Sampling

Sampling

Why should we draw samples?

There are 2 main reasons for sampling.
1.) It is impossible to study the entire population (usually); we also have finite resources so financially cannot study the entire population.
2.) It is unnecessary to study entire population because of probability.

What is the goal of sampling?

The goal of sampling is that it is representative, meaning the sample looks like the population from which it is drawn.

Vocabulary

population- the set of individuals or other entitites to which we wish to be able to generalize our findings

sampling frame- list from which population is selected (you may not always be able to have a sampling frame, example: the homeless)

population parameter*-

sample statistic*-

*It is important that the population parameter and the sample statistic are similar in order to have representativeness.

Representativeness- degree to which the sample accurately reflects the population from which it is drawn; how much your sample looks like the population; if your sample is representative, you can generalize the findings to the population

Sampling error- any difference between the characteristics of a sampl eand the characteristics of the population from which it was drawn; the larger the sampling error, the less representativ the sample is of the population.

Probability samples

A probability sample must:

1.) be able to calculate the likelihood of any single unit being selected (it doesn't need to be an equal amount... i.e. you may over sample... this would mean chances are not equal)
2.) the likelihood of being selected must be greater than 0 (probability runs from 0 to 1, that is, 0% to 100%)

Most of the time, a larger sample is better, but this is trumped by the process by which it is selected.

Randomization
- random selection refers to sampling
- random assignment refers to research design


Different types of sampling

   *Probability (SRS-SS-SRS-SRS-CS)*

1.) simple random sampling
each element has an equal chance of being selected
can be done with or without replacement

2.) systematic sampling
ex: taking every nth case

3.) stratified random sampling
relies on the distribution of characteristics of a population (often demographic characteristics such as age, gender, and race) in order to select a sample that is representative of those characteristics

4.) cluster sample
sampling in which elements are selected in 2 or more stages

   *Non-probability (P-C-Q-S*

1.) purposive
2.) convenience/availability
3.) quota
4.) snowball

We cannot generalize from non-probability studies.

What are pros to probability sampling?

Probability sampling reduces bias and allows us to calculate sampling error.

Reingle, et al

- obtained data from The Add Health Data
- looked at association between marijuana use and intimate partner violence
- unique because it uses a nationally representative sample of adolescent youth and follows them to adulthood (15-26); also unique as first study to evaluate the longitudinal effect of marijuana on intimate partner violence
- they hypothesized that consistent marijuana users are at greater risk of being both perpetrators and victims of IPV as opposed to non-users
- Dependent variable - IPV
- Independent variable - marijuana use

Results:

- substantial evidence for overlapping victimization/perpetration among marijuana and intimate partner violence
- marijuana use has a significant effect on being both a victim and perpetrator
- impacts legalization

Theory:

- none is given in the article!