Conceptualization
What is the difference between a conception and a concept?
A conception is a personal mental image of an idea; a concept is the shared word/symbol we use to communicate that idea.
Is “gender” nominal, ordinal, interval, or ratio?
Nominal. Gender is just a label with categories, no order.
Define reliability in your own words.
Reliability = consistency. A measure gives the same results when repeated under the same conditions.
Define face validity.
Face validity = Does the measure look like it's supposed to?
What is a composite measure?
A composite measure combines multiple indicators into one index or scale.
Spells out exactly how a concept will be measured in the real world (e.g., measuring SES through survey questions about income and education).
An operational definition.
Annual income is measured in dollars. What level of measurement is this, and why?
Ratio. Annual income has equal intervals and a true zero point. You can say one person makes twice as much as another.
What is the difference between the test-retest and interrater reliability methods?
Test-retest: Apply the same measure at two different times and compare.
Interrater: Two or more observers measure the same thing and check agreement.
Which type of validity compares a measure to an actual outcome, like predicting rearrest?
Criterion validity = Compares the measure to a concrete outcome (e.g., predicting rearrest).
Give one example of a composite measure in criminology or social science.
Examples:
Index of disorder (graffiti, litter, abandoned buildings).
Fear of crime scale.
Socioeconomic status index (income, education, occupation).
Give an example of indicators and dimensions for the concept “fear of crime.”
Indicators: avoiding certain areas, installing security systems, anxiety levels.
Dimensions: behavioral, physical, and emotional aspects of fear of crime.
Is “crime seriousness ranked from 1–10” ordinal or interval? Why?
Ordinal. Rankings show order (higher = more serious), but the distance between ranks isn't equal.
If a bathroom scale gives you the same weight every time but it’s always 5 pounds too heavy, is it reliable, valid, both, or neither?
It is reliable (consistent), but not valid (inaccurate).
A scale measuring depression also correlates with related constructs with low self-esteem. What type of validity is this?
Construct validity = Measure aligns with theory and correlates with related concepts.
What is a typology, and how is it different from a composite measure?
Typology = classification system (categories based on multiple dimensions).
Composite = one score/scale built from multiple indicators.
Why is it important to have exhaustive and exclusive categories when defining variables?
Categories must be exhaustive so every case fits somewhere, and exclusive so cases don’t fit into more than one category. If not, results become inconsistent and meaningless.
True or False: You can take a nominal variable and treat it as ratio data. Explain.
False. You can reduce data (ratio -> ordinal/nominal), but you cannot increase precision. You cannot treat nominal data as ratio because it lacks order and a true zero.
Give one real-world example from criminal justice research where reliability could be a problem.
Examples:
Survey responses change depending on interviewer style.
Police officer estimating car speed by eye instead of radar gun.
Forensic lab errors (sample contamination, analyst bias).
Explain why a measure can be reliable but not valid.
A measure can be reliable but not valid if it's consistent but inaccurate (e.g., shoe size predicting intelligence).
Create a simple typology for "neighborhood types" using at least two dimensions (e.g., crime rate and economic resources).
Example typology:
High-crime, low-income neighborhoods
High-crime, high-income neighborhoods
Low-crime, low-income neighborhoods
Low-crime, high-income neighborhoods
Why does conceptualization matter for policy decisions? Give an example of how different definitions of “recidivism” could lead to different policy outcomes.
If recidivism is defined as rearrest, rates may look higher than if defined as reincarceration. Policymakers might wrongly conclude rehabilitation programs are failing if they use one definition versus another. Consistent conceptualization ensures fair policy evaluation.
Police record the number of times officers stop people in a city. Later, researchers group the numbers into “low, medium, high” categories. What is the original level of measurement? What level is it after recoding? Why does this matter for analysis?
Original = Ratio (true zero, counts). Recoded = Ordinal (ranked categories). This matters because ratio allows advanced statistics, but recoding to ordinal limits analysis (e.g., can’t calculate means).
Imagine two probation officers using the same risk assessment tool give very different scores for the same person. What type of reliability problem does this represent, and how might it be fixed?
This is an interrater reliability problem. Fixes include better training, clearer scoring guidelines, or using automated scoring systems to reduce subjective interpretation.
A researcher develops a new measure of “community safety” that includes crime rates, perceptions of safety, and neighborhood disorder. To test validity, they check whether this measure predicts property values and migration patterns. Which type of validity is being tested and why?
Criterion validity — the measure is being validated by seeing if it predicts relevant real-world outcomes (property values, migration).
Create a typology for drug offenders using at least two dimensions (e.g., drug type and level of violence involved). Explain why typologies are useful for research.
Example:
- Nonviolent marijuana offenders
- Nonviolent hard drug offenders
- Violent drug offenders
- Trafficking focused offenders
Typologies help organize complex populations into meaningful groups, making analysis and policy more targeted.