Omnibus studies

An omnibus study (or survey) is an ongoing study through which a buyer can ask a limited number of questions at a very reasonable cost, because the general overhead costs of fielding the study are shared by a number of clients. Omnibus studies are often scheduled on a weekly, monthly or quarterly basis. Some are conducted through personal interviews but most omnibus services collect data by telephone interviews, so it is impossible to show display materials such as advertisements or products.

Omnibus studies assist in finding small segments of the population and the costs of locating a sufficient number of such respondents are low. If the incidence is too low to produce a sufficient number in one survey, the questions are retained in subsequent ones at a modest cost. Most omnibus services offer follow-up probes of those giving specific replies. These probes are usually conducted on a follow-up basis rather than at the time of original questioning (unless there are only a few questions) so as not to delay completion of fieldwork for the standard service.

Omnibus research is a hybrid between off-the-peg research, where data have already been collected and the researcher simply purchases whatever is available, and primary research, where the researcher collects data first hand. The term describes regular research surveys undertaken with a certain frequency and definite method, which use a set number of respondents and sampling points.

Omnibus research is particularly suitable for fairly robust data - for example, the usage and purchase of products - but it is less useful for attitudinal and opinion studies. Typical omnibus surveys might last for 40 minutes or more and during that time the respondent might be asked about six or more entirely different topics. The superficiality of some questions can prevent respondents from gaining the amount of concentration necessary to pursue any particular topic in depth. The Market Research Society Newsletter provides a list of organisations offering omnibus services.

Most consumer omnibus surveys reach at least 1000 adults per wave. This provides a substantial sample size for most purposes of analysis. It is also possible to increase the sample size by running questions over several waves of the omnibus. In the case of the more specialist omnibus survey, the sample size is much smaller (often around 100-200 respondents per wave) and they tend to be conducted less frequently.

Omnibus surveys employ one of three sampling designs. The most common approach uses a random location - with quotas set for age and working status within gender - although there are surveys utilising pre-selected and quota sampling. The standard output of most omnibuses is computer tables, although many suppliers provide data in various formats.

Continue reading here: Semantic differential scale

Was this article helpful?