1/83
Comprehensive practice flashcards covering Topics 12 through 16 of Criminology Research Methods, including sampling methods, experimental and quasi-experimental designs, survey research, qualitative research design, and descriptive and inferential statistics.
What is a census?
A systematic process of collecting data from all members of a given population.
What are population parameters?
Statistics for the entire population, often estimated through statistics from a sample.
What does it mean for a study to "generalize"?
When findings from those studied can extend to predicting, explaining, or describing the behaviors of those who were not in the study.
What is external validity?
How well the results of a study generalize beyond the specific conditions and participants involved in the study.
What is internal validity?
The extent to which a study can establish a causal relationship between its variables by controlling for alternative explanations.
What is ecological validity?
How realistic a study is; how closely it resembles what happens in the real world.
What is the sampling model?
The model proposing that utilizing proper sampling techniques will yield a sample representative of the population — used when we trust randomization/technique to generalize to the population.
What is the proximal similarity model, and when do we use it?
A way of judging generalizability by matching contexts more similar (vs. less similar) in time, place, and units to the original study; used when we can't assume full representativeness and instead reason about how "close" a new context is to the one studied.
What is the theoretical (target) population?
Everyone to whom the researcher assumes the findings of a study could generalize.
What is the study (accessible) population?
Everyone from the theoretical population whom the researcher can feasibly access.
What is the sampling frame?
The list of individuals or units who are part of the study population.
What is the sample?
The subset of the population under study through research.
Who are respondents in research?
The individuals in the sample who respond to a survey or other research requiring a response.
What is selection bias?
An occurrence when the sampling procedure yields a sample that is not representative of the population it purports to represent.
What is nonresponse bias?
The concern that those in the sample who do not respond to a survey are different from those who do reply.
What is probability sampling?
Sampling procedures that utilize random sampling as part of their process (considered the "gold standard," though hard to achieve).
What is nonprobability sampling?
Sampling procedures that do not involve random selection of elements from the sampling frame.
What is simple random sampling, and when would we use it?
A method in which each unit in the sampling frame has an equal chance of being selected; used when probability sampling is feasible and we want the strongest claim to a representative sample.
What is snowball sampling, and when is it used?
A nonprobability procedure where participants help recruit others like them; used when the population is small, hard to reach, or marginalized.
What is convenience sampling?
Accessing research participants who are easy to access with minimal logistical preparation.
What is quota sampling, and when is it used?
A nonprobability method limiting the number of people from given groups who can participate; used when the researcher wants a specific ratio of types of participants (e.g., 35% dog owners, 65% cat owners).
What is an experimental design?
A group design where the researcher manipulates the attributes (levels) of an independent variable and measures the dependent variable; involves random assignment.
What is a quasi-experimental design, and how does it differ from an experiment?
A design like an experimental design, but it lacks random assignment, and the researcher is not always the one directly manipulating the independent variable.
What does "group design" being categorical mean?
A group is a category or classification separating participants based on characteristics or conditions relevant to the research — the independent variable in a group design is categorical (distinct groups), not continuous.
What is a pretest / posttest measure?
Measurements of the dependent variable taken before (pretest) and/or after (posttest) a treatment or intervention is introduced.
What is the One-Group Pretest, Posttest Design?
A group design with a single group that receives a treatment; the dependent variable is measured both before and after the treatment (no comparison group, no random assignment).
What is the Nonequivalent Groups Design?
A quasi-experimental design comparing a treatment group with a comparison group, but with no random assignment to groups.
What do we mean by "treatment"?
Something that is changed or that we think might influence the dependent variable — i.e., the independent variable; also called an intervention.
Why do Two-Group Experimental Designs with random assignment show the strongest causality?
Two-Group Experimental Designs with random assignment show the strongest causality because they achieve high internal validity via probabilistic equivalence.
Why do One-Group Pretest/Posttest and Nonequivalent Groups designs demonstrate causality weakly?
Because they lack a true comparison group and/or random assignment, leaving third variables unruled out.
What is random assignment, and why do we use it?
A process placing participants into different groups by chance rather than by choice; used to create probabilistic equivalence (the assumption that groups are equal at the start), which strengthens internal validity/causal claims.
What is probabilistic equivalence?
The assumption that groups are equal at the start of a study, created through random assignment to strengthen internal validity and causal claims.
What is a one-way factorial design?
An experimental design with three or more conditions of a single independent variable (multiple treatment groups, potentially plus a control group).
What is a two-way factorial design?
An experimental design that utilizes two independent variables, each with two or more conditions.
What is an interaction?
When the effect of one variable changes depending on a condition of another variable.
What is a field experiment?
An experiment that takes place in the field (real world) rather than in a laboratory setting; uses random assignment but has less control over confounds than a lab experiment, though more ecological validity.
Who is "in" a survey — what do we call the people and their responses?
The people are respondents (those who complete the survey); their answers come from survey items (questions).
What is a focus group?
A qualitative method for gathering insights, opinions, and feedback from a small, diverse group guided by a moderator, where participants represent a target audience.
When would we use a questionnaire vs. an interview?
A questionnaire (self-administered survey) is efficient/cost-effective and best for large samples; an interview (administered by a person) is better for depth, clarification, and complex topics.
What is social desirability bias?
A survey bias where respondents give socially acceptable rather than honest answers.
What are interviewer effects?
Changes in survey answers caused by who the interviewer is or how they behave.
What is a structured vs. unstructured survey item?
Structured items offer predefined answer choices (easier to analyze, may oversimplify); unstructured items are open-ended, letting respondents answer freely (richer detail, harder to analyze).
How do you calculate response rate?
(Number of completed surveys÷number of people who received the survey)×100 — i.e., smallest number divided by biggest number, times 100.
What is a dichotomous survey item?
A survey item with only two possible choices.
What is a contingency question?
A question that directs certain respondents to relevant follow-up questions based on their previous answers.
What is a filter question?
A question used to determine whether a respondent qualifies for or is eligible to complete the survey.
What is a false respondent?
Someone who lies on a filter question and gets passed through when their data should not actually be included in the study.
What is a double-barreled question, and how do you fix it?
A question that asks about two different topics at once without letting the respondent address each separately; fix it by splitting it into two separate, single-topic questions.
How specific or general should a survey question be?
Clear and specific enough to be unambiguous — avoid vague wording while still being broad enough to apply to all respondents.
What does it mean for survey response options to be mutually exclusive?
Each option stands on its own with no overlap.
What does it mean for survey response options to be exhaustive?
All possible answers are included as options.
What is a leading question in survey research?
A question that tries to influence the respondent toward a particular answer by suggesting a preferred/expected response.
What is a loaded question in survey research?
A question that contains an assumption about the respondent that forces an unfair or non-neutral answer.
What should opening survey questions look like?
Easy to answer, non-sensitive questions placed first to build rapport.
Where should sensitive questions be placed in a survey?
Sensitive questions go later with clear transitions.
Can surveys demonstrate causality?
Surveys are best suited for descriptive and relational research; with proper design (e.g., longitudinal studies) they can support causal inferences, but much more weakly than experimental designs.
Is qualitative research more idiographic or nomothetic?
Idiographic — qualitative research (and case reports especially) focuses on the unique, individual aspects of what's studied rather than generalizing broadly.
Why are idiographic case reports rare in criminology?
Criminology as a field tends to be nomothetic, which is why idiographic case reports are rare in it.
What type of data does qualitative research obtain?
Qualitative data — data relying on language and words rather than numerical values.
What is the difference between an approach and a method?
An approach is the overarching framework or strategy guiding how research is conceptualized; a research method is the specific procedure used within an approach to collect and analyze data.
What is field research?
An approach to research that involves collecting data directly from real-world environments where people live, work, or interact.
What is phenomenology?
A qualitative approach focused on how individuals experience and make meaning of a particular phenomenon being studied.
What is ethnography?
A qualitative approach focused on the in-depth, holistic study of a culture or social group, involving prolonged engagement in the field, often as a participant-observer.
What is participant observation?
An observational method where the researcher takes part in the activities being observed.
What is non-participant observation?
An observational method where the researcher studies a group or setting without directly engaging in the activities.
What is controlled observation vs. naturalistic observation?
Controlled observation occurs in structured environments where researchers manipulate variables and hold conditions constant; naturalistic observation occurs in the field/real world, unmanipulated.
What is positivism, and how does it relate to phenomenology?
Positivism assumes reality is objective and can be observed, measured, and explained through empirical, scientific methods; phenomenology contrasts with positivism by focusing on subjective lived experience and meaning-making rather than objective measurement.
How do we summarize qualitative data?
Using themes — subjects of information drawn from the data, rather than reporting the information verbatim.
How do structured vs. unstructured interviews differ, and when is each appropriate?
Structured interviews use predetermined, uniform questions (good for comparability and consistency across respondents); unstructured/qualitative interviews use open-ended questions through in-depth conversation (better for depth and flexibility).
What is unobtrusive research?
Methods of conducting research without directly interacting with those being studied.
What is content analysis?
A qualitative method involving the systematic examination and interpretation of communication content to identify patterns, themes, or meanings.
What is a case report (case study)?
A type of qualitative research involving detailed information on a single or small number of cases, usually emphasizing outlier/rare cases rather than generalizing.
How do normal and skewed data differ in what they demonstrate?
A normal distribution is symmetrical and bell-shaped, with most values clustering around the middle; skewness measures how far a distribution's shape deviates from that bell curve (asymmetry), signaling that the data are not evenly distributed around the mean.
What is a percentile ranking?
An indicator of the percentage of scores within a distribution that fall below a particular value.
If data are normally distributed and I know the standard deviation, what do I know?
I know how spread out the data are around the mean and can estimate what percentage of values fall within a given number of standard deviations of the mean (the basis of percentile/probability estimates for normal distributions).
What is the range?
The difference between the largest value in a set and the smallest value in the set.
What is the median?
The value in the middle of a dataset when all values are ordered from smallest to largest (or largest to smallest).
What is the mean?
The arithmetic average of a set of data — the sum of all values divided by the number of values.
What is the mode?
The value that occurs most frequently within a dataset.
What are inferential statistics, and why do we use them?
Statistics that allow researchers to estimate population parameters or test causal/statistical hypotheses — used to determine whether a pattern or group difference is statistically significant rather than due to chance.
What is a p-value, and what does it tell us?
An estimate of the likelihood that a given finding is due to chance alone; smaller p-values indicate the finding is less likely to be due to chance (i.e., more likely statistically significant).
When do we use Pearson's r?
When examining a linear relationship between two quantitative variables.
When do we use a t-test?
When the independent variable is categorical and dichotomous and the dependent variable is quantitative — tests whether there's a significant difference between the two groups' means.
When do we use a chi-square test?
When both the independent variable and the dependent variable are categorical.