203-205
Introduction to Sampling in Criminal Justice Research
Importance of data collection in criminal justice research.
Research value depends on how data is collected.
Key considerations:
What will be observed and what won't.
Selection criteria for specific populations (e.g., drug users).
Overview of the logic of sampling and approaches for selecting subjects.
What is Sampling?
Definition: Sampling involves selecting observations from a larger population to study.
Two main reasons for sampling:
Practicality: Collecting data from all individuals is often impossible.
Example: Assessing proportions of drug users among arrested individuals in U.S. cities.
Efficiency: Not necessary to survey an entire population; can generalize from a sample.
Example: Surveying a sample of high school students regarding marijuana use can suffice instead of polling all students.
Probability Sampling Techniques
Central to criminal justice research but not always applicable.
Nonprobability sampling can be used in certain situations.
Goal: Reduce or understand bias in selecting subjects.
The Logic of Probability Sampling
Definition: A sampling method where each population member has a known chance of being selected.
Helps in generalizing findings from observed to unobserved cases.
Sampling process ensures representative samples that reflect the larger population.
Issues with Homogeneity and Heterogeneity
Identical populations require less rigorous sampling techniques.
Example: A single case can represent a uniform population.
Real populations are heterogeneous:
Variation in demographics, attitudes, experiences.
Needs to be captured in samples for accurate representation.
Sampling Bias
Definition: Occurs when selected subjects are not representative of the larger population.
Types of biases:
Conscious Bias: Researchers intentionally select a non-representative group.
Unconscious Bias: Researchers may select conveniently located individuals, skewing results.
Example: Interviewing 100 lawyers at a courthouse may favor those who frequently visit, not representing the whole.
Assessing Representativeness
Self-selection in surveys (e.g., blogs, email surveys) can lead to biased results.
A representative sample closely matches the characteristics of the population.
Example: If the population has 50% women, the sample should also reflect that proportion.
Limitations on representativeness depend on the study's substantive interests.