Sampling SocResearch Test2

Likely sampling strategies used by interviews

Non-probability strategies (i)

Likely sampling strategies used by surveys

Probability strategies like convenience

Likely sampling strategies used by field research

Non-probability sampling strategies

Definition of non-probability sampling

Sampling without any “random selection” processes of settings or groups of people; inductive approach, NO statistical inferences; qualitative

Definition of probability sampling

Sampling based on “random selection” processes of settings or groups of people; deductive approach, statistical arguments; quantitative

When should non-probability sampling be used?

No sampling frame exists, target population is hidden/unknown/rare, when testing a DV what are possible IV’s, propose/develop theory

When should probability sampling be used?

When examining large populations like large-scale surveys, census studies, or comparing different social groups, test if an IV affects DV, and assess to what extent theory plays out

What are types of non-probability sampling?

Convenience, quota, purposive, theoretical, and snowball sampling

Definition of convenience sampling

Selection of cases that are conveniently available.

Definition of quota sampling

Getting subgroups to match % of sample with some characteristic, to % with that characteristic in the population (ex: gender)

Definition of purposive/judgemental sampling

Sampling that involves the careful and informed selection of typical cases or of cases that represent relevant dimensions of the population to research

Definition of theoretical sampling

As data collection and ongoing analysis progresses, focusing on cases that help to inform developing theory

Definition of snowball sampling

Asking each participant to refer or introduce you to additional participants.

What are types of probability sampling?

Simple random, stratified (disproportionate and proportional), cluster sampling, and multistage cluster sampling

Definition of simple random sampling

Sampling design where every case and every possible combination of cases has an equal chance of being included in the sample; prob

Definition of stratified sampling

Selecting cases by first dividing population into strata (or categories) and drawing random samples from each stratum; prob

Definition of proportionate stratified sampling

Cases are selected from each strata proportionate to their presence in the population presence

Definition of disproportional stratified sampling

Cases are selected from each strata disproportionate to their presence in the population, and “weighting” corrects for it in statistical analysis

Definition of cluster sampling

Probability sampling design in which the population is broken down into natural groupings or areas, called clusters, and a random sample of clusters is drawn

Definition of multistage cluster sampling

Cluster sampling design in which sampling occurs at two or more stages.

When to use convenience sampling?

Quick, low-cost studies

When to use quota sampling?

Ensuring representation of important subgroups in surveys or beginning stages of field research/in-depth interviews

When to use theoretical sampling?

Exploratory research where theory is built during data collection

When to use purposive sampling?

In-depth qualitative research on specific experience

When to use snowball sampling?

Studying hidden, hard-to-reach groups (e.g., undocumented immigrants).

When to use simple random sampling?

Generalizable surveys and experiments.

When to use stratified sampling?

Comparing different groups (e.g., racial, income, gender differences); studying a diverse population with subgroups that must be accounted for separately.

What is strata?

Based on variables or characteristics you are interested in to ensure collection of enough cases

When to use disproportionate stratified sampling?

Need more cases from smaller subgroups to ensure valid comparisons.

When to use proportionate stratified sampling?

Want to maintain real-world proportions while ensuring each group is represented.

When to use cluster sampling?

Probalistic large-scale research when individual selection isn’t feasible.

What are clusters?

Natural groups based on geography, community, or administrative units to reduce costs

When to use multistage cluster sampling?

Nationwide or regional surveys (e.g., selecting cities, then schools, then students)

Strengths/weaknesses of convenience sampling?

Fast, easy, inexpensive, BUT high risk of bias and has no basis for statistical generalization

Strengths/weaknesses of quota sampling?

Really no strengths, it can be biased and is fake probability sampling

Strengths/weaknesses of theoretical sampling?

Follows emerging pattern, but requires continuous analysis

Strengths/weaknesses of purposive sampling?

Focuses on relevant participant for rich data, but can have researcher bias and not generalizable

Strengths/weaknesses of snowball sampling?

Helps access closed-off groups, but has high potential for bias and lacks randomization

Strengths/weaknesses of simple random sampling?

Reduces bias and representative, but requires full population list which may not be possible and may miss some cases you’re interested in

Strengths/weaknesses of stratified sampling?

Ensures representation of key groups, reduces sampling error, BUT needs a good sampling frame for each and every strata (must know characteristics in advance)

Strengths/weaknesses of disproportionate stratified sampling?

Ensures representation of small groups, improves precision for marginalized populations, BUT not representative of real world proportions

Strengths/weaknesses of proportionate stratified sampling?

Maintains real world accuracy, minimizes sampling bias, BUT have fewer cases to analyze subgroups and overlook disparities

Strengths/weaknesses of cluster sampling?

More practical for big populations, BUT less precise and may not represent whole population

Strengths/weaknesses of multistage cluster sampling?

Reduces cost and effort while maintaining randomization, BUT less accurate

Definition of coverage error

Error introduced by a faulty sampling frame that fails to match your target population sampling frame

Definition of selection bias

Favors certain cases or if the selection of one case increases or decreases the likelihood that another case will be selected

Sample size criteria for non-probability samples

Usually small amount of cases until saturation is reached

Methods to reduce sampling error

Increase sample size and minimize homogenous populations

How to calculate sampling error

Using the sampling distribution of sample size; “average distance of sample values from the population value” in a sampling distribution of a given sample size

Definition of sampling error

Difference between an actual population value (e.g., a percentage) and the population value estimated from a sample; both prob and non-prob.

Definition of sampling frame

Probability operational definition of the population that provides the basis for drawing a sample; ordinarily consists of a list of cases.

Definition of sampling distribution

Distribution of the probabilities for a variable, which indicates the likelihood that each category or value of the variable will occur.

Relationship between sample size, sampling distribution shapes, sampling error

As sample size increases, the sampling distribution narrows becoming more bell-shaped (normal), reducing sampling error, which improves the accuracy of probability sample estimates by decreasing variability

Importance of sample size vs. population size for sample precision/sampling error

A large sample size with sample precision can reduce standard error and ensure representation of population size

Methods to address non-response bias

Pursue sample of non-respondents, conduct analysis of last-minute respondents, and compare characteristics with broader population