The first step in a marketing research process is to define the problem chosen for investigation. This step is a very significant one since it is said, “A problem well defined is half solved.” On the other hand, if the problem is defined vaguely, a wrong problem is defined, or the uses of research are not clarified, then the research results may be completely useless for the management and research effort of the investigator is a futile exercise.

Generally, it has been observed that poor definition of research problem is the result of faulty assignment of problem by the top management. For example, if the top management of company is interested in investigating the reasons for declining sales, the marketing manager may ask the marketing research executive to investigate everything regarding the sales. The marketing manager, in the problem assignment has not spec tined which aspects (viz., production planning. pricing, distribution (channels, advertising, personal selling etc.) are to be studied. The marketing manager asks the researcher to study all aspects affecting sales. But all the aspects may not be resonator for the decline in sales. Therefore, the marketing manager should have specifically told the investigation that the latter should study product planning, pricing policies, distribution policies, promotion policies, or other policies of marketing management affecting the sales. Thus before assigning it to the marketing researcher, there problem at hand must be define clearly. It may be reiterated that a well defined problem leads to better solution.

In order to define a problem properly, we should determine the nature of research. The problem illustrated above is of exploratory nature, the major emphasis of which is on gathering new ideas or insights into it. After understanding the nature of the broblem a preliminary analysis may be carried out applying the techniques. viz., (i) situation analysis; and (ii) informal investigation. Unfortunately, the research situation is often ignored. But the situation analysis is virtual to obtain background information on the problem. Situation analysis means the circumstances under which the research is being conducted. Situation analysis is useful to take note of the factors affecting the marketing operations in a business organization. Six categories2 of information can be used in preliminary investigation ; (a) the product, (b) the company , industry and competition, (c) the market, (d)  the channel of distribution, (e) the sales organization, and (f) the advertising and sales promotion policies. For example, if Kirloskars have to find out reasons for the declining sales of their diesel engines, they must analyze the problem situation in view of above mentioned factors.

Informal investigation refers to the discussion with a few selected customers, dealers, top management personnel of the company, and other particles concerned with the problem. The information investigations may be designated as the pilot survey. Generally he problem and proposed techniques to solve it are thoroughly discussed with a panel of experts and respondents. The changes corporate in research proposal. For example, in the problem declining sales, we may select a few academicians, practitioners, customers, dealers, competitors, sales staff etc., and discuss the research proposal with them. There can be two purposes to carry out the informal investigation; (i) to develop and select the hypothesis to be used in the final study; and (ii) to estimate the complexity involved in marketplace problems.

Although the situation analysis and informal investigation are simple activities, yet they are of critical importance in the research process. These activities help in defining the research problem entire research exercise. Therefore, it is creditable to mention at this point the researcher should attentively carry out these research activities.

Once the problem is defined, the next step, that is the research design, becomes easier. Research design is the basic framework which provides guidelines for the rest of research process. It is a map or blueprint according to which the research is to be conducted. The analysis. The researcher specifically pinpoints that to carry our research properly; (i) how would the data be collected , (ii) which instruments for data collection would be sued, and (iii) what sampling plan would be used? The researcher has to carefully decide and make a choice from the group of different alternatives available to him.

There are two types of data: secondary and primary. The secondary data refer to those data which are gathered for some other purpose and are already available in the firm’s internal records and commercial, treade, or government publications (For details on secondary data see Chapter6). On the other hand, primary data do not exist already in records and publication. The researcher has to gather primary data afresh for the specific study undertaken by him. The primary data re explicitly gathered for a specific research project at hand.

It is always in the fitness of things that the researcher attempts to look into sources of secondary data before starting to collect primary data. All too often, it has been observed that firms (researchers) had wasted a lot of time and money on collection of primary data, while the secondary data were already available in time. and money can be saved if secondary data is first searched and used of the study at hand.

Of course, it is essential to collect primary data if the secondary is not sufficiently available. The primary data can colleted by three methods: (1) observation, (2) experimentation, and (3) surveys. Let us explain each or these methods in some detail in the underlying text.

1.    Observation: In observation method, we observe the action of the respondents, either directly missing with them or indirectly without getting mixed up with them. For example, in the case of declining sales of diesel engines, Kirloskar’s investigators may go and mix up with the farmers using the engine. After reveling the purpose of study to the respondents, enquiries are made and data are collected. In indirect observation method researcher does not are reveal the purpose of study and observes the respondents as a stranger. For example, if we are interested to know the shopping habits of the people, we can carry out observation of buyers in a general store. In direct observation, we tell the purpose of study to people being investigated and get informally mixed up with themes, as a buyer. Whereas in indirect observation, we do not reveal the purpose of investigation to the respondents. We tell it to the owner of shop and sit down silently to note the purchasing behavior of people. For example, influence of accompanying children on shopping patterns of a housewife can be observed in a general store.

2.    Experimentation: The experimentation method emphasizes the creation of a controlled environment where some variables are allowed to vary and cause-and-effect relationship is tidied. For example, in our preceding case of declining sales of Kirloskar diesel engines, we may consider that advertising used is not effective enough and is not able to stop the declining sales. We may conduct an experiment. For this , we may take three equally matched markets: Delhi, Chennai and Mumbai. In one market, we may doable the advertising expenditure, in second we may reduce it to half and in third we may keep it constant (normal). After the stipulated time period we may record the sales effect of advertising keeping all other variables constant (controlled). On this basis we may decide what action we should take to make advertising more effective.

3.    Survey: One of the most common and widely used primary data collection methods is the survey. With the use of surveys, we can gather a wide range of valuable information on overt behavior of the consumer, viz., attitudes, motive, and opinions, the earlier two methods (observation and experiment) are not capable to yield such information. The primary data, using any of these three above-mentioned methods, can be collected from any population comprised any type of respondent groups like customers, middlemen, competitors, salesmen, company employees, etc.

The researcher while collecting the information, is looking for, a good research design and suitable research instruments. When observation method for data collection is used, the researcher may make use of instruments, viz., cameras, tape recorders, VCRs, tally sheets, etc. In experimentation method also the same instruments may be useful. Whereas in survey method, questionnaire is the instrument most frequently used and yields the most satisfactory results. Though we have dealt with the construction of questionnaire in greater detail in a later chapter of this book yet, it is creditable to mention here that a great care is required to be taken while dealing with certain aspects of questionnaire such as: (i) types of questions to be asked, (ii) number and form of questions, (iii) wording of questions, and (iv) sequencing of questions. The construction of a good questionnaire is an art and takes considerable time for a professional researcher.

First step in the sampling plan is to decide the universe or population (the whole collection of items to be studied by the researcher). Once the universe is decided, the researcher must concern himself to find; the answers to the following four questions:

(i)    What sampling unit should be studied?
(ii)    What should be the sample size?
(iii)    What sampling procedure should be used?
(iv)    What contact method should be used?

Sampling unit: Perhaps the basic thing to be decided is the sampling unit (who is to be surveyed?). For example, in a survey on the attitudes towards the use of toilet soaps with reference to a specific brand; whiter husband, wives, children, or a combination of all of them is to be surveyed?

Sample size: The second issue to be decided is the sample size (how many units to be surveyed?). Of course, the whole of the universe cannot be studied in a single research project. The researchers has to select a relevant fraction of the population which is a representative of the entire population or universe. Generally, in exploratory study a sample less than 1 per cent is sufficient to provide reliable results. Whereas in motivation research, a sample of fewer than thirty respondents for indepth interview is usually sufficient to reveal the significant aptitudes or a large group of consumers. In any situation , the sample size must not be increased more than 10 per cent of the universes.

Sampling procedure: How to conduct the survey? It is the next step in sampling plan. Sampling procedure depends upon the research objectives to be accomplished through the investigation. There can be choice between probability and non-probability sampling producer may be justified. However, for an accurate estimate of the population characteristics a random or probability sampling procedure may be used because it lends itself or the application of various statistical tests. It depends upon he nature of the research project and accuracy demanded that whether the probability or non-probability sampling procedure is better. The researcher must carefully weigh the two available alternatives while deciding about the sampling procedure.


Contact Method: The last but not the least issue to be decided in the sampling plan is to determine the contact method (media) (How the selected units be approached?). The sampling method or media refers to contact methods, e.g., telephone surveys, mail surveys and personal interviews. Each of these methods has got its merits and demerits. However, the most commonly used methods in the Indian conditions are mail surveys and personal interview. The personal interviews, though costly, yield the most satisfactory results in a research work. In India, the telephone survey does not yield good results because many homes do not have telephones and respondents do not reveal detains on the telephone owing to their personal security reasons.

Once the researcher has finalized the problem definition and research design, he must conduct the actual data collection operation. This step is also called fieldwork. Fieldwork is the most expensive of all the steps in a research project. Also the fieldwork is more prone to errors. The researcher may face a number of problems during the fieldwork. The most common problems during a fieldwork are as follows:

The people (sampling units) whom the researcher wants to interview during his visit may not be available at home. The researcher in such a situation may either call back later or substitute another house just adjacent or next door. The latter alternative is better because the former is very expensive. But whether the subsititituted house is just replica of the sampled house, is a question which puzzles the marketing research experts. It is generally accepted that a bias may intrude into the results of the research owing to substitution.

The researcher may find the respondents at home but he may not obtain the desired co-operation from them. If the researcher is carefully selected and trained, he amy use his tactics to persuade the respondents.

The respondents may decline to cooperate. The researcher must establish a rapport with the respondent to avoid this. Even if they cooperate, respondents may sometimes give blazed or misleading answers to the questions asked by the researcher. They do so either to finish the interview quickly or for not revealing the actual personal details. The researcher must establish a good rapport by talking informally on other issues of mutual interest first and encourage benefit both parties. And assuring hem that nothing will be made public and data given by them will be used purely for academic purposes only.

The interviewers may also introduce different type of bias in the answers owing to mere fact of their age, sex, manner, or intention. In addition, there may be conscious bias created by the interviewer owing to his dishonesty. Every interviewer tries to finish his quota of interview as quickly as possible and also at a cheapest possible cost. In doing so, he may falsify the interviews. He may fill up all his questionnaires by himself sitting at his own residence claiming not-at-homes or refusal to co-operate.

All these fieldwork problems can minimized or eliminated even, if we select, train, compensate, control, evaluate, and maintain the field force though a carefully designed method. The research director can check on telephone or y personal visit to a few of the persons claimed to have been interviewed by the researcher. If the interviewers are told beforehand that such checks will be conducted to test their sincerity, the interviewers may avoid the intentional falsification of interviews.

In order to extract meaningful information from the data collected the data analysis interpretation is carried out. The data analysis is the fourth step in the research process. The data are first edited, coded and tabulating for the purpose of analyzing them. The editing coding, and tabulating is a must when the interviewer has amassed a huge amount of data concerning the research project at hand.

The analysis is basically aimed at giving inferences of association of differences between the various variables present in the research. The analysis can be conducted by using simple statistical tools like percentages, averages and measures fo dispersion. Alter natively, the collected data may be analyzed by using simple statistics graphs, charts, pictures, etc. Data may be cross-tabulated to produce useful relationships among the variables involved. To conduct a more sophisticated analysis, correlation or egression analysis may be used. Furthermore, various statistical tests, t,F, Z, X2 etc., may be applied. The most complicated and sophisticated analysis is to attempt multivariate analysis on the available data by using advanced statistical techniques, viz., multiple-regression analysis, multiple correlation analysis, discriminate analysis, factor of data may be extracted from the analysis thus conducted. The conclusion, summary and recommendations of research are based on the statistical analysis and inferences drawn.