Sampling Methods, Errors, and Units of Analysis

Vocabulary flashcards covering core sampling concepts, probability vs non-probability sampling methods, sampling errors, fallacies, and structural diagrams from the lecture notes.

Population

The entire set of entities, such as individuals, cities, states, neighborhoods, or schools, in which a researcher is interested.

Sample

A subset of entities selected from a population that is studied to make conclusions about the entire population.

Representativeness

The extent to which a sample is similar to the population in all respects relevant to the study.

Generalizability

The validity of conclusions or generalizations that can be drawn about a population based on studying a sample.

Sample Generalizability

The ability to generalize sample results back to the specific population from which the sample was selected.

Cross-Population Generalizability

The ability to generalize findings from a study of one sample or population to another setting or population.

Target Population

The specific population to which a researcher wishes to generalize study findings.

Sampling Frame

The list of cases or units from which a researcher selects a sample.

Probability Sampling

A sampling method where each entity in the population has a known and non-zero probability of selection that is random and unrelated to the study variables.

Non-Probability Sampling

A sampling method where the probability of selecting each entity is unknown and systematic, potentially relating to the study variables.

Simple Random Sampling

A probability sampling method in which cases are randomly selected from a full list of entities in the population.

Systematic Random Sampling

A probability sampling method in which elements are selected by picking every $n\text{th}$ case from an unsorted list of population entities.

Stratified Random Sampling

A probability sampling method where the population is sorted into distinct groups (strata) and an exact number of cases is randomly selected from each group.

Proportionate Stratified Sampling

A stratified sampling method where elements are selected from each stratum so that the sample matches the exact proportions of those strata in the overall population.

Disproportionate Stratified Sampling

A stratified sampling method where elements are selected from strata in proportions different from the population, useful for analyzing smaller subpopulations.

Multi-Stage Cluster Sampling

A probability sampling method where elements are selected in two or more stages, beginning with the random selection of naturally occurring aggregate groups (clusters) and ending with random selection within clusters.

Clusters

Naturally occurring aggregate groups of elements, such as schools, cities, or states, used as sampling units in multi-stage cluster sampling.

Sampling Error

Differences in a value between a sample and the population that occur by chance in random samples and tend to cancel out across many samples.

Systematic Error

Differences between sample and population values resulting from non-random sampling methods or bias, which do not cancel out or decrease with larger sample sizes.

Convenience Sampling

A non-probability sampling method in which elements are selected based on availability or convenience, yielding a low likelihood of representativeness.

Quota Sampling

A non-probability sampling method using convenience sampling where quotas are set to ensure representativeness for specific chosen characteristics.

Purposive Sampling

A non-probability sampling method (also called judgment sampling) where each entity is selected based on the researcher's judgment of its unique position or purpose.

Snowball Sampling

A non-probability sampling method in which initial population members are interviewed and asked to identify other members, useful for hard-to-reach interconnected populations.

Respondent-Driven Sampling

A structured version of snowball sampling that uses financial incentives and recruitment rules across successive waves to produce a more representative sample.

Units of Analysis

The level of social life or entity about which data is collected and conclusions are drawn, classified as individual-level or group-level.

Reductionist Fallacy

An error in causal reasoning that occurs when conclusions about group-level processes are drawn based on data collected from individual-level units.

Ecological Fallacy

An error in causal reasoning that occurs when conclusions about individual-level processes are drawn based on data collected from group-level units.