Special forms of Non-probability sampling

Group Interview sample: Several specialized forms of non-probability sampling have achieved wide usage. One is the group interview sample, which is used on focus group studies. A ‘group’ is usually a quota sample of 5-10 consumers (possibly all product users, non-users and so on) assembled for a one to two hour joint interview by a person specially trained in group dynamics. Usually several such groups are conducted.

The number of ‘groups’ used is often small for two reasons: they are expensive and the information obtained from each group is highly unstructured because it is based on a simultaneous, free form interview with a number of people. Researchers usually learn as much as this method will provide from just a few group interviews. The number of respondents per group is restricted to about 10, because of the practical difficulties experienced in maintaining control over a larger group.

One study involved three groups of male tequila drinkers younger than 30, with six respondents per group. The groups were geographically widely dispersed to elicit a broad range of reactions. The interviews, which included tasting a new brand, lasted about two hours. Discussion subjects included brand usage and perceptions, frequency and circumstances of purchase and use, and reactions toward advertising. The respondents for this study were recruited via telephone screening and offered $25 each for their participation.

This kind of sampling is typically done in exploratory research to learn something about the range of consumer beliefs and practices with respect to the subject at issue. Each group is a ‘fishing expedition’, rather than a rigorous attempt to estimate the values of some universe. Used in this way, group samples can contribute importantly to the development of hypotheses. They should never, however, be regarded as providing evidence on the size of some universe value. One reason is the difficulty of defining what universe is sampled by this technique. A second reason is that each group interview is essentially a sample of one, consisting usually of a vaguely defined cluster of responses. In practice, it is common to vary the form and content of the questioning from group to group. The respondents within a group are interviewed at the same time and each person’s responses, therefore, are conditioned by those of the other respondents. Because groups are quota cluster samples, the data from four groups of five respondents each – for example cannot be regarded as the equivalent of 20 independent randomly chosen observations.

Shopping Mall Intercept Samples: Another widely used non-probability sampling technique is the shopping mall intercept sample. Respondents are recruited for individual interviews (using assigned ‘quotas’) at fixed locations in shopping malls. It is common in a given study to use several malls, each serving different socioeconomic populations, in each of several cities.

Such samples are frequently used in moderate size studies (100-200) that are experimental in character, for example, to compare responses to two or more TV commercials, two or more product concepts and so forth rather than to estimate the parameters of a single existent universe. Mall samples can be informative or such purposes, assuming that the differences in experimental results do not depend importantly on the characteristics of the sample. If, for example the difference in effectiveness of two commercials varies with frequently of mall shopping, demographic characteristics of mall shoppers versus non-shoppers, or any characteristic related to availability for mall interview, then mall samples should not be used.

All sample advantages are speed and economy relative to a probability sample of personal interviews. Their limitations derive from the expediency of selecting the sample and then a typical character of the population sampled (i.e. mall customers).