Group Characteristics Sampling

Group characteristics sampling focuses on discovering something meaningful about the nature of a group, or just a particular narrow stratum within a group. It is meant to demonstrate that a particular challenge (or success factor) exists regardless of other characteristics. It can also be used to demonstrate that a phenomenon exists by demonstrating its existence within a small randomly chosen sample.

Types of group characteristics sampling include:

• Maximum-variation sample: Selecting a wide range of cases to demonstrate a pattern that cuts across an entire population
• Homogenous sampling: Selecting only cases within a narrow range of variation due to their importance as an area for examination or due to a lack of pre-existing knowledge about a specific subgroup
• Typical cases: Using typical cases to help us understand challenges and difficulties experienced in most cases
• Key informants: Selecting individuals that, due to experience and training, have better-than-most understanding of a particular issue or problem
• Complete target population: Conducting an observation of every single case that meets a specific set of criteria
• Purposeful random sample: Using a small random sample to add credibility to findings by reducing selection bias (not to be confused with a generalizable sample)

Maximum-Variation Sample

With maximum-variation sampling, auditors select cases that represent the full variation that exists in the population: the cases are purposefully as different from each other as possible. The goal is simply to have at least one example of each type of case that exists in the population. This type of sample is useful because it vividly demonstrates any similarities across cases despite the obvious heterogeneity. This type of sampling is useful for examining large national or global programs.

Homogenous Sampling

As the name implies, a homogenous sample includes cases that show very little variation. Homogenous sampling is used to focus on a single, but important, subgroup within a population. This group might be new and emerging, and the nature of risk and controls associated with it may be unique. Alternatively, it may be a subgroup that represents a high-risk area and requires attention.

Typical Cases

Typical cases are best used for illustrative purposes. An examination of typical cases will help an audience understand situations they would otherwise not be knowledgeable of.

Key Informants

A key informants sampling method involves sampling from a population of individuals. Key informants, by virtue of education, training, and/or experience, have a better-than-most appreciation and understanding of the situation or program of interest. Key informant evidence is most powerful when independent expert opinions converge and describe problems and potential solutions in a similar manner.

Generic Example #3: Key Informants Case

Complete Target Population

A complete target population method is ideal in a situation where an unexpected and new incident occurs and either affects or is related to a limited number of cases. In a situation such as this, all affected or related cases are examined.

Purposeful Random Sample

The unbiased nature of a randomly selected sample can provide powerful and persuasive evidence. While the sample sizes are usually too small to be considered generalizable, these samples can demonstrate the existence of pervasive problems. This method is also useful in combination with other methods to increase the level of objectivity and validity of findings.

Generic Example #4: Purposeful Random Sample

• PreviousComparison-Focused Sampling
• NextMixed, Stratified, and Nested Sampling Methods