Schaffer Online Library of Drug Policy Sign the Resolution
Contents | Feedback | Search
DRCNet Home
| Join DRCNet
DRCNet Library | Schaffer Library | Government Publications


Sourcebook of Criminal Justice Statistics

Appendix 9

National Crime Victimization Survey

Survey methodology and definitions of terms

Note: This information was excerpted from U.S. Department of Justice, Bureau of Justice Statistics, Criminal Victimization in the United States, 1995, NCJ-171129 (Washington, DC: U.S. Department of Justice, 1998); and Kathryn A. Chandler, et al., Students' Reports of School Crime: 1989 and 1995, NCES 98-241/NCJ-169607 (Washington, DC: U.S. Departments of Education and Justice, 1998), pp. 1, 2, 22, 23. Non-substantive editorial adaptations have been made.

Survey methodology

The National Crime Victimization Survey (NCVS) collects data from residents living throughout the United States, including persons living in group quarters, such as dormitories, rooming houses, and religious group dwellings. Crew members of merchant vessels, Armed Forces personnel living in military barracks, and institutionalized persons, such as correctional facility inmates, were not included in the survey. Similarly, U.S. citizens residing abroad and foreign visitors to this country were excluded. With these exceptions, individuals age 12 and older living in units selected for the sample were eligible to be interviewed.

Data collection

Each housing unit selected for the NCVS remains in the sample for 3 years, with each of seven interviews taking place at 6-month intervals. An NCVS interviewer's first contact with a housing unit selected for the survey is in person. The interviewer may then conduct subsequent visits, except for the fifth, by telephone.

To elicit more accurate reporting of incidents, NCVS uses the self-respondent method which calls for the direct interviewing of each person 12 years and older in the household. An exception is made to use proxy interviewing instead of direct interviewing for the following three cases: 12- and 13-year-old persons when a knowledgeable household member insists they not be interviewed directly, incapacitated persons, and individuals absent from the household during the entire field-interviewing period. In the case of temporarily absent household members and persons who are physically or mentally incapable of granting interviews, interviewers may accept other household members as proxy respondents, and in certain situations non-household members may provide information for incapacitated persons.

Approximately 30% of the interviews in the 1995 sample were conducted using Computer-Assisted Telephone Interviewing (CATI), a data collection mode that involves interviewing from centralized facilities and using a computerized instrument. In the CATI-eligible part of the sample, all interviews are done by telephone whenever possible, except for the first and fifth interviews, which are still primarily conducted in person. The telephone interviews are conducted by the CATI facilities (Hagerstown, MD and Tucson, AZ).

Sample design and size

Survey estimates are derived from a stratified, multi-stage cluster sample. The primary sampling units (PSUs) composing the first stage of the sample were counties, groups of counties, or large metropolitan areas. Large PSUs were included in the sample automatically and are considered to be self-representing (SR) since all of them were selected. The remaining PSUs, called non-self-representing (NSR) because only a subset of them was selected, were combined into strata by grouping PSUs with similar geographic and demographic characteristics, as determined by the 1990 census.

The 1995 NCVS sample households were drawn from both the 1980- and 1990-based sample designs. The 1980 design consists of 84 SR PSUs and 153 NSR strata, with one PSU per stratum selected with probability proportionate to population size. The 1990 design consists of 92 SR PSUs and 153 NSR strata, with one PSU per stratum selected with probability proportionate to population size. The NCVS sample design continued use of both the 1980- and 1990-based samples through 1997. Beginning in 1998 only the 1990-based sample remains.

In the second stage of sampling, each selected stratification PSU is divided into four frames (unit, area, permit, and group quarter) from which NCVS independently selects its sample. From each selected stratification PSU, clusters of approximately four housing units or housing unit equivalents are selected from each frame. For the unit and group quarter frames, addresses come from the 1990 census files. For the permit frame, addresses come from building permit data obtained from building permit offices. For the area frame, sample blocks come from the 1990 census files. Then, addresses are listed and sampled in the field.

Approximately 58,520 housing units and other living quarters were designated for the sample. In order to conduct field interviews, the sample is divided into six groups, or rotations, and each group of households is interviewed once every 6 months over a period of 3 years. The initial interview is used to bound the interviews (bounding establishes a timeframe to avoid duplication of crimes on subsequent interviews), but is not used to compute the annual estimates. Each rotation group is further divided into six panels. A different panel of households, corresponding to one sixth of each rotation group, is interviewed each month during the 6-month period. Because the survey is continuous, newly constructed housing units are selected as described, and assigned to rotation groups and panels for subsequent incorporation into the sample. A new rotation group enters the sample every 6 months, replacing a group phased out after being in the sample for 3 years.

For these 58,520 sample households, complete interviews were obtained for approximately 47,750 households (95.1% of eligible housing units). Within interviewed households approximately 89,900 persons (91.1%) provided responses. Of the remaining 10,770 housing units, 8,010 were determined to be ineligible (i.e., vacant, demolished, etc.), and the occupants could not be reached or refused to participate in approximately 2,660 of the units.

Selection of cases for CATI

About 30% of the 47,750 households obtained in the 1995 sample were interviewed using the CATI technique. Currently, the NCVS sample PSUs fall into three groups of CATI usage: maximum-CATI PSUs, where all the segments in the PSU are CATI-eligible; half-CATI PSUs, where half of the segments in the PSU are randomly designated to be CATI-eligible; and no-CATI PSUs, where none of the segments are CATI-eligible. The level of CATI usage for each PSU was established with concern toward an optimal workload for the field interviewers. In the "half-CATI" PSUs, a random sample of about 50% of the segments in each PSU is taken and designated as CATI-eligible. The sample cases in CATI-eligible segments from the max-CATI and the half-CATI PSUs are interviewed from CATI facilities while the other sample cases are interviewed by the standard NCVS field procedures.

Estimation procedure

Annual estimates of the levels and rates of victimization are derived by accumulating six quarterly estimates, which in turn are obtained from 17 months of field interviewing, ranging from February of one year through June of the following year. The population and household figures shown on victimization rate tables are based on an average for these 17 months, centering on the ninth month of the data collection period, in this case October 1995.

Sample data from 8 months of field interviewing are required to produce estimates for each quarter. (Quarterly estimates are not published since there may not be sufficient observations to ensure their reliability.) For example, data collected between February and September are required to estimate the first quarter of any given calendar year. Each quarterly estimate is composed of equal numbers of field observations from the months during the half-year interval prior to the time of interview. Therefore, incidents occurring in January may be reported in a February interview (1 month between the crime and the interview), in a March interview (2 months), and so on up to 6 months ago for interviews conducted in July. This arrangement minimizes expected biases associated with the tendency of respondents to place victimizations in more recent months of a 6-month reference period rather than the month in which they actually occurred.

The estimation procedure begins with the application of a base weight to the data from each individual interviewed. The base weight is the reciprocal of the probability of each unit's selection for the sample, and provides a rough measure of the population represented by each person in the sample. Next, an adjustment is made to account for households and individuals in occupied units who were selected for the survey but unavailable for interview.

In addition to adjusting for unequal probabilities of selection and observation, the final weight also includes a ratio adjustment to known population totals based on the adjusted counts from the 1990 Decennial Census. Readers interested in a detailed discussion of the estimation and weighting procedures should consult the original source.

Series victimizations

A series victimization is defined as six or more similar but separate crimes that the victim is unable to recall individually or describe in detail to an interviewer. These series crimes have been excluded from the tables because victims were unable to provide details for each separate event.

Accuracy of estimates

The accuracy of an estimate is a measure of its total error, that is, the sum of all the errors affecting the estimate: sampling error as well as nonsampling error.

The sample used for the NCVS is one of a large number of possible samples of equal size that could have been obtained by using the same sample design and selection procedures. Estimates derived from different samples would differ from one another due to sampling variability, or sampling error.

The standard error of a survey estimate is a measure of the variation among the estimates from all possible samples. Therefore, it is a measure of the precision (reliability) with which a particular estimate approximates the average result of all possible samples. The estimate and its associated standard error may be used to construct a confidence interval. A confidence interval is a range of numbers which has a specified probability that the average of all possible samples, which is the true unknown value of interest in an unbiased design, is contained within the interval. About 68% of the time, the survey estimate will differ from the true average by less than one standard error. Only 10% of the time will the difference be more than 1.6 standard errors, and just 1 time in 100 will it be greater than 2.5 standard errors. A 95% confidence interval is the estimate plus or minus twice the standard error. Thus there is a 95% chance that the result of a complete census would fall within the confidence interval.

In addition to sampling error, the estimates are subject to nonsampling error. While substantial care is taken in the NCVS to reduce the sources of nonsampling error throughout all the survey operations, by means of a quality assurance program, quality controls, operational controls, and error-correcting procedures, an unquantified amount of nonsampling error remains.

Major sources of nonsampling error are related to the ability of the respondents to recall in detail the crimes that occurred during the 6 months prior to the interview. Research based on interviews of victims obtained from police files indicates that assault is recalled with the least accuracy of any crime measured by the NCVS. This may be related to the tendency of victims to not report crimes committed by offenders who are not strangers, especially if they are relatives. In addition, among certain groups, crimes that contain elements of assault could be a part of everyday life, and are therefore forgotten or not considered important enough to mention to a survey interviewer. These recall problems may result in an understatement of the actual rate of assault.

However, as part of the 1992 redesign of the survey, substantial improvements were made to measure crime more accurately and, therefore, reduce the nonsampling error. The NCVS now includes improved questions and cues that aid victims in recalling victimizations, more explicit questions are now asked about sexual victimizations, and new components have been added to measure victimizations by nonstrangers. As a result, victims are reporting more crime incidents.

Another source of nonsampling error is the inability of some respondents to recall the exact month a crime occurred, even though it was placed in the correct reference period. This error source is partially offset by interviewing monthly and using the estimation procedure described earlier. Telescoping is another problem in which incidents that occurred before the reference period are placed within the period. The effect of telescoping is minimized by using the bounding procedure previously described. The interviewer is provided with a summary of the incidents reported in the preceding interview and, if a similar incident is reported, it can be determined whether or not it is a new one by discussing it with the victim. Events that occurred after the reference period are set aside for inclusion with the data from the following interview.

Other sources of nonsampling error can result from other types of response mistakes, including errors in reporting incidents as crimes, misclassification of crimes, systematic data errors introduced by the interviewer, errors made in coding and processing the data. Quality control and editing procedures were used to minimize the number of errors made by the respondents and the interviewers.

Since field representatives conducting the interviews usually reside in the area in which they interview, the race and ethnicity of the field representatives generally matches that of the local population. Special efforts are made to further match field representatives and the people they interview in areas where English is not commonly spoken. About 90% of all NCVS field representatives are female.

Standard errors measure only those nonsampling errors arising from transient factors affecting individual responses completely at random (simple response variance); they do not reveal any systematic biases in the data. As calculated in the NCVS, the standard errors would partially measure nonsampling error arising from some of the above sources, such as transient memory errors, or accidental errors in recording or coding answers, for example.

Definitions of terms

Age--The appropriate age category is determined by the respondent's age on the last day of the month before the interview.

Aggravated assault--Attack or attempted attack with a weapon, regardless of whether an injury occurred, and attack without a weapon when serious injury results.

Annual family income--The total income of the household head and all members of the household for the 12 months preceding the interview. Includes wages, salaries, net income from businesses or farms, pensions, interest, dividends, rent, and any other form of monetary income.

Assault--An unlawful physical attack or threat of attack. Assaults may be classified as aggravated or simple. Rape, attempted rape, and sexual assaults are excluded from this category, as well as robbery and attempted robbery. The severity of assaults ranges from minor threat to incidents which are nearly fatal.

Ethnicity--A classification based on Hispanic culture and origin, regardless of race.

Head of household--A classification that defines one and only one person in each housing unit as the head. Head of household implies that the person rents or owns (or is in the process of buying), the household unit. The head of household must be at least 18, unless all members of the household are under 18, or the head is married to someone 18 or older.

Hispanic--Persons who describe themselves as Mexican-American, Chicano, Mexican, Mexicano, Puerto Rican, Cuban, Central American, South American, or from some other Spanish culture or origin, regardless of race.

Household--A person or group of people meeting either of the following criteria: (1) people whose usual place of residence is the same housing unit, even if they are temporarily absent; (2) people staying in a housing unit who have no usual place of residence elsewhere.

Household burglary--Unlawful or forcible entry or attempted entry of a residence. This crime usually, but not always, involves theft. The illegal entry may be by force, such as breaking a window or slashing a screen, or may be without force by entering through an unlocked door or an open window. If the person entering has no legal right to be present in the structure a burglary has occurred. The structure need not be the house itself for a burglary to take place; illegal entry of a garage, shed, or any other structure on the premises also constitutes household burglary. If breaking and entering occurs in a hotel or vacation residence, it is still classified as a burglary for the household whose member or members were staying there at the time the entry occurred.

Incident--A specific criminal act involving one or more victims and offenders. For example, if two people are robbed at the same time and place, this is classified as two robbery victimizations but only one robbery incident.

Marital status--Every person is assigned to one of the following classifications: (1) married, which includes persons in common-law unions and those who are currently living apart for reasons other than marital discord (employment, military service, etc.); (2) separated or divorced, which includes married persons who are legally separated and those who are not living together because of marital discord; (3) widowed; and (4) never married, which includes persons whose marriages have been annulled and those who are living together and not in a common-law union.

Metropolitan Statistical Area (MSA)--Office of Management and Budget defines this as a population nucleus of 50,000 or more, generally consisting of a city and its immediate suburbs, along with adjacent communities having a high degree of economic and social integration with the nucleus. MSA's are designated by counties, the smallest geographic units for which a wide range of statistical data can be obtained. However, in New England, MSA's are designated by cities and towns since these subcounty units are of great local significance and considerable data is available for them. Currently, an area is defined as an MSA if it meets one of two standards: (1) a city has a population of at least 50,000; (2) the Census Bureau defines an urbanized area of at least 50,000 people with a total metropolitan population of at least 100,000 (or 75,000 in New England). The Census Bureau's definition of urbanized areas, data on commuting to work, and the strength of the economic and social ties between the surrounding counties and the central city determine which counties not containing a main city are included in an MSA. For New England, MSA's are determined by a core area and related cities and towns, not counties. A metropolitan statistical area may contain more than one city of 50,000 and may cross State lines. Within this general classification unit, there are three subclassifications: urban, suburban, and rural. They are defined as follows:

Motor vehicle--An automobile, truck, motorcycle, or any other motorized vehicle legally allowed on public roads and highways.

Motor vehicle theft--Stealing or unauthorized taking of a motor vehicle, including attempted thefts.

Non-Hispanic--Persons who report their culture or origin as something other than "Hispanic" as defined above. This distinction is made regardless of race.

Nonstranger--A classification of a crime victim's relationship to the offender. An offender who is either related to, well known to, or casually acquainted with the victim is a nonstranger. For crimes with more than one offender, if any of the offenders are nonstrangers, then the group of offenders as a whole is classified as nonstranger. This category only applies to crimes that involve contact between the victim and the offender; the distinction is not made for crimes of theft since victims of this offense rarely see the offenders.

Offender--The perpetrator of a crime; this term usually applies to crimes involving contact between the victim and the offender.

Offense--A crime. When referring to personal crimes, the term can be used to refer to both victimizations and incidents.

Personal crimes--Rape, sexual assault, personal robbery, assault, purse snatching and pocket picking. Includes both attempted and completed crimes.

Personal crimes of violence--Rape, sexual assault, personal robbery, or assault. Includes both attempted and completed crimes; does not include purse snatching and pocket picking. Murder is not measured by the NCVS because of the inability to question the victim.

Property crimes--Burglary, motor vehicle theft, or theft. Includes both attempted and completed crimes.

Purse snatching/pocket picking--Theft or attempted theft of property or cash directly from the victim by stealth, without force or threat of force.

Race--Racial categories for this survey are white, black, and other. The category "other" is composed mainly of Asians, Pacific Islanders, American Indians, Aleuts, and Eskimos. The race of the head of household is used in determining the race of the household for computing household crime demographics.

Rape--Forced sexual intercourse including both psychological coercion as well as physical force. Forced sexual intercourse means vaginal, anal, or oral penetration by the offender(s). This category also includes incidents involving penetration using a foreign object such as a bottle. Includes attempted rapes, male as well as female victims, and both heterosexual and homosexual rape. Attempted rape includes verbal threats of rape.

Rate of victimization--See "Victimization rate."

Robbery--Completed or attempted theft, directly from a person, of property or cash by force or threat of force, with or without a weapon, and with or without injury.

Sexual assault--A wide range of victimizations, separate from rape or attempted rape. Includes attacks or attempted attacks generally involving unwanted sexual contact between victim and offender. Sexual assaults may or may not involve force and include such things as grabbing or fondling. Sexual assault also includes verbal threats.

Simple assault--Attack without a weapon resulting either in no injury, minor injury (for example, bruises, black eyes, cuts, scratches, or swelling), or in undetermined injury requiring less than 2 days of hospitalization. Also includes attempted assault without a weapon.

Stranger--A classification of the victim's relationship to the offender for crimes involving direct contact between the two. Incidents are classified as involving strangers if the victim identifies the offender as a stranger, did not see or recognize the offender, or knew the offender only by sight. Crimes involving multiple offenders are classified as involving nonstrangers if any of the offenders was a nonstranger. Since victims of theft without contact rarely see the offender, no distinction is made between strangers and nonstrangers for this crime.

Tenure--The NCVS recognizes two forms of household tenancy: (1) owned, which includes dwellings that are mortgaged, and (2) rented, which includes rent-free quarters belonging to a party other than the occupants, and situations where rental payments are in kind or in services.

Theft--Completed or attempted theft of property or cash without personal contact. Incidents involving theft of property from within the sample household would classify as theft if the offender has a legal right to be in the house (such as a maid, delivery person, or guest). If the offender has no legal right to be in the house, the incident would classify as a burglary.

Victim--The recipient of a criminal act, usually used in relation to personal crimes, but also applicable to households.

Victimization--A crime as it affects one individual person or household. For personal crimes, the number of victimizations is equal to the number of victims involved. The number of victimizations may be greater than the number of incidents because more than one person may be victimized during an incident. Each crime against a household is assumed to involve a single victim, the affected household.

Victimization rate--A measure of the occurrence of victimizations among a specified population group. For personal crimes, this is based on the number of victimizations per 1,000 residents age 12 and older. For household crimes, the victimization rates are calculated using the number of incidents per 1,000 households.

Victimize--To commit a crime against a person or household.

School Crime Supplement

The School Crime Supplement (SCS) was jointly designed by the U.S. Department of Education's National Center for Education Statistics and the U.S. Department of Justice's Bureau of Justice Statistics (BJS). The SCS was created as an occasional supplement to the annual NCVS and was fielded in 1989 and 1995. NCVS interviews were conducted first with each household member 12 years of age or older. Once all NCVS interviews were completed, household members between ages of 12 and 19 were given an SCS interview. Only those 12- to 19-year-olds who were in primary or secondary education programs leading to a high school diploma, and who had been enrolled sometime during the 6 months prior to the interview, were administered the SCS questionnaire. Students who were home schooled were not included.

The SCS questionnaire was designed to record the incidence of crime and criminal activity occurring inside a school, on school grounds, or on a school bus during the 6 months preceding the interview. There were 10,449 SCS interviews completed in 1989 and 9,954 in 1995.

Data were collected by the U.S. Bureau of the Census. In both 1989 and 1995, SCS surveys were conducted between January and June, with one-sixth of the sample being covered each month. Interviews were conducted with the subject student over the telephone or in person. In both years, efforts were made to assure that interviews about student experiences at school were conducted with the students themselves. However, under certain circumstances, interviews with proxy respondents were accepted. These circumstances included interviews scheduled with a child between the ages of 12 and 13 where the parents refused to allow an interview with the child, interviews where the subject child was unavailable during the period of data collection, and interviews where the child was physically or emotionally unable to answer for him or herself.

Responses to both the NCVS and SCS are confidential by law. Interviewers are instructed to conduct interviews in privacy unless respondents specifically agree to permit others to be present. Most interviews for the NCVS and SCS are conducted by telephone, and most questions require "yes" or "no" answers, thereby affording respondents a further measure of privacy.

Unit response rates indicate how many sampled units have completed interviews. Because interviews with students could only be completed after households had responded to the NCVS, the unit completion rate for the SCS reflects both the household interview completion rate and the student interview completion rate. In the 1989 and 1995 SCS, the household completion rates were 96.5% and 95.1%, respectively. The student completion rates were 86.5% and 77.5%, respectively. Multiplying the household completion rate times the student completion rate produced an overall SCS response rate of 83.5% in 1989 and 73.7% in 1995. The rate at which the respondents provide a valid response to a given item is referred to as its item response rate. Most items were answered by over 95% of all eligible respondents. The only exception was the household income question which was answered by approximately 90% of all households in both years.

Readers should be aware that the 1989 SCS estimates on victimization at school shown in this edition of SOURCEBOOK do not match the estimates presented in earlier BJS analyses of the 1989 SCS. In both the 1989 and 1995 SCS collections, persons 12 to 19 years of age were asked to respond to the NCVS and the SCS, and victimization information was captured in both questionnaires. The earlier researchers elected to use the victimization information reported in the NCVS, rather than the SCS, in the development of their school crime estimates. Because of a redesign of the NCVS in 1992, the 1995 victimization estimates from the NCVS cannot readily be compared to those developed before 1993. Therefore, the researchers performing the current analyses elected to reanalyze the 1989 data to compare estimates of victimization in 1995 to 1989 using the SCS data in both cases. It is possible that the redesign of the NCVS also had implications on responses to the SCS. However, it is not possible to measure the extent of the impact.

Sourcebook of Criminal Justice Statistics