CRIM EXAM 2

Popular literature

Literature written primarily for the general public rather than an academic audience, usually aimed at informing, entertaining, persuading, or attracting readers, and typically not peer-reviewed.

Scholarly literature

Literature produced for an academic or professional audience that presents research, theory, or scholarly analysis and typically follows academic conventions such as citations and references.

Academic journal

A publication that regularly publishes scholarly work within an academic discipline, such as research articles, reviews, and theoretical articles.

Peer review

A process in which experts in the same or a related field evaluate scholarly work before publication to assess its quality and suitability for publication.

Abstract

A brief summary of an empirical research paper, usually including the research purpose/question, basic methodology, major findings, and conclusions.

Introduction / Literature review

The section of a research paper containing background information, previous research, relevant theories and concepts, conceptual definitions, and the research problem or question.

Methods section

The section of a research paper describing how the study was conducted, including details about the sample or participants, measures, operationalization of variables, procedures, and research design.

Results section

The section of a research paper that presents the study's findings, including statistical analyses or other results generated from the data.

Discussion section

The section of a research paper that interprets the findings, discusses implications and ties to theory or past research, identifies limitations, and suggests future research directions.

Hypothesis

A specific and testable prediction about an expected relationship or difference involving variables.

A priori

A research term meaning specified or determined before examining a study's results. Hypothesize before data collection

Research question

An inquiry that asks what relationship or phenomenon will be investigated, contrasting with a hypothesis which predicts what will be found.

Theory

An explanation of how or why phenomena occur, which researchers use to generate specific, testable predictions.

Rival hypothesis

An alternative prediction or explanation that competes with another hypothesis about the same phenomenon.

Multivariate research

Research examining relationships among more than two variables, allowing researchers to evaluate additional factors that might influence the primary relationship.

Bivariate research

Research that examines relationships between exactly two variables.

Covariability

A required causal criterion indicating that two variables are related, meaning values of one variable systematically differ as values of the other change.

Temporal precedence

A required causal criterion stating that the proposed cause must occur before the proposed effect.

Spurious relationship

An observed statistical relationship between two variables that does not reflect a true causal relationship, often because a third variable or process explains the association.

True experiment

A research design that employs manipulation and random assignment to conditions, providing strong internal validity by addressing alternative explanations.

Quasi-experimental design

A research design that lacks random assignment, resulting in generally weaker internal validity because groups may differ systematically prior to the treatment or event.

Secondary data

Data that were previously collected rather than gathered directly by the researcher for the specific study at hand.

Population

The larger set of units or cases that researchers intend to study or make conclusions about.

Census

An attempt to collect information from every single unit or member of a target population rather than selecting a subset.

Sampling

The process of selecting a subset of units from a population for inclusion in a study.

Haphazard sampling

A weak nonprobability sampling method in which cases are selected without any systematic strategy, relying on whatever cases happen to be available.

Measurement reliability

the consistency or stability of a measure.

Measurement validity

how well a tool or instrument measures what it claims to measure.

The ideal measure of a concept…

is both reliable and valid

Test-retest reliability

the extent to which a particular measure produces similar results when administered to the same sample under the same conditions but at different times

Some measures require one or more individuals to act as raters

individuals who evaluate, score, judge, or classify data, behavior, or responses according to a set of criteria

Intra-rater reliability

the consistency of a single rater reviewing the same data at different times

Inter-rater reliability

the consistency of multiple raters reviewing the same data at the same time.

parallel-forms reliability

is how well two different versions of a measure designed to measure the same variable produce similar results when given to the same group or sample

internal consistency

how well each item on a scale measures the same underlying concept.

Face Validity

Whether a given measure appears to measure what it claims to measure

Content validity

how well a measure adequately represents the entire domain of the concept it is intended to measure

Conceptual validity

how true a given measure or operational definition is to the conceptual definition of what it is intended to measure.

Convergent validity

a type of construct validity: where a given measure positively correlates with other measures that are supposed to be related.

Discriminant Validity

A type of construct validity whereby a given measure fails to correlate with other measures that should not be related

Criterion Validity

How well a measure correlates with a known benchmark (i.e., criterion) that represents the concept one is trying to measure

Concurrent Validity

A type of criterion validity that examines whether a measure agrees with an existing validated measure at the same time

Predictive Validity

A type of criterion validity that examines whether a measure can forecast outcomes that occur later in time

Strong measurements of concepts are…

both reliable and valid

A measure has good reliability when…

it gets consistent results

A measure has good validity when it measures…

what it is supposed to be measuring

Some measures are reliable without…

being valid, meaning that they are consistently measuring something but not what they were supposed to be measuring.

Conceptual validity subsect

convergent validity and discriminant validity

Criterion validity subsects

concurrent validity and predictive validity

Parallel forms requirements

Measure the same concept or construct (e.g., anger, trust in police, delinquency, and likelihood of recidivism)

Have comparable difficulty, structure, and content

Be administered under the same conditions

Parallel forms can….

also avoid test fatigue

If there is a lack of internal consistency…

then that tool is measuring more than a single concept

High content validity is…

directly associated with high conceptual validity.

Convergent validity matters….

because it is a way to demonstrate that a measure is not just made up and arbitrary.

Discriminant validity demonstrates…

that a given measure is not just overlapping with other random measures.

Which of the following is a means of increasing inter-rater reliability?

Making sure that the instructions to the raters are clear and objective.

Which type of validity or reliability involves comparing a measure to a gold standard measure of the same concept?

concurrent validity

“This concept has six different dimensions. I need to make sure that my measure of the concept includes all six dimensions.” This person is talking about what type of reliability or validity?

content validty

In what situation, if any, would a researcher want to avoid high face validity of a measure?

to prevent participants from influencing their responses.

What is one reason why a researcher would have parallel forms of a given measure?

to ensure the same construct is measured in a similar way across different versions, preventing practice effects or bias.

What is the main purpose of sampling?

To study a subset of a population and use the findings to generalize to the larger population.

What is external validity?

How well a study’s findings generalize beyond the specific sample and conditions studied.

What is the Sampling Model?

The idea that if researchers use proper sampling techniques to obtain a representative sample, findings should generalize to the population.

What is the Proximal Similarity Model?

Findings generalize best to people, places, times, and situations that are most similar to those actually studied.

Proximal Similarity Model Example

A study of criminology students in 1975 probably generalizes better to criminology students today than to toddlers.

Sampling Model vs. Proximal Similarity Model?

• Sampling Model: Generalization depends on whether the sample represents the population.

• Proximal Similarity: Generalization depends on how similar the new context is to the original study context.

What is the order of the major sampling terms?

Theoretical population → Study population → Sampling frame → Sample → Respondents

Theoretical population:

Everyone you want to generalize to

Study population:

Everyone you can feasibly access

Sampling frame:

The list of accessible units

Sample:

Units selected/contacted for the study

Respondents:

Members of the sample who actually respond

What is selection bias?

When the sampling process produces a sample that is not representative of the population.

What is nonresponse bias?

When people who respond are meaningfully different from those who do not respond, potentially skewing the findings.

What is response rate?

The percentage of the sample that actually responds to or completes the study.

What is the difference between probability and nonprobability sampling?

• Probability sampling: Uses randomization

• Nonprobability sampling: Does not use randomization

Why is probability sampling the “gold standard”?

Because random selection helps produce a representative sample and reduces the likelihood of selection bias, strengthening generalizability.

What is simple random sampling?

Probability sampling where every unit in the sampling frame has an equal chance of being selected.

Convenience Non probability

Choose people who are easiest to access.

Snowball Non probability

Participants recruit other participants like themselves. Useful for small or marginalized/hard-to-reach populations.

Quota non probability

Recruit until predetermined amounts/ratios of certain groups are reached.

Do sampling concepts apply only to people?

No. The unit of analysis can be anything being studied, such as court cases, neighborhoods, police departments, or objects.

Bivariate research

research involving only two variables

Categorical data

data placed into individual categories or groups

Causal relationship

the relationship where one event (the cause) directly leads to another event (the effect)

Causal research question

research questions that examine whether a statistical relationship between variables exists and whether that is causal. Asks the question, “Does a change in X cause a change in Y?

Causality

the relationship where one event (the cause) directly leads to another event (the effect)

Cause

a person or thing that acts, happens, or exists in such a way that some specific thing happens as a result; the producer of an effect

Complexity

having many interconnected or intricate parts or causes

Confirmation bias

the tendency to seek, interpret, and favor information that supports one’s preexisting beliefs, while ignoring or dismissing contradictory evidence

Consumers of research

someone who utilizes research in some way; a reader of published research

Covariability

when one variable changes at approximately the same time as another variable; a statistical relationship between variables

Illusory correlation

when someone views variables as associated or correlated when there is no association between the two, or when people assume that an actual association is larger than it really is

Inferential statistics

use of a sample to make generalizations or predictions about a larger population; tests used to determine if findings are statistically significant

Probabilistic

a likelihood or chance of something occurring rather than it occurring with absolute certainty

Research design

the way that research is conducted in a study, including the methodology employed; ways of collecting or analyzing data for studies

Ruling out alternative explanations

demonstrating that the relationship between two or more variables is not due to some third variable that is unaccounted for

Significance testing

a statistical method used to determine whether the results of a study are likely to have occurred by chance alone

Significant correlation

a relationship between two or more quantitative variables that is statistically significant based on significance testing

Significant difference

a difference between two or more categorical variables that is statistically significant based on significance testing