Statistical Measures
This lesson explores the different sampling techniques using random and non-random sampling. The lesson also distinguishes among different study techniques. In addition, it provides simulations that compare results with expected outcomes.
Types of Probability Sampling Review
Probability and Non-Probability Sampling
A population includes all items within a set of data, while a sample consists of one or more observations from a population.
The collection of data samples from a population is an important part of research and helps researcher draw conclusions related to populations. Probability sampling creates a sample from a population by using random sampling techniques. Every person within a population has an equal chance of being selected for a sample. Non-probability sampling creates a sample from a population without using random sampling techniques.
KEEP IN MIND
Probability sampling is random, and non-probability sampling is not random.
There are four types of probability sampling. Simple random sampling is assigning a number to each member of a population and randomly selecting numbers. Stratified sampling uses simple random sampling after the population is split into equal groups. Systematic sampling chooses every nth member from a list or a group. Cluster random sampling uses natural groups in a population: the population is divided into groups, and random samples are collected from groups.
Each type of probability sampling has an advantage and a disadvantage when finding an appropriate sample.
| Probability Sampling | Advantage | Disadvantage |
|---|---|---|
| Simple random sampling | Most cases have a sample representative of a population | Not efficient for large samples |
| Stratified random sampling | Creates layers of random samples from different groups representative of a population | Not efficient for large samples |
| Systematic sampling | Creates a sample representative of population without a random number selection | Not as random as simple random sampling |
| Cluster random sampling | Relatively easy and convenient to implement | Might not work if clusters are different from one another |
There are four types of non-probability sampling. Convenience sampling produces samples that are easy to access. Volunteer sampling asks for volunteers or recommendations for a sample. Purposive sampling bases samples on specific characteristics by selecting samples from a group that meets the qualifications of the study. Quota sampling is choosing samples of groups of the sub-population.
Examples
Census, Surveys, Experiments, Observational Studies
Various sampling techniques are used to collect data from a population. These are in the form of a census, a survey, observational studies, or experiments.
A census collects data by asking everyone in a population the same question. Asking everyone at school or everyone at work are examples of a census. A survey collects data on every subject within a sample. The subjects can be determined by convenience sampling or by simple random sampling. Examples of surveys are asking sophomores at school or first shift workers at work.
In an observational study, data collection occurs by watching or observing an event. Watching children who play outside and observing if they drink water or sports drinks is an example. An experiment is way of finding information by assigning people to groups and collecting data on observations. Assigning one group of children to drink water and another group to drink sports drinks after playing and making comparisons is an example of an experiment.
KEEP IN MIND
A census includes everyone within a population, and a survey includes every subject of a sample. An observational study involves watching groups randomly, and an experiment involves assigning groups.
Example
Simulations
A simulation enables researchers to study real-world events by modeling events. Advantages of simulations are that they are quick, easy, and inexpensive; the disadvantage is that the results are approximations. The steps to complete a simulation are as follows:
- Describe the outcomes.
- Assign a random value to the outcomes.
- Choose a source to generate the outcomes.
- Generate values for the outcomes until a consistent pattern emerges.
- Analyze the results.
KEEP IN MIND
A simulation is only useful if the results closely mirror real-world outcomes.
Examples
Let’s Review!
- Probability (random) sampling and non-probability (not random) sampling are ways to collect data.
- Censuses, surveys, experiments, and observational studies are ways to collect data from a population.
- A simulation is way to model random events and compare the results to real-world outcomes.
Subscribe to the online course to gain access to the full lesson content.
If your not ready for a subscription yet, be sure to check out our free practice tests and sample lesson at this link