criminology -2025-08-19T14:40 .:33.740Z
Methods of scientific research (three main categories)
Quantitative research
Focus: numbers, data, and measurement
Characteristic: objective; “the numbers are the numbers” with little gray area
Uses: statistical analysis, testing hypotheses, large sample sizes
Qualitative research
Focus: text, language, and observations
Characteristic: more subjective; interpretive analysis
Uses: interviews, focus groups, text analysis, case studies, literature reviews
Process: interview participants with the same or similar questions, transcribe into written form, extract themes and word patterns to build a word map and draw conclusions
Mixed methods research
Combines quantitative and qualitative in one study
Rare due to the workload and complexity; involves collecting and integrating both data types
How these methods relate to research practice
All research starts with one of these three approaches
Choice depends on goals: measurement and testing (quantitative), understanding and exploration (qualitative), or both (mixed)
Quick comparison (video summary)
Quantitative:
Test hypotheses; data in numbers and graphs; requires larger samples
Data analyzed via math/statistics
Qualitative:
Formulate hypotheses; data in words; smaller samples
Data analyzed by summarizing, categorizing, and interpreting
Mixed methods:
Use both approaches in one study; increases methodological rigor but is resource-intensive
Example: student satisfaction in a university
Quantitative: survey 300 students; scale 1–5; result e.g., average score (e.g., 4.4/5)
Qualitative: interviews with 15 students; open-ended questions; identify recurring themes (e.g., need for more one-on-one guidance)
Mixed: start with interviews to generate hypotheses, then test with a larger survey; or start with a survey and follow up with interviews to understand reasons
Data collection for the three approaches
Quantitative data collection methods
Online surveys, in-person surveys, phone surveys
Experiments
Observations (quantified)
Qualitative data collection methods
Interviews, focus groups, case studies, literature reviews
Mixed methods sequencing
Start with qualitative interviews to gain insights, then test with a quantitative survey (larger sample)
Or start with a survey to capture the larger picture, then use interviews to understand the reasons behind results
Transition to criminology data sources (three primary sources)
Criminology relies on multiple data streams, commonly from government agencies and research surveys
Three primary sources (core focus of the lecture):
Uniform Crime Reports (UCR) / National Incident-Based Reporting System (NIBRS)
National Crime Victimization Survey (NCVS)
Self-report surveys (often by academics)
Ethical issues and data provenance matter when interpreting these sources
Deep dive: UCR/NIBRS (Uniform Crime Reports → National Incident-Based Reporting System)
What is UCR/NIBRS?
The UCR program is the long-standing, well-known crime data source collected from police reports across the U.S.
It has evolved into NIBRS, a more detailed, incident-based system that traces crimes more comprehensively.
How data are collected and reported
Data come from over 17,000 police agencies across the United States; not every agency reports every month
Data flow: local agency collects reports → state agency aggregates (e.g., SLED in SC, Alabama SBI) → FBI collects and compiles into national datasets
Reports are vetted before submission to FBI; the FBI then publishes national data via the Crime Data Explorer (CDE)
Part I (index) crimes vs Part II crimes
Part I crimes (index crimes) are historically the focus; eight categories total:
Violent crimes: Homicide, Rape, Robbery, Aggravated Assault
Property crimes: Burglary, Larceny, Motor Vehicle Theft, Arson
1979 addition: Arson added to the index list
Part II crimes: crimes other than index crimes plus minor traffic offenses
Key properties and limitations
Strengths:
Large-scale, nationwide baseline based on police reports
Good for trend analysis of reported crimes, arrest data, demographics at arrest
Data are collected monthly and updated (vetting and standardization occur up the chain)
Limitations and common critiques:
Underreporting: not all crimes are reported to the police; common in many crime categories (the “dark figure of crime”)
Reporting practices: differences in local reporting, charging, and coding practices can distort national totals
Definitions and coding differences: state laws and FBI/NIBRS definitions don’t always align; e.g., a crime may be coded differently in different jurisdictions
Not a complete picture: not all agencies participate in NIBRS; federal offenses are not counted in NIBRS data
Misclassification: e.g., a simple burglary vs trespassing vs vandalism can be coded differently depending on jurisdiction
Apparent reliability vs reality: a high “clearance rate” can be misinterpreted if the denominator (reported incidents) is biased
The shift from UCR to NIBRS
1982: UCR redesign began due to holes in the old system; aim to provide more detail and accuracy
Old UCR: if an incident involved multiple crimes, only the highest offense might be counted (hierarchical reporting)
NIBRS: counts all crimes within an incident (e.g., if a fight included vandalism and assault, both are reported)
NIBRS collects incident-level data: location, victim and offender data, relationships, weapons, drugs/alcohol involvement, etc.
46 offenses in NIBRS (including the eight index crimes and many additional categories) and 46+ offenses broken down with subcategories
1982–2021: gradual adoption; by 2021 (01/01/2021) NIBRS became the standard
Coding, offenses, and examples
Codes exist for each offense (example codes):
13a: Aggravated Assault
13b: Simple Assault
13c: Intimidation
Robbery reclassification: now considered a crime against property rather than a personal offense in certain classifications
NIBRS adds many minor offenses and more detailed classifications (e.g., breaking down “Burglary” into specific categories like breaking into a building vs. a vehicle, etc.)
Data accessibility and use in coursework
The Crime Data Explorer (CDE) is the public interface to view NIBRS data and analyze trends by month, location, victim/offender demographics, etc.
You can customize views by month and offense, and extract data for analysis
Instructors often require students to pull data from CDE to compute totals over a period and to report per 100,000 people when comparing rates across years
Example usage (CDE snapshot in class demonstration)
Viewing homicides in a given month (e.g., February 2022): total homicides, percentage of population coverage, and the monthly or annual clearance rate
With CDE, you can see: location type (where it happened), victim-offender relationship, offender age, victim age, sex, race/ethnicity, etc.
Common finding: most homicides occur in residences (including homes) and relationships often unknown; many cases have unknown relationships due to ongoing investigations
Weapon type distributions commonly show handguns as the leading weapon in homicides and aggravated assaults
Crime Data Explorer (CDE) specifics
CDE provides a public interface to national and state data; you can select a crime (e.g., Homicide) and a time window to view monthly counts and related demographics
Key metrics shown in CDE visuals
Monthly counts of crimes reported to police (blue line in the demo)
Population coverage (top line; essentially the percentage of the population included in the data)
Clearance rate (gray line; cases with an arrest and charge or exceptional means)
Important caveats when using CDE
The national totals may require summing month-by-month data to get a period total since the public interface emphasizes per-month views
Population coverage is not the same as sample size; it indicates geographic and agency coverage
Clearance rates are influenced by multiple factors (arrests, extradition, jurisdictional issues, deaths, etc.)
Practical tips for students
When writing papers, compute total offenses over the specified years by adding monthly counts
Use per 100,000 population rates to compare across years or jurisdictions
Include location and demographic breakdowns (e.g., victim-offender relationship, age, sex, race) to add depth to analysis
National Crime Victimization Survey (NCVS)
Purpose and design
Created by the Bureau of Justice Statistics (BJS) in 1973 to estimate the “dark figure” of crime—the crimes not reported to police
It is a victim-focused survey, meaning it asks respondents about their victimization experiences
It captures information not typically present in police reports, including the frequency and consequences of crimes
Sample and methodology
Approximately 95,000 households in a three-year panel; ages 12 and up in those households are eligible
The panel is followed for seven interviews over three years (an intake followed by six follow-ups)
Data collection has evolved from mail surveys to computer-assisted, web-enabled or interview-based approaches
What NCVS collects
Victim demographics: age, race, gender, income, education
Crime characteristics: type of crime, time, location, weapons used, injury, economic impact (e.g., job loss, hospitalization)
Experiences with the criminal justice system and the victim’s perception of justice response
Information on whether victims used protective measures
Coverage and limitations
Not all crimes are surveyed; notably, homicide is excluded by design because you cannot interview a deceased person
Advantages: provides a broader view of crime beyond police reports, helps quantify the dark figure of crime, reveals victim experiences and economic impact
Limitations: relies on self-report; potential misreporting or misinterpretation by respondents; memory biases; some crimes may be misclassified; memory decay over time; respondents may fear repercussions or fail to report certain details
Examples and insights from NCVS comparisons
Rape/Sexual assault: historically a high rate of underreporting in police data; NCVS often reveals higher victimization levels than police reports alone
In 2012, NCVS figures for rape suggested substantially higher victimization than UCR police counts (rape is one of the most underreported crimes in police data; NCVS helps illustrate the true scale)
Limitations and interpretation cautions
The NCVS cannot measure homicide (deceased victims cannot be surveyed)
Self-reported data require careful interpretation due to potential misclassification and recall biases
Self-report surveys (academic research surveys)
What they are
Surveys designed and conducted by university researchers