When conducting an evaluation, it is critical to select a sample of participants that are representative of the population you wish to study.
A population is the entire group of people, items, or units of analysis that you intend to study (e.g., classrooms, schools, programs, etc.).
- A sample is a subset of a particular population.
- Samples are representative when they provide an accurate reflection of the variations and diversity represented within a population.
- If a sample is representative, it can be assumed that the results of the evaluation are generalizable or applicable to the greater population (Krathwohl, 1998).
There are two main types of sampling methods (i.e., probability and non-probability) and several techniques that can be used to select a sample for your evaluation.
- The method and technique you choose should depend on your research questions, resources, and desired level of accuracy.
- You should review your options carefully and select the sampling technique that is most compatible with your evaluation purpose, design, and resources (Galloway, 1997).
Probability sampling involves the random selection of study participants in a manner that gives each member of the population an equal chance of being selected for the sample.
- The benefit to using probability sampling is that your sample will be fully representative of your population, and your results will be generalizable to the population.
- The disadvantage to using probability sampling techniques is that they can be very time-consuming and costly.
Types of Probability Sampling
Simple Random Sampling – This technique gives all units in the population an equal opportunity of being selected by using a method that will select units completely at random (Krathwohl, 1998).
- Simple random sampling methods that are commonly used in evaluation for selecting units include using a basic lottery system or drawing numbers/names from a hat. Both methods are very effective if you have a small or moderate sized population.
- If the population is large you may want to consider using sampling software to select your sample or a random numbers table (usually found in statistics books).
Stratified Sampling – This technique divides the population into meaningful homogenous or similar groups based on a certain characteristic (e.g., gender, race, socioeconomic status) and then selects a simple random sample from each group. [For example, if you were interested in the affects of student motivation on academic achievement, particularly by grade level, you would divide the population into their respective grade levels and then randomly select an equal number of 9th, 10th, 11th, and 12th graders.]
- Stratified sampling ensures greater representativeness on a characteristic of interest within the population.
- This method is a more commonly used probability sampling technique than simple random sampling because it allows you to study wider range of the population without a larger sample size.
Nonprobability Sampling – This technique doesn’t use random sampling at any stage of the selection process, so some members of the population may have a greater chance of being selected (Krathwohl, 1998).
When using nonprobability sampling techniques the potential for bias is high and applicability of results is low because the sample is made up of units that were selected specifically by the evaluator, referred by others, or were simply convenient or available at the time.
- The benefits, however, to using nonprobability sampling techniques are that they are less resource intensive and are generally more accessible and convenient to work with.
- Nonprobability sampling is less rigid than probability sampling and has a wider range of techniques that can be used to select your sample.
- Five commonly used nonprobability sampling techniques in evaluation are (Galloway, 1997; Krathwohl, 1998):
Types of Non-probability Sampling
Convenience Sampling – This technique uses whatever units from the population that are available to participate at a given time.
- Convenience sampling is often called grab sampling.
- This technique has very little structure. The only criterion for selection is that the unit you select is a member of the population and is available to participate in the study at the time required.
- For example, using a convenience sampling technique, you would select the first 100 people who received a copy of your program’s college planning handbook to take a survey on its user-friendliness.
Purposive Sampling – This technique is conducted by allowing the evaluator to select a sample that he/she feels is representative of the population.
- Purposive sampling relies on your knowledge as the evaluator and the design of the program to choose the most appropriate and representative sample.
- You must subjectively choose units which you believe are representative of your population and try to ensure that the full spectrum of variation and diversity of your population is represented in the sample.
Quota Sampling – This technique predefines specific groups within the population to investigate and then samples from the population to fill a quota for each group.
- Quota sampling is similar to stratified sampling but does not use random sampling to choose units from each group.
- The goal, however, is to still choose a sample that closely matches the population ratio/breakdown on a characteristic or set of characteristics of interest (e.g., age, gender, race, etc.).
- You should continue to select units to participate in your sample until all groups are filled (note: you may fill your quota for some groups earlier than others).
Snowball Sampling – This technique uses the process of referrals to build the sample size.
- With snowball sampling you begin with a few key individuals who you would like to include in your sample.
- You ask them to participate in the study, as well as recommend other people they might know with similar characteristics to the population you are studying that could also participate in the study.
- Then you contact the individuals you were referred to and ask them to recommend others for the study as well, thus growing your sample size through referrals.
Self-selection Sampling – This technique allows participants to volunteer and decide that they would like to be part of the study sample.
- Self-selection sampling is usually done by advertising the study and asking for volunteers.
- This allows potential participants to contact you and to volunteer to participate in the study.