stuart trend test to examine short-term trends in delinquency and drug use. Overall, the trends for 81 percent of the NYS offenses are not statistically significant and about half of the MTF trends are. But an examination of the trends for the 16 items included in their Table 2 indicates that there are more declining trends in the NYS data, 9 of 16 for the 1976-1980 comparisons and 7 of 16 for the 1976-1983 comparisons, than there are for the MTF data, 3 of 16 in both cases. Menard and Elliott focus on the statistically significant effects, which do indicate fewer declining trends in the NYS than is evident when one focuses on all trends, regardless of the magnitude of the change.

More recently, Lauritsen (1998) examined this topic using hierarchical linear models to estimate growth curve models for general delinquency and serious delinquency. She limited her analysis to four of the seven cohorts in the NYS, those who were 11, 13, 15, and 17 years old at wave 1. For those who were 13, 15, or 17 at the start of the NYS, involvement in both general delinquency and serious delinquency decreased significantly over the next four years. For the 11-year-old cohort, the rate of change was also negative but not statistically significant. This downward trajectory in the rate of delinquent behavior for all age cohorts is not consistent with theoretical expectations or with what is generally known about the age-crime curve. Also, as Lauritsen points out, it is not consistent with other data on secular trends for the same time period (see also Thornberry, 1989; Osgood et al., 1989).

Finally, Lauritsen examined whether this testing effect is due to the introduction of detailed follow-up questions during wave 4 of the NYS or whether it appeared to be produced by general panel fatigue. Her analysis of individual growth trajectories indicates that the decline is observed across all waves. Thus she concludes, as Thornberry did, that the reduced reporting is unlikely to have been produced by the addition of follow-up questions.

Overall, Lauritsen offers two explanations for the observed testing effects. One concerns generalized panel fatigue, suggesting that as respondents are asked the same inventory at repeated surveys they become less willing to respond affirmatively to screening questions. The second expla

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

nation concerns a maturation effect in which there is change in the content validity of the self-report questions with age. For example, how respondents interpret a question on simple assault and the type of behavior they consider relevant for responding to the question may be quite different for 11 and 17 year olds. This would not account for the drop in the age-specific rates observed by Thornberry (1989), however.

The studies by Thornberry and Lauritsen suggest that it is likely there is some degree of panel bias in self-report data collected in longitudinal panel studies. The analysis by Menard and Elliott indicates that this is indeed just a suggestion at this point, as the necessary comparisons between panel studies and cross-sectional trend studies are severely hampered by the lack of comparability in item wording, administration, and other methodological differences. Also, if there are testing effects, neither Thornberry nor Lauritsen argues that they are unique to the NYS. It just so happens that the sequential cohort design of the NYS makes it a good vehicle for examining this issue. The presumption, unfortunately, is that if testing effects interfere with the validity of the NYS data, they also interfere with the validity of other longitudinal data containing self-report information. This is obviously a serious matter because etiological research has focused almost exclusively on longitudinal designs during the past 20 years. Additional research to identify the extensiveness of testing effects, their sources, and ways to remedy them are certainly a high priority.

Validity of Self-Reports Across Developmental Stages

Earlier we reviewed the literature that assessed the criterion validity of self-report data. Almost all of those studies assess criterion validity at a single point in time. There has been little systematic investigation of validity at different ages, especially for the same subjects followed over time. Because of that, we have begun to assess this issue using the self-report and official data collected in the Rochester Youth Development Study. As in previous studies, two comparisons can be made: (1) the prevalence of self-reported arrests versus the prevalence of official arrests and (2) the prevalence of self-reported delinquency and drug use versus the prevalence of official arrests. We combine the delinquency and drug use items into one self-report inventory since youth can be, and are, arrested for this full range of illegal behaviors. We expect, of course, positive correlations across these alternative measures of involvement in crime.

Table 3-1 presents the results for the total Rochester sample at each of

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

TABLE 3-1 Yule’s Q Comparing the Prevalence of Self-Reported and Official Data, Rochester Youth Development Study, Total Panel

 

Wave

 

2

3

4

5

6

7

8

9

10

11

12

Mean

Self-reported arrestsa with official arrests (n = 834 to 940)

0.83

0.89

0.79

0.86

0.80

0.68

0.78

0.80

0.80

0.86

0.84

0.81

Self-reported delinquency/ drug use with official arrests (n = 836 to 943)

0.48

0.64

0.61

0.57

0.50

0.41

0.52

0.44

0.44

0.45

0.45

0.50

NOTE: The self-report data and the official data for waves 10, 11, and 12 are annual rates. The data for waves 2 through 9 cover six-month periods.

aThe method of asking about arrests changed during the course of the study. In waves 2 and 3, respondents were asked if they had been arrested or picked up by the police in the last six months. In waves 4 through 11, the arrest questions were presented as follow-up questions to the self-reported delinquency/drug use inventory and a global arrest question (arrested or picked up for anything else) was included at the end. In wave 12 only the global question was asked.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

11 waves of data, waves 2 through 12. This allows us to assess criterion validity from early adolescence (the average age at wave 2 is 14) to early adulthood (the average age at wave 12 is 22). The self-reported arrest measure asks respondents if they had been arrested or picked up by the police since the last interview. The self-reported delinquency data are for our general delinquency index, which includes a variety of offenses from trivial to serious. The official arrest file contains information on arrests and official warnings during the juvenile years and arrests during the adult years. Rochester city, Monroe County, and New York state files were searched. Each arrest was assigned to an interview wave.

There is a high degree of concordance between the official arrest histories and the self-reported arrest histories for the Rochester subjects. We use Yule’s Q, a standard measure of association for two-by-two contingency tables, that varies from 0 to 1 (Christensen, 1997). The average Yule’s Q is 0.81 across the 11 waves, and the range is from 0.68 to 0.89. Subjects who have an official contact or arrest were, generally speaking, willing to report that to their interviewers. There does not appear to be a strong developmental trend in the validity of these data.

The second panel in Table 3-1 presents the association between official arrests and self-reported general delinquency and drug use. If the self-report data are valid, it can be expected that subjects who report offending will be more apt to have an official record than subjects who do not. This is generally what we see, although consistent with the literature, these coefficients are somewhat lower than those in the top panel.

The average Yule’s Q across the 11 waves is 0.50, with a range between 0.41 and 0.64. Here there does seem to be a slight downward drift in the size of the relationship over time. During the first few waves, the correlations are in the 0.5 to 0.6 range, but by the last four waves they are in the 0.40 to 0.45 range. The coefficients for the early waves are similar to those reported in previous studies of adolescents (e.g., Hindelang et al., 1981).

It is not yet clear why these coefficients decline over time. The drop in the validity estimates for self-reported delinquency is consistent with a testing effect, although the major decline does not occur until the last few waves. The absence of a strong trend in the self-reported arrest data argues against a testing effect, however, since for most waves these questions were embedded in the self-report follow-up questions. An alternative explanation concerns the changing nature of criminal behavior. It is possible that offenses committed at these ages (early 20s) are less public and therefore somewhat less well correlated with arrest data.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

A major question about the validity of self-report data concerns differential levels of reporting by race/ethnicity and gender. Table 3-2 presents comparisons for male and female respondents separately. Overall, there is a somewhat higher degree of validity for the female respondents than the males. The average Yule’s Q for the comparison between official arrests and self-reported arrests is 0.74 for males and 0.84 for females. There is no evidence of a strong developmental trend for these data. For the comparison between self-reported delinquency/drug use and official arrests, the average association is 0.43 for the males and 0.52 for the females. For the male respondents, the size of the coefficients tails off somewhat at the older ages. The results for females are unstable, probably because of the low number of females who were arrested at these six-month intervals.

Table 3-3 presents the results by race/ethnicity. When attention focuses on the association between self-reported arrests and official arrests, there is no evidence of differential validity. The mean Yule’s Q for African Americans is 0.82, for Hispanics 0.80, and for whites 0.83. There are no strong developmental trends across time for any of the three groups.

The comparison between self-reported delinquency/drug use and official arrests is hampered by our inability to estimate Yule’s Q for the white subjects. At 9 of the 11 waves there are empty cells and/or expected cell frequencies of less than 5. Nevertheless, there does seem to be some evidence of differential validity across racial groups. The mean for African Americans is 0.47; for Hispanics, 0.67.

Overall, when self-reported arrests and official arrests are compared, there is little evidence of differential attrition by gender or race/ethnicity and all the coefficients are reasonably high. For the comparison between self-reported delinquency/drug use and official arrests, however, validity is lower for African Americans than Hispanics. This finding is consistent with previous research and must be taken into account when using self-report data.

Similarity of Results

In the past quarter century criminological research has increasingly relied on longitudinal studies to describe and explain patterns of criminal behavior. Much of this research, especially the descriptive studies, has used official measures of crime, but there has been growing use of self-report data, especially in etiological studies. An important but understudied topic

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

is the extent to which these two measures provide the same or different results with respect to key criminal career parameters.

Farrington and colleagues (2000) have begun to address this issue with data from the Seattle Social Development Project. Focusing on the juvenile years, ages 11 to 17, they compared results based on self-reports to those based on court referrals. There was a good deal of similarity across the methods. In particular, similar patterns were found for variations in prevalence by age, the level of continuity in commission of offenses, and the relationship between age of onset and later frequency of committing offenses. There were also some notable differences. “In self-reports, prevalence and individual offending frequency were higher, the age of onset was earlier, and the concentration of offending was greater” (Farrington et al., 2000:21). Also, there was less variation in the individual offending rate by age for the official data compared to self-reports.

While this study is a good first step to take in exploring the issue, it is not yet clear whether the glass is half empty or half full. Additional investigation is needed to identify which criminal career parameters are similar and which are different, across a variety of data sets.

Longitudinal research has demonstrated a substantial degree of continuity in offending. Past offending is related to future offending in both official and self-report data (Farrington et al., 2000). A current controversy in the criminological literature is the source of this continuity. Some argue that it is generated by static, time-stable characteristics (persistent population heterogeneity); others argue that it is generated by dynamic, time-varying processes (state dependence). A number of studies have empirically examined whether the association between past and future offending persists after stable individual differences are taken into account (Nagin and Paternoster, 1991, 2000). If it does, we assume the association is due in part to dynamic processes. Previous studies have used both self-reported data and official data, but typically not on the same individuals. Unfortunately, the results vary somewhat by type of data. Studies based on self-reports are more apt to find a state dependence effect than are studies based on official data.

To examine this issue more systematically, Brame, Bushway, Paternoster, and Thornberry (2001) used both self-report data and official data on subjects in the Rochester Youth Development Study. Separate models were estimated for violent and property offenses, for self-report and official data, and for the younger (<13) and older (>14) groups in the Rochester sample, yielding a total of eight models.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

TABLE 3-2 Yule’s Q Comparing the Prevalence of Self-Reported and Official Data, Rochester Youth Development Study, by Gender

 

Wave

 

2

3

4

5

6

7

8

9

10

11

12

Mean

Male

 

Self-Reported Arrests with Official Arrests (n = 597 to 683)

0.84

0.84

0.73

0.83

0.61

0.67

0.68

0.75

0.66

0.80

0.77

0.74

Self-Reported Delinquency/ Drug Use with Official Arrests (n = 599 to 686)

0.44

0.62

0.45

0.49

0.33

0.56

0.28

0.53

0.30

0.43

0.27

0.43

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Female

 

Self-Reported Arrests with Official Arrests (n = 234 to 257)

0.80

0.93

0.85

0.88

0.95

0.49

0.89

0.76

0.92

0.92

0.89

0.84

Self-Reported Delinquency/ Drug Use with Official Arrests (n = 234 to 257)

0.54

0.66

0.87

0.65

0.69

0.00

0.84

0.08

0.52

0.37

0.55

0.52

NOTE: The Yule’s Q values in Table 3-1 for the total sample are not simple arithmetic means of the comparable Yule’s Q values by gender because of the sample weights that are applied.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

TABLE 3-3 Yule’s Q Comparing the Prevalence of Self-Reported and Official Data, Rochester Youth Development Study, by Race/Ethnicity

 

Wave

 

2

3

4

5

6

7

8

9

10

11

12

Mean

African American

 

Self-Reported Arrests with Official Arrests (n = 324 to 383)

0.85

0.92

0.78

0.85

0.79

0.63

0.78

0.80

0.82

0.90

0.88

0.82

Self-Reported Delinquency/ Drug Use with Official Arrests (n = 326 to 385)

0.44

0.54

0.54

0.58

0.39

0.35

0.57

0.26

0.45

0.45

0.56

0.47

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Hispanic

 

Self-Reported Arrests with Official Arrests (n = 104 to 115)

0.70

0.71

0.83

0.86

0.88

0.75

0.83

0.86

0.90

0.76

0.76

0.80

Self-Reported Delinquency/ Drug Use with Official Arrests (n = 104 to 115)

0.82

0.79

0.86

0.43

0.66

0.87

0.55

0.77

0.60

0.64

0.35

0.67

Whitea

 

Self-Reported Arrests with Official Arrests (n = 158 to 174)

0.87

0.88

0.73

0.85

0.97

0.83

0.76

0.81

0.70

0.78

0.91

0.83

aThe associations between self-reported delinquency/drug use and official arrests are not reported for white subjects due to empty cells and/or expected cell frequencies of less than 5 in 9 of the 11 waves.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

In all but one case (violent offenses for the older group measured by official data) there is a positive effect of past offenses on future offenses after unobserved heterogeneity is held constant. In the exceptional case the number of arrests for violent offenses is so sparse over time that we do not think the estimate is reliable.

Overall, therefore, these results based on the same subjects suggest that self-report and official data yield the same substantive conclusion on this central issue. Both data sources indicate there are both static and dynamic processes at work that produce the observed association between past and future offenses.

CONCLUSIONS

The self-report method for measuring crime and delinquency has developed substantially since it was introduced a half century ago. It is now one of the fundamental ways to scientifically measure criminality, and it forms the bedrock of etiological studies. The challenges confronting this approach to measurement are daunting; after all, individuals are asked to tell about their own undetected criminality. Despite this fundamental challenge, the technique seems to be successful and capable of producing valid and reliable data.

Early self-report scales had substantial weaknesses, containing few items and producing an assessment of only minor forms of offending. Gradually, as the underlying validity of the approach became evident, the scales expanded in terms of breadth, seriousness, and comprehensiveness. Contemporary measures typically cover a wide portion of the behavioral domain included under the construct of crime and delinquency. These scales are able to measure serious as well as minor forms of crime, major subdomains (such as violence, property crimes, and drug use), and different parameters of criminal careers (such as prevalence, frequency, and seriousness) and identify high-rate as well as low-rate offenders. This is substantial progress for a measurement approach that began with a half dozen items and a four-category response set.

The self-report approach to measuring crime has acceptable, albeit far from perfect, reliability and validity. Of these two basic psychometric properties, the evidence for reliability is stronger. There are no fundamental challenges to the reliability of these data. Test-retest measures (and internal consistency measures) indicate that self-reported measures of delinquency are as reliable as, if not more reliable than, most social science measures.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Validity, as noted above, is much harder to assess as there is no gold standard by which to judge self-reports. Nevertheless, current scales seem to have acceptable levels of content and construct validity. The evidence for criterion validity is less clear-cut. At an overall level, criterion validity seems to be in the moderate-to-strong range. While there is certainly room for improvement, the validity appears acceptable for most analytical tasks. At a more specific level, however, there is a potentially serious problem with differential validity in that African American males have lower validity than do Hispanic males. Additional research on this topic is imperative.

While basic self-report surveys appear to be reliable and valid, researchers have experimented with a variety of data collection methods to improve the quality of reporting. Several of these attempts have produced ambiguous results; for example, there is no clear-cut benefit to the mode of administration (interview vs. questionnaire) or the use of randomized response techniques. There is one approach that appears to hold great promise— audio-assisted computerized interviews, which produce increased reporting of many sensitive topics, including delinquency and drug use. Greater use of this approach is warranted.

In the end, the available data indicate that the self-report method is an important and useful way to collect information about criminal behavior. The skepticism of early critics like Nettler (1978) and Gibbons (1979) has not been realized. Nevertheless, the self-report technique can clearly be improved. The final topic addressed in this chapter concerns suggestions for future research.

Future Directions

Much of our research on reliability and validity simply assesses these characteristics; there is much less research on improving their levels. For example, it is likely that both validity and reliability would be improved if we experimented with alternative items for measuring the same behavior and identified the strongest ones. Similarly, reliability and validity vary across subscales (e.g., Huizinga and Elliott, 1986); improving subscales will not only help them but also the overall scale as they are aggregated.

This chapter raised the issue of differential validity for African American males. It is crucial that more is learned about the magnitude of this bias and, if it exists, its source. Future research should address this issue directly and attempt to identify techniques for eliminating it. These re

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

search efforts should not lose sight of the fact that the problem may be with the criterion variable (official records) and not the self-reports.

The self-report method was developed in and for cross-sectional studies. Using it in longitudinal studies, especially ones that cover major portions of the life course, creates a new set of challenges. Maintaining the age appropriateness of the items while at the same time ensuring content validity is a knotty problem that we have just begun to address. There is some evidence that repeated measures may create testing effects. More research is needed to measure the size of this effect and its sources and to identify methods to reduce its threat to the validity of self-report data in the longitudinal studies so crucial to etiological investigation.

The similarities and differences in our understanding of criminal career parameters in self-report data and official data are just beginning to be investigated. This approach began with official data but is increasingly coming to rely on self-report data. It is important that we understand more about the validity of both types of data for these purposes.

Finally, we recommend that methodological studies be done in a cross-cutting fashion so that several of these issues—reliability and validity, improved item selection, assessing panel bias—can be investigated simultaneously. In particular it is important to examine all of these methodological issues when data are collected using audio-assisted computerized interviewing. For example, studies that have found differential validity or testing effects have all used paper-and-pencil interviews. Whether these same problems are evident under the enhanced confidentiality of audio interviews is an open question. It is clearly a high-priority one as well.

There is no dearth of work that can be done to assess and improve the self-report method. If the progress of the past half century is any guide, we are optimistic that the necessary studies will be conducted and that they will improve this basic way of collecting data on criminal behavior.

REFERENCES

Achenbach, T.M. 1992 Manual for the Child Behavior Checklist/2-3 and 1992 Profile. Burlington: University of Vermont.

Akers, R.L. 1964 Socio-economic status and delinquent behavior: A retest. Journal of Research in Crime and Delinquency 1:38-46.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Akers, R.L., M.D. Krohn, L. Lanza-Kaduce, and M. Radosevich 1979 Social learning and deviant behavior: A specific test of a general theory. American Sociological Review 44:636-655.

Akers, R.L., J. Massey, W. Clarke, and R.M. Lauer 1983 Are self-reports of adolescent deviance valid? Biochemical measures, randomized response, and the bogue pipeline in smoking behavior. Social Forces 62(September):234-251.

Anderson, L.S., T.G. Chiricos, and G.P. Waldo 1977 Formal and informal sanctions: A comparison of deterrent effects. Social Problems 25:103-112.

Aquilino, W.S. 1994 Interview mode effects in surveys of drug and alcohol use. Public Opinion Quarterly 58:210-240.

Aquilino, W.S., and L. LoSciuto 1990 Effects of interview mode on self-reported drug use. Public Opinion Quarterly 54:362-395.

Beebe, T.J., P.A. Harrison, J.A. McRae, Jr., R.E. Anderson and J.A. Fulkerson 1998 An evaluation of computer-assisted self-interviews in a school setting. Public Opinion Quarterly 62:623-632.

Belsky, J., S. Woodworth, and K. Crnic 1996 Troubled family interaction during toddlerhood. Development and Psychopathology 8:477-495.

Belson, W.A. 1968 The extent of stealing by London boys and some of its origins. Advancement of Science 25:171-184.

Blackmore, J. 1974 The relationship between self reported delinquency and official convictions amongst adolescent boys. British Journal of Criminology 14:172-176.

Blumstein, A., J. Cohen, J.A. Roth, and C.A. Visher 1986 Criminal Careers and Career Criminals. Washington, DC: National Academy Press.

Brame, R., S. Bushway, R. Paternoster, and T.P. Thornberry 2001 Temporal Linkages in Violent and Nonviolent Criminal Activity. Unpublished manuscript, University of South Carolina, Columbia.

Braukmann, C.J., K.A. Kirigin, and M.M. Wolf 1979 Social Learning and Social Control Perspectives in Group Home Delinquency Treatment Research. Paper presented to the American Society of Criminology, Philadelphia.

Broder, P.K., and J. Zimmerman 1978 Establishing the Reliability of Self-Reported Delinquency Data. Williamsburg, VA: National Center for State Courts.

Browning, K., D. Huizinga, R. Loeber, and T.P. Thornberry 1999 Causes and Correlates of Delinquency Program, Fact Sheet, Washington, DC: U.S. Department of Justice, Office of Juvenile Justice and Delinquency Prevention.


Campbell, S.B. 1987 Parent-referred problem three-year-olds: Developmental changes in symptoms. Journal of Child Psychology and Psychiatry 28:835-845.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Campbell, S.B. 1990 Behavioral Problems in Preschool Children: Clinical and Developmental Issues. New York: Guilford Press.

Christensen, R. 1997 Log-Linear Models and Logistic Regression, 2nd ed. New York: Springer-Verlag.

Clark, J.P., and L.L. Tifft 1966 Polygraph and interview validation of self-reported delinquent behavior. American Sociological Review 31:516-523.

Clark, J.P., and E.P. Wenninger 1962 Socioeconomic class and area as correlates of illegal behavior among juveniles. American Sociological Review 28:826-834.

Conger, R. 1976 Social control and social learning models of delinquency: A synthesis. Criminology 14:17-40.

Drug Use Forecasting 1990 Drug Use Forecasting Annual Report: Drugs and Crime in America. Washington, DC: U.S. Department of Justice.

Dentler, R.A., and L.J. Monroe 1961 Social correlates of early adolescent theft. American Sociological Review 26:733-743.


Elliott, D.S. 1966 Delinquency school attendance and dropout. Social Problems 13:306-318.

1994 Serious violent offenders: Onset, developmental course, and termination. Criminology 32:1-21.

Elliott, D.S., and S.S. Ageton 1980 Reconciling race and class differences in self-reported and official estimates of delinquency. American Sociological Review 45:95-110.

Elliott, D.S., and H.L. Voss 1974 Delinquency and Dropout. Lexington, MA: D.C. Heath.

Elliott, D.S., D. Huizinga, and S.S. Ageton 1985 Explaining Delinquency and Drug Use. Beverly Hills, CA: Sage.

Empey, L.T., and M. Erickson 1966 Hidden delinquency and social status. Social Forces 44(June):546-554.

Erickson, M. and L.T. Empey 1963 Court records, undetected delinquency and decision-making. Journal of Criminal Law, Criminology, and Police Science 54:456-469.


Farrington, D.P. 1973 Self-reports of deviant behavior: Predictive and stable? Journal of Criminal Law and Criminology 64:99-110.

1986 Stepping Stones to Adult Criminal Careers. In Development of Antisocial and Prosocial Behavior, D. Olweus, J. Block, and M.R. Yarrow, eds. New York: Academic Press.

1989a Early predictors of adolescent aggression and adult violence. In Violence and Victims. Washington, DC: Springer.

1989b Self-reported and official offending from adolescence to adulthood. In Cross-

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

National Research in Self-Reported Crime and Delinquency, M.W. Klein, ed. Los Angeles: Kluwer Academic Publishers.

Farrington, D.P., R. Loeber, M. Stouthamer-Loeber, W.B. Van Kammen, and L. Schmidt 1996 Self-reported delinquency and a combined delinquency seriousness scale based on boys, mothers, and teachers: Concurrent and predictive validity for African-American and Caucasians. Criminology 34:493-517.

Farrington, D.P., D. Jolliffe, J.D. Hawkins, R.F. Catalano, K.G. Hill, and R. Kosterman 2000 Comparing Delinquency Careers in Court Records and Self-reports. Unpublished manuscript, Cambridge University, United Kingdom.

Gibbons, D.C. 1979 The Criminological Enterprise: Theories and Perspectives. Englewood Cliffs, NJ: Prentice-Hall.

Gold, M. 1966 Undetected delinquent behavior. Journal of Research in Crime and Delinquency 3:27-46.

1970 Delinquent Behavior in an American City. Belmont, CA: Brooks/Cole.

Gottfredson, M.R., and T. Hirschi 1990 A General Theory of Crime. Stanford, CA: Stanford University Press.


Hardt, R.H., and S. Petersen-Hardt 1977 On determining the quality of the delinquency self-report method. Journal of Research in Crime and Delinquency 14:247-261.

Hathaway, R.S., E.D. Monachesi, and L.A. Young 1960 Delinquency rates and personality. Journal of Criminal Law, Criminology, and Police Science 50:433-440.

Hawkins, J.D., R.F. Catalano, and J.Y. Miller 1992 Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin 112:64-105.

Hepburn, J.R. 1976 Testing alternative models of delinquency causation. Journal of Criminal Law and Criminology 67:450-460.

Hindelang, M.J. 1973 Causes of delinquency: A partial replication and extension. Social Problems 20: 471-487.

Hindelang, M.J., T. Hirschi, and J.G. Weis 1979 Correlates of delinquency: The illusion of discrepancy between self-report and official measures. American Sociological Review 44:995-1014.

1981 Measuring Delinquency. Beverly Hills, CA: Sage.

Hirschi, T. 1969 Causes of Delinquency. Berkeley: University of California Press.

Huesmann, L.R., L.D. Eron, M.M. Lefkowitz, and L.O. Walder 1984 The stability of aggression over time and generations. Developmental Psychology 20:1120-1134.

Huizinga, D., and D.S. Elliott 1983 A Preliminary Examination of the Reliability and Validity of the National Youth

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Survey Self-Reported Delinquency Indices. National Youth Survey Project Report 27. Boulder, CO: Behavioral Research Institute.

1986 Reassessing the reliability and validity of self-report delinquent measures. Journal of Quantitative Criminology 2:293-327.

Huizinga, D., A.W. Weiher, S. Menard, R. Espiritu, and F.A. Esbensen 1998 Some not so boring findings from the Denver Youth Survey. Paper presented at the annual meeting of American Society of Criminology, Washington, D.C.

Jensen, G.F. 1973 Inner containment and delinquency. Journal of Criminal Law and Criminology 64:464-470.

Jensen, G.F., and R. Eve 1976 Sex differences in delinquency. Criminology 13:427-448.

Jensen, G.F., M.L. Erickson, and J.P. Gibbs 1978 Perceived risk of punishment and self-reported delinquency. Social Forces 57:57-78.

Jessor, R. 1998 New Perspectives on Adolescent Risk Behavior. New York: Cambridge University Press.

Jessor, R., J.E. Donovan, and F.M. Costa 1991 Beyond Adolescence: Problem Behavior and Young Adult Development. Cambridge, England: Cambridge University Press.

Johnson, R.E. 1979 Juvenile Delinquency and Its Origins. Cambridge, England: Cambridge University Press.

Johnston, L.D., P.M. O’Malley, and J.G. Bachman 1996 National Survey Results on Drug Use from the Monitoring the Future Study, 1975-1995. Washington, DC: U.S. Government Printing Office.


Kaplan, H.B. 1972 Toward a general theory of psychosocial deviance: The case of aggressive behavior. Social Science and Medicine 6:593-617.

Kelly, D.H. 1974 Track position and delinquent involvement: A preliminary analysis. Sociology and Social Research 58:380-386.

Klein, M.W., ed. 1989 Cross-National Research in Self-Reported Crime and Delinquency. Los Angeles: Kluwer Academic Publishers.

Krohn, M.D., G.P. Waldo, and T.G. Chiricos 1974 Reported delinquency: A comparison of structured interviews and self-administered check-lists. Journal of Criminal Law and Criminology 65:545-553.

Krohn, M.D., A.J. Lizotte, and C.M. Perez 1997 The interrelationship between substance use and precocious transitions to adult statutes. Journal of Health and Social Behavior 38:87-103.

Krohn, M.D., T.P. Thornberry, C. Rivera, and M. LeBlanc 2001 Later delinquency careers. In Child Delinquents: Development, Intervention, and Service Needs, R. Loeber and D.P. Farrington, eds. Thousand Oaks, CA: Sage.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Kulik, J.A., K.B. Stein, and T.R. Sarbin 1968 Disclosure of delinquent behavior under conditions of anonymity and nonanonymity. Journal of Consulting and Clinical Psychology 32:506-509.

Lauritsen, J.L. 1998 The age-crime debate: Assessing the limits of longitudinal self-report data. Social Forces 76:1-29.

LeBlanc, M. 1989 Designing a self-report instrument for the study of the development of offending from childhood to adulthood: Issues and problems. Pp. 371-398 in Cross-National Research in Self-Reported Crime and Delinquency, M.W. Klein, ed. Los Angeles: Kluwer Academic Publishers.

Lehnen, R.G., and A.J. Reiss 1978 Response effects in the National Crime Survey. Victimology: An International Journal 3:110-124.

Loeber, R., M. Stouthamer-Loeber, W.B. Van Kammen, and D.P. Farrington 1989 Development of a new measure of self-reported antisocial behavior for young children: Prevalence and reliability. Pp. 203-225 in Cross-National Research in Self-Reported Crime and Delinquency, M.W. Klein, ed. Los Angeles: Kluwer Academic Publishers.

Loeber, R., P. Wung, K. Keenan, B. Giroux, and M. Stouthamer-Loeber 1993 Developmental pathways in disruptive child behavior. Development and Psychopathology 5:101-133.

Loeber, R., D.P. Farrington, M. Stouthamer-Loeber, T.E. Moffitt, and A. Caspi 1998 The development of male offending: Key findings from the first decade of the Pittsburgh Youth Study. Studies on Crime and Crime Prevention 7:1-31.


Matthews, V.M. 1968 Differential identification: An empirical note. Social Problems 14:376-383.

Maxfield, M.G., B.L. Weiler, and C.S. Widom 2000 Comparing self-reports and official records of arrests. Journal of Quantitative Criminology 16:87-100.

Menard, S. and D.S. Elliott 1993 Data set comparability and short-term trends in crime and delinquency. Journal of Criminal Justice 21:433-445.

Moffitt, T.E. 1993 Life-course-persistent and adolescence-limited antisocial behavior: A developmental taxonomy. Psychological Review 100:674-701.

1997 Adolescence-limited and life-course-persistent offending: A complementary pair of developmental theories. Pp. 11-54 in Developmental Theories of Crime and Delinquency, Volume 7: Advances in Criminological Theory, T.P. Thornberry, ed. New Brunswick, NJ: Transaction Publishers.


Nagin, D.S., and R. Paternoster 1991 On the relationship of past and future participation in delinquency. Criminology 29:163-190.

2000 Population heterogeneity and state dependence: State of the evidence and directions for future research. Journal of Quantitative Criminology 16:117-145.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Nettler, G. 1978 Explaining Crime. New York: McGraw-Hill.

Nye, F.I., J.F. Short, and V.J. Olson 1958 Socioeconomic status and delinquent behavior. American Journal of Sociology 63:381-389.

Olweus, D. 1979 Stability and aggressive reaction patterns in males: A review. Psychological Bulletin 86:852-875.

Osgood, D.W., P. O’Malley, J. Bachman, and L. Johnston 1989 Time trends and age trends in arrests and self-reported illegal behavior. Criminology 27:389-418.


Patterson, G.R. 1993 Orderly change in a stable world: The antisocial trait as a chimera. Journal of Consulting and Clinical Psychology 61:911-919.

Patterson, G.R., and R. Loeber 1982 The Understanding and Prediction of Delinquent Child Behavior. Research proposal to National Institute of Mental Health from the Oregon Social Learning Center, Eugene.

Polk, K. 1969 Class strain and rebellion among adolescents. Social Problems 17:214-224.

Porterfield, A.L. 1943 Delinquency and outcome in court and college. American Journal of Sociology 49:199-208.

1946 Youth in Trouble. Fort Worth, TX: Leo Potishman Foundation.


Reiss, A.J., Jr., and A.L. Rhodes 1959 A Socio-Psychological Study of Adolescent Conformity and Deviation. Washington, DC: U.S. Office of Education.

1963 Status deprivation and delinquent behavior. Sociological Quarterly 4:135-149.

1964 An empirical test of differential association theory. Journal of Research in Crime and Delinquency 1:5-18.

Richman, N., J. Stevenson, and P.J. Graham 1982 Preschool to School: A Behavioural Study. London: Academic Press.

Rojek, D.G. 1983 Social status and delinquency: Do self-reports and official reports match? In Measurement Issues in Criminal Justice, G.P. Waldo, ed. Beverly Hills, CA: Sage.


Sampson, R.J., and J.H. Laub 1990 Crime and deviance over the life course: The salience of adult social bonds. American Sociological Review 55:609-627.

Saris, W.E. 1991 Computer-Assisted Interviewing. Beverly Hills, CA: Sage.

Sellin, T. 1931 The basis of a crime index. Journal of Criminal Law and Criminology 22:335-356.

Shaw, D.S., and R.Q. Bell 1993 Developmental theories of parental contributors to antisocial behavior. Journal of Abnormal Child Psychology 21:35-49.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Short, J.F. 1957 Differential association and delinquency. Social Problems 4:233-239.

Short, J.F., Jr., and F.I. Nye 1957 Reported behavior as a criterion of deviant behavior. Social Problems 5:207-213.

1958 Extent of unrecorded juvenile delinquency: Tentative conclusions. Journal of Criminal Law and Criminology 49:296-302.

Silberman, M. 1976 Toward a theory of criminal deterrence. American Sociological Review 41:442-461. Slocum, W.L., and C.L. Stone

1963 Family culture patterns and delinquent-type behavior. Marriage and Family Living 25:202-208.

Skolnick, J.V., C.J. Braukmann, M.M. Bedlington, K.A. Kirigin, and M.M. Wolf 1981 Parent-youth interaction and delinquency in group homes. Journal of Abnormal Child Psychology 9:107-119.

Stanfield, R. 1966 The interaction of family variables and gang variables in the aetiology of delinquency. Social Problems 13:411-417.

Thornberry, T.P. 1987 Toward an interactional theory of delinquency. Criminology 25:863-891.

1989 Panel effects and the use of self-reported measures of delinquency in longitudinal studies. Pp. 347-369 in Cross-National Research in Self-Reported Crime and Delinquency, M.W. Klein, ed. Los Angeles: Kluwer Academic Publishers.

Thornberry, T.P., ed. 1997 Developmental Theories of Crime and Delinquency. New Brunswick, NJ: Transaction Publishers.

Thornberry, T.P., and M.D. Krohn 2001 The development of delinquency: An interactional perspective. Pp. 289-305 in Handbook of Law and Social Science: Youth and Justice, S.O. White, ed. New York: Plenum.

Thornberry, T.P., M.D. Krohn, A.J. Lizotte, C.A. Smith, and K. Tobin In Press The Toll of Gang Membership: Gangs and Delinquency in Developmental Perspective. New York: Cambridge University Press.

Thrasher, F. 1927 The Gang: A Study of 1,313 Gangs in Chicago. Chicago: University of Chicago Press.

Tourangeau, R. and T.W. Smith 1996 Asking sensitive questions: The impact of data collection, mode, question format, and question context. Public Opinion Quarterly 60:275-304.

Tracy, P.E., and J.A. Fox 1981 The validity of randomized response for sensitive measurements. American Sociological Review 46:187-200.

Tremblay, R.E., C. Japel, D. Perusse, P. McDuff, M. Boivin, M. Zoccolillo, and J. Montplaisir 1999 The search for the age of ‘onset’ of physical aggression: Rousseau and Bandura revisited. Criminal Behaviour and Mental Health 9:8-23.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×

Turner, C.F., J.T. Lessler, and J. Devore 1992 Effects of mode of administration and wording on reporting of drug use. Pp. 177-220 in Survey Measurement of Drug Use: Methodological Studies, C.F. Turner, J.T. Lessler, and J.C. Gfroerer, eds. Washington, DC: U.S. Department of Health and Human Services.

Vaz, E.W. 1966 Self-reported juvenile delinquency and social status. Canadian Journal of Corrections 8:20-27.

Voss, H.L. 1963 Ethnic differentials in delinquency in Honolulu. Journal of Criminal Law and Criminology 54:322-327.

1964 Differential association and reported delinquent behavior: A replication. Social Problems 12:78-85.

1966 Socio-economic status and reported delinquent behavior. Social Problems 13:314-324.


Waldo, G.P., and T.G. Chiricos 1972 Perceived penal sanction and self-reported criminality: A neglected approach to deterrence research. Social Problems 19:522-540.

Wallerstein, J.S., and C.J. Wylie 1947 Our law-abiding law-breakers. Probation 25:107-112.

Weis, J.G., and D.V. Van Alstyne 1979 The Measurement of Delinquency by the Randomized Response Method. Paper presented at the meeting of the American Society of Criminology, Philadelphia.

Weitekamp, E. 1989 Some problems with the use of self-reports in longitudinal research. Pp. 329-346 in Cross-National Research in Self-Reported Crime and Delinquency, M.W. Klein, ed. Los Angeles: Kluwer Academic Publishers.

Widom, C.S. 1989 Child abuse, neglect, and violent criminal behavior. Criminology 27:251-271.

Williams, J.R., and M. Gold 1972 From delinquent behavior to official delinquency. Social Problems 20(2):209-229.

Wolfgang, M.E., R.M. Figlio, and T. Sellin 1972 Delinquency in a Birth Cohort.Chicago: University of Chicago Press.

Wolfgang, M.E., T.P. Thornberry, and R.M. Figlio 1987 From Boy to Man, From Delinquency to Crime.Chicago: University of Chicago Press.

Wright, D.L., W.S. Aquilino and A.J. Supple 1998 A comparison of computer-assisted and paper-and-pencil self-administered questionnaires in a survey on smoking, alcohol, and drug use. Public Opinion Quarterly 62:331-353.

Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 43
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 44
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 45
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 46
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 47
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 48
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 49
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 50
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 51
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 52
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 53
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 54
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 55
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 56
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 57
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 58
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 59
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 60
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 61
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 62
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 63
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 64
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 65
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 66
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 67
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 68
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 69
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 70
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 71
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 72
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 73
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 74
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 75
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 76
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 77
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 78
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 79
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 80
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 81
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 82
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 83
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 84
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 85
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 86
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 87
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 88
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 89
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 90
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 91
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 92
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 93
Suggested Citation:"3. Comparison of Self-Report and Official Data for Measuring Crime." National Research Council. 2003. Measurement Problems in Criminal Justice Research: Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/10581.
×
Page 94
Next: Appendix A: Workshop Agenda »
Measurement Problems in Criminal Justice Research: Workshop Summary Get This Book
×
Buy Paperback | $45.00 Buy Ebook | $35.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Most major crime in this country emanates from two major data sources. The FBI’s Uniform Crime Reports has collected information on crimes known to the police and arrests from local and state jurisdictions throughout the country. The National Crime Victimization Survey, a general population survey designed to cover the extent, nature, and consequences of criminal victimization, has been conducted annually since the early1970s. This workshop was designed to consider similarities and differences in the methodological problems encountered by the survey and criminal justice research communities and what might be the best focus for the research community. In addition to comparing and contrasting the methodological issues associated with self-report surveys and official records, the workshop explored methods for obtaining accurate self-reports on sensitive questions about crime events, estimating crime and victimization in rural counties and townships and developing unbiased prevalence and incidence rates for rate events among population subgroups.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!