Criminal Justice Research Methods: Design, Measurement, and Ethics

Comprehensive practice flashcards covering research design, experimental methods, crime measurement systems (UCR, NIBRS, NCVS), conceptualization and operationalization, levels of measurement, validity and reliability, and research ethics.

Experiment

A process of observation carried out in a situation expressly brought about for that purpose, especially suited for hypothesis testing, explanation, and evaluation.

Basic Experimental Design

The foundational experimental architecture where an experimental group and a control group are measured on a dependent variable, an experimental stimulus is administered to the experimental group, and both groups are remeasured to compare outcomes.

Hawthorne Effect

The phenomenon demonstrated by Roethlisberger and Dickenson (1939) where observed improvements in subjects' behavior or productivity occur due to the attention of being studied rather than the experimental manipulation.

Placebo

A dummy, inactive independent variable administered to a control group so that all participants receive identical procedural treatment, ensuring psychological expectations do not confound results.

Double-Blind Experiment

An experimental procedure in which neither the participants nor the experimenters know which individuals belong to the experimental group and which belong to the control group.

Random Assignment

A central feature of the classical experiment (randomization) that assigns subjects to groups by chance, producing statistically equivalent experimental and control groups and eliminating systematic selection bias.

Threats to Internal Validity

Nonrandom or confounding factors—such as history, maturation, testing, instrumentation, statistical regression, selection bias, experimental mortality, and causal time order—that cast doubt on whether the independent variable caused the change in the dependent variable.

Diabolical Dilemma

John Eck's (2002) concept describing the methodological trade-off where tightly controlled lab conditions maximize internal validity while reducing external validity, whereas field experiments enhance external validity at the expense of internal validity.

Statistical Conclusion Validity

The degree to which an empirical study has sufficient statistical power—dependent on sample size and magnitude of posttest differences—to correctly determine whether two variables covary.

Nonequivalent-Groups Design

A quasi-experimental design used when random assignment cannot occur, utilizing a comparison group rather than a control group and matching subjects or aggregate characteristics on key variables.

Interrupted Time-Series Design

A quasi-experimental longitudinal design in which a sequence of observations across time is compared before and after the introduction of an intervention.

Incident

One or more offenses committed by the same offender, or group of offenders acting in concert, at the same time and place.

Type I Offenses

Index offenses in the FBI's summary Uniform Crime Reporting system: murder, rape, robbery, larceny, burglary, aggravated assault, motor vehicle theft, and arson.

Type II Offenses

A compilation of less serious criminal offenses in the summary UCR that are only recorded and counted if an arrest is officially made.

Hierarchy Rule

A reporting convention in the summary UCR dictating that when multiple offenses occur within a single criminal incident, only the single most serious offense is counted.

National Incident-Based Reporting System (NIBRS)

An incident-level crime reporting system that replaced the summary UCR on January 1, 2021, capturing detailed incident, victim, offender, and arrestee data across Group A and Group B offenses without using the hierarchy rule.

NIBRS Data Structure

The relational system in NIBRS that links administrative data, multiple specific offenses, property records, victim characteristics, offender details, and arrestee profiles to a single unifying incident identifier.

Zero Reports

Reports submitted in NIBRS indicating that no Group A incidents or arrests, and no Group B arrests, occurred in a reporting jurisdiction during a given reporting period.

National Crime Victimization Survey (NCVS)

A longitudinal panel study conducted since 1972 by the U.S. Census Bureau for the BJS and DOJ that surveys households every 6 months for 3 years (7 total interviews) to illuminate the dark figure of unreported crime.

Telescoping

A cognitive recall error in victimization surveys where respondents mistakenly remember past events as having occurred more recently than they actually did.

Series Victimization

Six or more similar but separate victimizations that a respondent cannot distinctly separate or recall individually, recorded as a single collective measurement in surveys.

National Survey on Drug Use and Health (NSDUH)

An ongoing nationwide household self-report survey conducted since 1971 measuring lifetime, past-year, and past-month illicit substance use and misuse across household and institutionalized populations.

Monitoring the Future (MTF)

An ongoing self-report survey conducted since 1975 by the National Institute on Drug Abuse that monitors drug, alcohol, and delinquency behaviors and values among high school students and young adults.

Drug Abuse Warning Network (DAWN)

A long-term surveillance system established in 1972 that compiles emergency room reports and medical examiner data on drug episodes to monitor acute consequences of substance use.

Global Terrorism Database (GTD)

A specialized surveillance system developed by START that collects, screens, and codes open-source information to document worldwide terrorist incidents.

Conception

A mental image or subjective perception that an individual holds in their mind about a particular phenomenon.

Concept

Words, abstract terms, or symbols in language used to represent mental images or conceptions (such as gender, punishment, delinquency, and racism).

Direct Observables

Qualities or physical characteristics that can be observed directly and simply with the senses, such as color and shape.

Indirect Observables

Characteristics that cannot be directly perceived in real time and require more subtle, complex, or archival observations, such as court transcripts or criminal history records.

Constructs

Theoretical creations based on mental imagery that cannot be observed either directly or indirectly, such as intelligence quotient or crime seriousness.

Conceptualization

The process of specifying precisely what a researcher means when using particular terms, resulting in an agreed-upon working definition, dimensions, and indicators.

Reification

The error of regarding abstract concepts or mental creations as real, physical, concrete things.

Operationalization

The process of developing operational definitions that spell out the exact empirical procedures and data-collection methods used to measure a concept in the real world.

Exhaustive Measurement

A measurement quality ensuring that every observation can be classified into at least one attribute composing the variable.

Mutually Exclusive Measurement

A measurement quality ensuring that every observation can be classified into one, and only one, attribute composing the variable.

Nominal Measures

Variables whose attributes represent exhaustive and mutually exclusive names or categorical labels without any intrinsic mathematical ordering (e.g., race, gender, state of residence).

Ordinal Measures

Variables whose attributes can be logically ranked or ordered in terms of more or less, but the exact distance between attributes is unequal or unknown (e.g., education level, occupational status).

Interval Measures

Variables whose attributes have meaningful, equal, and standardized distances between ranks, but lack a true absolute zero point (e.g., temperature in Fahrenheit or Celsius, IQ).

Ratio Measures

The highest level of measurement, possessing equal distances between attributes and a true absolute zero point that enables proportional mathematical operations (e.g., age, number of prior arrests, sentence length, income).

Reliability

The consistency with which a particular measurement technique yields the same result each time it is repeatedly applied to the same object or phenomenon.

Face Validity

The assessment of whether an empirical indicator seems, on its surface and according to common understanding, to be an appropriate measure of the concept.

Criterion-Related Validity

The degree to which a measure relates to or predicts an external criterion, often demonstrated through convergent or discriminant empirical testing.

Content Validity

The degree to which an empirical measure covers the full breadth and range of meanings and dimensions contained within a theoretical concept.

Construct Validity

The extent to which an empirical test corresponds to the underlying theoretical concept and causal process it is intended to represent, evaluated via logical relationships between variables.

Typology

A taxonomy or classification system created by the intersection of two or more nominal or categorical variables to produce distinct types.

Nomothetic Causation

A probabilistic causal model (Shadish, Cook, & Campbell, 2002) requiring three specific criteria: empirical correlation, temporal ordering (cause precedes effect), and non-spuriousness.

Spurious Correlation

A coincidental or apparent statistical relationship between two variables that is actually caused by a third, unexamined confounding variable rather than a true cause-and-effect link.

Necessary Cause

A condition that must be present for the effect to occur (e.g., being officially charged is a necessary cause for being convicted).

Sufficient Cause

A condition that, if present, essentially guarantees that the effect will occur (e.g., pleading guilty is a sufficient cause for being convicted).

Ecological Fallacy

The methodological error of drawing conclusions about individual-level behaviors based solely on observations of aggregate or group-level data.

Individualistic Fallacy

The error of rejecting general, aggregate patterns or probabilistic relationships because of known, atypical individual exceptions.

Reductionism

The error of overly restricting a study's conceptual framework to a single narrow unit of analysis or discipline, failing to recognize multiple complex contributing factors.

Cross-Sectional Study

A study design based on observations collected at a single point in time, serving as an analytical snapshot of a population.

Trend Study

A longitudinal research design that tracks shifts or patterns within a general population over an extended period of time using multiple time points.

Cohort Study

A longitudinal study that tracks a specific, defined group of individuals who experience a common life event or enter an institution at the same time as they change over time.

Panel Study

A longitudinal study design in which data are collected from the exact same sample of individuals on two or more distinct occasions over time.

Retrospective Study

A research design that looks backward in time, asking subjects to recall past experiences or analyzing past institutional records.

Prospective Study

A longitudinal study that selects subjects in the present and follows them forward through time to observe future outcomes.

Belmont Report

A 1979 ethical statement establishing three core ethical principles for human subjects research: respect for persons, beneficence, and justice.

Anonymity

An ethical safeguard achieved when neither the public nor the researcher can identify a given response or piece of data with a specific participant.

Confidentiality

An ethical assurance where the researcher can identify a participant's responses but agrees never to release or disclose those identifying details publicly.

Institutional Review Board (IRB)

An institutional administrative committee charged with evaluating research proposals to assess participant risk, review ethics, and ensure adequate welfare safeguards.

Informed Consent

The ethical requirement that subjects possess the legal capacity to understand and do understand the study's purpose, risks, and procedures, consenting without coercion.

Assent

The formal agreement process for minors who cannot legally grant informed consent, granting them the choice to participate or refuse research involvement.

Tearoom Trade Study

A controversial 1975 field study by Laud Humphreys examining homosexual encounters in public park restrooms, severely criticized for invading privacy by tracking participants' vehicle license plates to conduct deceptive home interviews.

Stanford Prison Experiment

A 1973 simulated prison study conducted by Haney, Banks, and Zimbardo that assigned 21 male students as guards or prisoners; it was terminated after 6 days due to extreme psychological distress and role-induced abuse.

Situational Hypothesis

The theoretical premise tested in the Stanford Prison Experiment asserting that the structural environment and institutional conditions of a prison produce brutal, dehumanizing behaviors independently of individual personalities.

Dispositional Hypothesis

The perspective that prisons are brutal and dehumanizing primarily because of the innate personal characteristics and aggressive dispositions of inmates and correctional officers.