This example illustrates an extremely rewarding method of collaboration between psycho-sociological specialists and econometric specialists. Such collaboration should be developed more end more in the study of many forecasting problems. In this connection, the rule to be followed in approaching the problems appears to be the following: where statistics are available (trade statistics, company sales, or the results of consumer research, for example), it seems advisable to make a preliminary statistical and econometric analysis which comes up with significant results, warns about some peculiarities, and suggests some explanations. This is where the psychologist comes in on already cleared ground where he may better âcenter" his work. Lastly, collaboration between the two approaches should occur to establish a synthesis of the results. On the other hand, where there are practically no statistics, they must be reconstituted by investigation. However, before undertaking it, is extremely useful for the psychologist to make a preliminarily exploratory study of the problems, whereby he can establish the fundamental hypotheses. Then an extensive investigation may be perfected which permits the psychologist to verify his hypotheses, and supplies econometric specialist the quantified elements which he needs.
This example illustrates an extremely rewarding method of collaboration between psycho-sociological specialists and econometric specialists. Such collaboration should be developed more end more in the study of many forecasting problems. In this connection, the rule to be followed in approaching the problems appears to be the following: where statistics are available (trade statistics, company sales, or the results of consumer research, for example), it seems advisable to make a preliminary statistical and econometric analysis which comes up with significant results, warns about some peculiarities, and suggests some explanations. This is where the psychologist comes in on already cleared ground where he may better âcenter" his work. Lastly, collaboration between the two approaches should occur to establish a synthesis of the results. On the other hand, where there are practically no statistics, they must be reconstituted by investigation. However, before undertaking it, is extremely useful for the psychologist to make a preliminarily exploratory study of the problems, whereby he can establish the fundamental hypotheses. Then an extensive investigation may be perfected which permits the psychologist to verify his hypotheses, and supplies econometric specialist the quantified elements which he needs.
This paper has two objects. One of these is to present some examples of research based on the semantic differential technique. The other is to make some general observations about the use of the technique, based on our own experience. The semantic differential has already been quite widely used in market research, but we believe that our use of it in media research has some novelty. At. the same time, we think that our conclusions about the use of the technique may apply in other branches of consumer, and possibly retailer, research. The paper starts with a description of the semantic differential technique. We then present two case studies of the use of the semantic differential in media research. In presenting these case studies we outline the research problems involved, describe briefly the research design, and give some indications of the findings. After this, we discuss methods of analysis and briefly review some technical problems. The paper ends with a summary of our research methods and some concluding remarks.
A major problem of market research units, in any field but we suggest particularly in industrial market research is to achieve the optimum rather than the maximum amount of data required, but with the maximum degree of precision necessary for the problem under consideration to be solved. We do not feel it necessary to say anything further about the question of achieving optima with respect to amounts of information. All we would add is that the postal questionnaire imposes a very severe discipline on the researcher in this respect. It may well be the case that the cost of asking for too much in the case of a postal questionnaire- when measured in terms of non-response rates - is greater than in the case of personal interviews. But what we believe to be more important is the possibility that, with respect to much of the quantitative data required in industrial market surveys, the postal questionnaire affords a better opportunity to the respondent to provide accurate information than is the case in very many personal interview situations. Indeed, not infrequently recourse is necessary to the provision of some sort of form, left with the respondent during a personal interview, by means of which, he subsequently returns the quantitative data required.
Based upon our good experiences with the use of a panel, we wondered whether it would be possible to use this same panel systematically for psychological research. From Mr. Koponen's literature we understood that he has achieved quite a lot in this field. Furthermore, it was known to us that large research-agencies, specialised in psychological research, are also frequently using panels. This, however, is generally done on a more limited scale and by way of verbal interviewing.
Based upon our good experiences with the use of a panel, we wondered whether it would be possible to use this same panel systematically for psychological research. From Mr. Koponen's literature we understood that he has achieved quite a lot in this field. Furthermore, it was known to us that large research-agencies, specialised in psychological research, are also frequently using panels. This, however, is generally done on a more limited scale and by way of verbal interviewing.
Can any market researcher think of a plausible reason why research in industrial markets lags so far behind research amongst consumers? Let us examine some facts, a literature about industrial market research (abbreviated IMR) is almost non-existent. The classical textbooks either neglect the subject, or devote a meagre last chapter to its most glaring problems. The Journal of Marketing has published-, in the last six years, some real contributions to IME literature. Not counting some short and simple features on industrial marketing, the Journal has done practically nothing to remove IME from the attic in Esomar-circles the industrial market researchers donât fill a bus at the annual outing. Their papers are few, their specialty is regarded upon as a queer and demanding hobby. This might seem to be any specialist's sob-story, but for the economic factors to be observed. Granted that marketing outlay, including market research, is correlated to added value in an industry, we can say. that industrial, market research will be correlated to the aggregate added value minus the added value of the consumer goods industry. So there must be a reason why the hinterland of the modern consumer goods industry is such a barren territory for both researchers and marketeers. We intend to review some possible explanations.
The structure of the interviewer organisation should follow; from the functions this organisation has to perform. These functions are of a very special kind, not comparable with other, occupational functions. They are also quite different from those of the researcher and analyst. The interviewer as an individual has to be a conscientious mechanic who carefully follows instructions. The less left to his initiative, the better. For, as a whole, the interviewer organisation must function with the smallest possible variance, that is to say, interviewers should ideally be interchangeable. In order to achieve these aims, a strongly centralised interviewer organisation must be set up. A uniform procedure for the selection of interviewers must be developed which operates on the principle of statistical probability in selecting promising interviewer material. Once selected, the interviewers must be controlled closely and continuously and all possible sources of deviation from the norm must be weeded out as soon as they are detected. For selection as well as for control, test systems must be developed. A number of psychological tests now in use for various purposes can be adapted for interviewer selection, while others have to be now set up. Control tests can often be built into the surveys, for example by split-ballot techniques. Also, selection and control tests can be combined with a training program.
This paper appears to he immediately concerned with advertising penetration. I propose to suggest that more formal analysis of the general notions used in marketing is extremely relevant and practical. There is a special reason for taking advertising penetration as an example of my theme. The subject is one which for me is associated with a peculiar problem (the nature of which is immaterial) over the technique to be used for its measurement. To resolve this problem, a closer study of the nature of the concept of advertising penetration was made, and certain investigations undertaken to show some of the implications of faulty definition. Some of this material is included in a condensed form in the paper which follows. The paper itself is presented in two parts. The first part consists of an examination of the term advertising penetration, from which it will be evident that I believe that what appears to be obvious should hot be accepted at its face value. The second part relates to the advantages of two possible measurement techniques.
In this paper I want to outline two applications of a mathematical model for consumer purchasing data. Apart from their own direct interest, these applications may serve to illustrate how a mathematical model can he useful for two somewhat different purposes, namely: A. To provide insight into a situation previously not understood - in this case the effect of particular patterns of purchasing on sampling errors; B. To allow the prediction of certain quantities instead of having to observe them directly - here the prediction of the market penetration in a longer period of time than has actually been observed. The model which will be used involves the so-called Negative Binomial distribution. Earlier work on the fit of this distribution to consumer purchasing data has already been described, but it may be helpful to summarize it briefly here before outlining the more recent developments.
In recent years, many objections have been raised against existing conceptions and theoretical propositions in the field of social stratification. This paper does not attempt to add another appraisal in this respect, however scientifically warranted such an endeavour might prove to be. The purpose of this paper is to focus critically on some rather widely employed practices in sample surveys attempting to ascertain social strata. Reference to conceptual and theoretical categories is made only, insofar as such an examination is a prerequisite for the criticism directed at some common practices used in survey research. In short, after briefly examining some conceptual aspects and generally employed criteria of social stratification, and the so-called objective and subjective methods, we shall mainly be concerned with two of the most commonly applied devices for stratifying a population, i.e., interviewer rating and index construction or scaling.