U.S. flag

An official website of the United States government, Department of Justice.

Sexual Assaults Recorded by Law Enforcement, 2015 - Methodology

Kimberly H. Martin, Ph.D., BJS Statistician
Return to the 2015 Report

—Special Report—

 July 2021
NCJ 301235 
 July 2021

—Special Report—

NCJ 301235 
logotext image: NCS-X NIBERS. Collect. Analyze. Empower.

National Incident-Based Reporting System



The National Incident-Based Reporting System (NIBRS) is an annual data collection system that compiles information on criminal incidents and arrests reported by participating law enforcement agencies. NIBRS is collected and maintained by the Federal Bureau of Investigation (FBI). In 2015, NIBRS captured detailed incident characteristics for 49 Group A offenses, including four offenses that capture multiple forms of sexual assault: rape, sodomy, sexual assault with an object, and fondling.1

NIBRS provides the most detailed information available for sexual assault incidents that are handled by law enforcement each year. NIBRS includes demographic characteristics of sexual assault victims, offenders, and arrestees. Because NIBRS collects information on each criminal incident and each offense that occurs within an incident, NIBRS also provides crucial details about sexual assault incidents, such as whether the victim experienced multiple crime types; the presence of multiple victims and offenders; the relationship of the victim(s) to the offender(s); the time of day and type of location at which the incident occurred; whether a weapon was present during the offense and the type of weapon; the severity of injuries to the victim; and whether the incident was cleared and resulted in an arrest.

NIBRS represents a significant improvement over the FBI’s Summary Reporting System (SRS), which collected aggregate monthly counts of nine offenses and employed a hierarchy rule that required agencies to only report information on the most serious crime in the incident. On January 1, 2021, the FBI transitioned to NIBRS as the nation’s official law enforcement crime reporting standard and discontinued collection of Summary Crime Reporting program data.

Law enforcement participation in NIBRS is voluntary, and participating states and agencies are required to meet the FBI’s NIBRS certification requirements before they can submit their crime statistics in the NIBRS format. In 2015, the FBI reported that NIBRS data were provided by approximately 6,648 law enforcement agencies that collectively served more than 96 million persons. These 6,648 agencies represented approximately 36% of the more than 18,400 agencies that submitted Uniform Crime Reporting (UCR) data to the FBI in 2015.

Additional information about the status of the nation’s transition to NIBRS can be found at https://ucr.fbi.gov/nibrs-overview and at National Crime Statistics Exchange | Bureau of Justice Statistics (ojp.gov).

1 https://ucr.fbi.gov/nibrs/nibrs-user-manual

NIBRS Data Used for this Report

Agency participation and population coverage in the states

Though NIBRS is not yet nationally representative, a number of states submit all of their annual crime statistics to the UCR via NIBRS. In states where NIBRS is the primary crime data reporting standard, the NIBRS data provide the most reliable representation of all crimes and arrests handled by law enforcement within the state. Law enforcement records of sexual assault victimization are not necessarily representative of all sexual assaults. For information on sexual assaults not reported to the police, see Criminal Victimization, 2015 (NCJ 250180).

This report presents state level statistics on sexual assaults recorded by law enforcement. We relied primarily on the percentage of population covered when determining which states were eligible for inclusion in this report. Agency participation among the full certification states ranged from 55% (West Virginia) to almost 99% (Delaware) (Appendix Table 1). Although the submission rate for West Virginia is significantly lower than the other fully certified states (55%), most of the non-participating agencies in West Virginia cover a fraction of the state’s overall population. The West Virginia agencies that did submit NIBRS data in 2015 represent over three-fourths of the state’s resident population (78%). Vermont is a fully certified NIBRS state, but was excluded from this report due to agency non-participation. Across the 15 states selected for this report, a total of 4,543 agencies submitted NIBRS data in 2015. This report excludes crime data from 235 agencies that did not submit 12 months of data.

Appendix Table 1 displays the states selected for this report and each state’s law enforcement agency NIBRS submission rates for 2015. The data used to calculate NIBRS agency response rates and state NIBRS population coverage were downloaded from files located on the FBI’s Crime Data Explorer (CDE).2 CDE values may differ slightly from NIBRS reported values shown above. These data are provided by the FBI in downloadable files that contain the law enforcement agencies that are certified to submit NIBRS data and an indicator whether these agencies submitted a complete year of data. Agency participation rates were calculated by dividing the total number of certified agencies by the total number of agencies that submitted NIBRS data to the state, and multiplying by 100. The same files used to calculate agency participation rates also contain the size of the population served by NIBRS-submitting agencies. NIBRS population coverage was calculated by summing the total population of all eligible agencies in the state and dividing that total by the sum of the population served by NIBRS-submitting agencies, and multiplying by 100.

Units of count

Incident is the unit of count used for analyses of clearance and arrest. The FBI’s UCR program defines an incident as one or more offenses committed by the same offender or group of offenders acting in concert at the same time and place. NIBRS requires law enforcement agencies to count as offenders persons who knowingly commit or assist in the commission of the offense(s) in the incident. For each incident, NIBRS captures information on up to 10 offenses, whether the incident was cleared, and whether the incident was cleared by an arrest or by exceptional means. Because NIBRS considers each offender in an incident to have committed or assisted in the commission of all offenses in an incident, an exceptional clearance or the arrest of one offender clears the entire incident, even if there were multiple offenses, victims, and offenders present.

A victimization represents one or more victim segment(s) connected to an incident. For crimes against persons, such as sexual assault, each victim in the incident will generate a separate victim segment, each of which contains information about the offense(s) committed against the victim, victim demographic characteristics, and the relationship of the victim to the offender(s). NIBRS requires that law enforcement agencies submit at least one victim segment for an incident in which a Group A offense was committed.

For more information on the data elements in the victim and incident segments of NIBRS, see the FBI NIBRS user manual.

NIBRS data extracts and measures

This report is based on analyses of NIBRS victim and incident-level extract data files for 2015, which were obtained from the Inter-university Consortium for Political and Social Research (ICPSR). NIBRS Group A incident reports are a series of data elements that are stored in six separate data segments linked together by incident identifiers: the administrative, offense, property, victim, offender, and arrestee segments. The NIBRS data extract file created for this report combined data elements from these segments into a single flat file that can be analyzed using standard statistical software packages. The victim extract consists of a data file for which each victimization is a unique row with the victim’s demographic characteristics, injury information, and how the victim was related to the offender, along with data elements from other segments that are linked to each unique incident and victim record, such as the time of day in which the incident occurred, type of location, and whether the offender possessed a weapon and the weapon type, and whether an arrest was made.

The victim extract file used for this report contained victimizations involving all serious violent crimes to allow for comparisons of sexual assault characteristics with those observed in other serious violent crimes. NIBRS measures sexual assault crimes using four offense types: rape, sodomy, sexual assault with an object, and fondling. Analyses of other serious violent crime types included all victimizations and incidents that involved at least one of the following offenses: murder/non-negligent manslaughter, negligent manslaughter, kidnapping/abduction, robbery, and aggravated assault.

An incident-level NIBRS extract data file was used to produce clearance and arrest statistics for sexual assault and other serious violent crime types. NIBRS contains indicators of whether an incident was cleared by arrest or cleared by “exceptional means.” Agencies may record cases as cleared by exceptional means when probable cause exists to support the arrest of a known offender and the offender’s whereabouts are known, but law enforcement is prevented from making an arrest. NIBRS includes data values indicating why an incident was cleared by exceptional means: the offender died; prosecution was declined; the victim refused to cooperate; the offender was already in custody in another jurisdiction; or the offender was a juvenile and not taken into custody.

Victimization rates shown in this report are calculated per 100,000 residents. Rates of sexual assault victimization for each state were calculated by dividing the state’s 2015 population by the total number of sexual assault victimizations recorded by that state’s law enforcement agencies, and then multiplying by 100,000. Rates for specific demographic groups were calculated by dividing the total number of residents belonging to that group, according to 2015 Census population estimates, dividing by the number of rapes and sexual assaults committed against that demographic group, and then multiplied by 100,000. Rates for persons of Hispanic origin were not calculated due to insufficient reporting on the Hispanic origin of victims, offenders, and arrestees. All rate calculations are based on state population estimates from the U.S. Census Bureau’s 2015 population estimates program.3

3Annual State Resident Population Estimates for 6 Race groups (5 Race Alone Groups and Two or More Races) by Age, Sex, and Hispanic Origin: July 1, 2010 to July 1, 2019. U.S. Census Bureau, Population Division.
Missing and unknown data values

Missing data patterns vary by state and NIBRS data element. Missing values are noted under each figure and table in the report. Missing data values were highest for offender demographic characteristics and the relationship of victims to offenders since these rely on witness, victim, or law enforcement accounts, which may not be possible if an offender’s identity is unknown. Depending on the state, the rate of missing data for offender age, race, and sex ranged from 4% to 20% of sexual assault victimizations. On average, about 17% of all sexual assault victimizations were missing information pertaining to the relationship between victims and offenders. This ranged from 7% in Kentucky to 45% in Arkansas. Incidents that were not cleared by law enforcement were more likely to have unknown or invalid data values for offender demographics and victim-offender relationship. In each of the 15 states in this report, fewer than 16% of sexual assault victimizations had incomplete data for victim age, race, or sex.

Incident time of day was not uniformly recorded by law enforcement agencies. The NIBRS reporting specifications allow participating agencies and states to submit information on the time of day in which an incident is reported to law enforcement in lieu of the time of day in which the incident occurred. Most agencies and states submit information on the time of day in which the incident occurred. Statistics for incident time of day were not available for the state of Delaware, which submits information on the day and time in which an incident was reported to police rather than when the incident occurred.

Cases with missing or invalid data values were excluded from this report, with one exception – victim race. Part 3 of this report displays statistics on rates of victimization for specific race groups, adjusted for each group’s population size. States varied in the extent to which records were missing victim race. Among the states included in this report, the percentage of sexual assault records missing victim race ranged from 1% to 16%. Cases for which victim race was missing were not excluded from calculations of victimization rates, and were instead included in the victim race category of Other. This approach to handling missing data when calculating victimization rates reduced the variability in state victimization rates that was driven primarily by variability in the number of victimizations that were removed from a state’s victimization count, i.e., the numerator, due to missing victim data.

Terms and Definitions

  • Change in the UCR definition of rape
      In 2013, the FBI UCR Program began collecting summary counts of rape using a new definition: the penetration, no matter how slight, of the vagina or anus with any body part or object, or oral penetration by a sex organ of another person, without the consent of the victim. Attempts or assaults to commit rape are also included; however, statutory rape and incest are excluded. Prior to 2013, the summary reporting program defined rape offenses as the carnal knowledge of a female forcibly and against her will.
  • Clearance and arrest
      In the FBI’s UCR program, agencies may report an incident cleared (i.e., solved or closed) in one of two ways: by an arrest or by exceptional means. Under UCR data guidelines, incidents that are administratively closed do not necessarily meet the criteria to be reported as cleared. An incident cannot be cleared by exceptional means if it was previously or concurrently reported as cleared by arrest. For the purposes of NIBRS, any incident that is not cleared by an arrest or exceptional means is considered not cleared.
    • Cleared by arrest includes any on-view arrest that occurs absent a warrant or previous incident report, arrests where an offender is taken into custody based on a warrant or previously submitted incident report, and incidents for which an offender is issued a summons to appear in court but not taken into police custody.
    • Exceptional clearance occurs when an incident is cleared without an arrest due to circumstances outside of the control of law enforcement. To clear an offense by exceptional means, agencies must meet all of the following criteria: an investigation must have clearly established the identity of a least one offender; there must be sufficient probable cause to support the arrest, charging, and prosecution of the offender; law enforcement must know the exact location of the offender so that an arrest could be made had circumstances not prevented the arrest; and there must be a reason outside the control of law enforcement that prevents law enforcement from making an arrest and charging the offender, such as death of the offender, the prosecutor declining prosecution for reasons other than lack of probable cause, the victim refusing to cooperate in the prosecution, the offender already being in custody in another jurisdiction (includes extradition denied), or if the offender was a juvenile and not taken into police custody.
  • Incident time of day
      About 12% of the sexual assault victimizations analyzed for this report occurred at midnight, resulting in a spike of victimizations at that hour. A spike in victimizations that occur at 12 a.m. may be related to how time of day data are entered into law enforcement agency records management systems (RMS). NIBRS data validation protocols are designed to reject incidents that are missing a valid value for the time of day in which the incident occurred. To avoid incidents being rejected for missing incident time of day, some agencies may default to entering a value of 0 in their records management systems (RMS), which would help account for a spike in incidents that occur at 12 a.m. To correct for this, the average number of victimizations that occurred at 11 p.m. and 1 a.m. was assigned to 12 a.m., and victimizations that were removed from 12 a.m. were distributed evenly across the hours of 1 a.m. to 11 p.m.
  • Juvenile and adult age categories
    • Juvenile victims are age 17 or younger.
    • Adult victims are defined as victims ages 18 and older.
  • Location type
      In 2019, NIBRS contained 57 separate location types that could be recorded. These location types were collapsed into 6 categories used for this report.
    • Residence/hotel includes victimizations that occurred at hotels, motels, or residences.
    • Restaurant, bar, or sports/entertainment venue includes victimizations that occurred at bars/nightclubs, restaurants, amusement parks, arenas/stadiums/fairgrounds/coliseums, and gambling facilities/casino/racetracks.
    • Retail/financial/other commercial establishment includes victimizations that occurred at banks or savings and loan institutions; commercial or office buildings; convenience stores; department or discount stores; drug stores, doctor’s offices, or hospitals; grocery stores or supermarkets; liquor stores; rental storage facilities; service or gas stations; specialty stores; ATMs separate from a bank; new or used auto dealerships; and shopping malls.
    • Schools, day cares & universities includes victimizations that occurred at schools/colleges, day care facilities, colleges/universities, and elementary/secondary schools.
    • Transportation hub/outdoor public locations includes victimizations that occurred at air, bus, or train terminals; in fields or woods; on highways, roads, alleys, streets, or sidewalks; lakes, waterways, or beaches, parking lots or garages; camps or campgrounds; docks, wharves, or freight or modal terminals; parks or playgrounds; and rest areas.
    • Other recorded location location includes victimizations that occurred at businesses and commercial establishments, government buildings and properties, churches and religious centers, other public places, and cyberspace.
  • Race of victim
    • The category of Other includes Asians, Native Hawaiians, Other Pacific Islanders, and victims for whom race information was unknown.
    • Native Hawaiian does not include individuals who are native to the state of Hawaii only by virtue of being born there.
  • Relationship of victim to the offender
    • Intimate partner includes victims who are the spouse, common-law spouse, ex-spouse, or the boyfriend/girlfriend of the offender.4
    • Other family includes victims who are the parent, sibling, child, grandparent, grandchild, in-law, stepparent, stepchild, stepsibling, or other family member of the offender.
    • Friend/acquaintance includes victims who are an acquaintance, friend, neighbor, babysittee, child of boyfriend/girlfriend, employee, or employer, or who are otherwise known to the offender.
  • Sexual assault offense types
    • NIBRS defines rape (except statutory rape) as the carnal knowledge of a person without the consent of the victim, including instances where the victim is incapable of giving consent because of his/her age or because of his/her temporary or permanent mental or physical incapacity.
    • NIBRS defines sodomy as oral or anal sexual intercourse with another person, without the consent of the victim, including instances where the victim is incapable of giving consent because of his/her age or because of his/her temporary or permanent mental or physical incapacity.
    • NIBRS defines sexual assault with an object as the use of an object or instrument to unlawfully penetrate, however slightly, the genital or anal opening of the body of another person without the consent of the victim, including instances where the victim is incapable of giving consent because of his/her age or because of his/her temporary or permanent mental or physical incapacity
    • NIBRS defines fondling as the touching of the private body parts of another person for the purpose of sexual gratification without the consent of the victim, including instances where the victim is incapable of giving consent because of his/her age or because of his/her temporary or permanent mental or physical incapacity.
  • Victimization type
    • Non-sexual violent crime victimizations consist of murder and non-negligent manslaughter, negligent manslaughter, kidnapping/abduction, robbery, or aggravated assault.
    • The terms sexual assault and sexual assault victimization are used interchangeably in this report to describe instances in which an individual was the victim of rape, sexual assault with an object, sodomy, or fondling.
  • Violent victimizations
      Violent victimizations are defined as any victimization in which an individual was the victim of murder, non-negligent manslaughter, negligent manslaughter, kidnapping/abduction, robbery, aggravated assault, rape, sexual assault with an object, sodomy, or fondling.
  • Weapon type
      NIBRS captured 21 weapon types in 2015, including whether no weapon was present. Weapon types were combined into 5 categories:
    • Firearm includes: firearm (type not stated), firearm-automatic, handgun, handgun automatic, rifle, rifle automatic, shotgun, shotgun automatic, other firearm, and other firearm-automatic.
    • Personal weapons includes using his or her hands, fist, feet, arms, teeth, etc. during an offense.
    • Other weapon includes blunt object, motor vehicle, explosives, fire/incendiary device, asphyxiation, and other weapon.
    • Drug or poison includes poison/gas and drugs/narcotics/sleeping pills.
    • None
4 The FBI is no longer using “homosexual relationship” code “HR” as of Jan 1, 2019. See FBI's UCR Program Quarterly, July 2017 for more information (p6).