Four major elements are involved in undertaking marketing research. The first element is a preliminary investigation. This initial study permits the researcher to determine the purpose and scope of his research as well as to identify tentative questions.
Creating a research design to test the questions is the most important and most complicated aspect of marketing research. It commences with the selection of the approach to be taken. The three most commonly used are the experimental, the observational, and the survey approaches. Any given project may use one or more of the three.
It is also necessary to determine the types of data that will be needed to solve the marketing problem and to locate sources where this information can be obtained. Data sources are generally classified as either primary or secondary. Secondary data are made up of previously collected information and are obtained from historical records, publications, government documents, and the like. Primary data are gathered for the first time. The survey method is probably the most frequently used method for collecting primary data. Data are gathered by mail, by telephone, by personal interviewing, and online.
Another critical aspect of most marketing research projects is the selection of the sample. A probability sample involves the selection of respondents in such a way that every unit in the pool has the same chance of being selected. One method of drawing a probability sample is by the use of a table of random digits. A nonprobability sample is drawn on a judgmental basis; respondents are selected because they are considered to be representative of the group from which they are drawn.
The final aspect of the research design is the anticipation of the results and the decision as to how the data will be summarized and reported. It is becoming more and more common in large marketing research projects to make use of a computer for the processing and tabulation of the research results. Some problems usually arise, however, and careful supervision and control of the data-collection activities are important. It is particularly critical to guard against various kinds of survey bias that can creep into a project.