This paper illustrates the effects of Internet usage on attitudinal and behavioral aspects and describes different ways in which usersâ mode of relationship with the Internet can be profiled and classified for segmentation and analysis purposes in Brazil. First, differences between groups of users and non-users are examined based on a series of general population sample survey studies. Following this, a new categorization of modes of relationship with the Internet is proposed using information of usage patterns. Finally, the relative leverage of this approach is tested visà -vis more standard classifications, explaining usersâ habits and postures on issues related to the spheres of consumption and politics.
In order to clarify our thinking about questionnaire design, it will be helpful to break down the issues associated with it into distinct groupings. A useful four way classification of these groups is as follows: a) meeting research objectives b) obtaining valid and reliable data from respondents c) facilitating the interviewerâs task and subsequent data processing d) achieving and maintaining respondent involvement. Groupings b), c) and d) are dealt with in the individual sections of this chapter. However, the first group of issues connected with meeting research objectives deserves special mention at this preliminary stage.
Hayashiâs quantification methods are multivariate analysis techniques dealing with qualitative data, which were developed by Dr. Chikio Hayashi, former Director-General of the Institute of Statistical Mathematics (Tokei-Suri Kenkyuzyo), in the late 1940s and 1950s. The methods have been, and are, widely used for academic and business fields including marketing research in Japan, hence their inclusion here. The methods traditionally employ the following terminology: 1. âexternal criterionâ for dependent variable; 2. âexplanatory variablesâ for independent variables; 3. âitemsâ for question items in the questionnaire; and 4. âcategoriesâ for alternative answers to the question.
Since the earliest days of marketing and social research, attempts have been made to group populations and survey respondents into classes: groupings which discriminate between people in ways which are considered likely to be relevant to the research objectives. We all know from our own experience that people from different social groups differ not only in what they can afford but in their preferences and habits. Such information has always been used to set quotas for samples and in the analysis of survey results; frequently the classification variables are âproxiesâ for what the researcher is really interested in but is unable to measure directly, such as their tastes and attitudes towards the product in question. Classification variables such as sex and age are straightforward to obtain. In this chapter we look at one of the oldest, but still most vexed and problematic, classifications: social class or socio-economic grading.
In this paper, a classification method (the index method) is proposed which aims to discriminate between members of two groups (e.g. buyers and non-buyers). The index method is compared with two-group discriminant analysis which is a standard and common technique of exploratory data analysis for selecting relevant variables and specifying the relationships between variables. It is shown how the index methods accounts for the problems met in applying two-group discriminant analysis.
The study from which this paper has been written was conducted on behalf of Radio Luxembourg (London) Limited, part of Compagnie Luxembourgeoise de Teledifussion's television and radio service. In 1979 Radio Luxembourg commissioned Social Surveys (Gallup Poll) Limited to conduct a programming study to assess the requirement from an evening radio station among persons available to listen at that time. As a result of the study the station was re-programmed and with the aid of advertising/marketing the decline in audience was reversed. Similar studies were conducted in 1980 and 1981 with subsequent minor amendments to programming to take account of changing tastes, and the station has maintained its higher audience levels.
This paper describes some recent developments in the area of consumer classification. No attempt has been made to be comprehensive and coverage is confined mainly to the UK, where the author has most first hand experience. The paper was intended as a base for discussions and cross- country comparison during the seminar
This paper discusses current and future developments in the ways pre-classification can be used. The paper particularly examines the use of small area statistics linked to other data bases. A suggestion is also made as to how pre- classification can be used, to optimally stratify and allocate samples, when the estimation of more than one variable is considered.
It could be argued that market and media research have already reached a high level of sophistication. That the consumer classification systems we have evolved over the past years keep the information we need flowing smoothly and effectively. So why rethink anything? Because we in Europe are experiencing the first ripples of an irresistible wave of change. Change in society. Change in markets. Change in the media. The backdrop against which we classify consumers, whether we are marketers, researchers, media owners or advertising people, is no longer a comfortable and predictable constant. Increasingly, we find ourselves classifying a moving target. The classifications we developed for yesterday's reality start collapsing when we try to apply them to tomorrow's.
Due to the potential impact and significance of the undecideds and the widespread use of telephone surveys, it became imperative that an improved technique be developed which would result in a reliable forecast of the likely voting patterns of respondents who claimed to be undecided. Such a technique, which can be used with any method of data collection, was developed by the authors. This paper will discuss:-1. the new technique -2. a specific application of the technique in the 1980 US presidential election and -3. the results of a validation study in which post-election follow-up interviews were conducted with previously undecided voters. In this validation, a comparison of the technique's predictions and reported voting behavior revealed that the model correctly classified approximately 80% of the undecided respondents in the three-way race for President. In addition, we will also discuss potential uses and applications of the procedure within the European political scene.
In a study for a local daily newspaper we have developed a method to classify readers according to their interests for several types of news, such as local news, sport news, sensational items, international, political, cultural, social news, etc. The study was commissioned after a merger between two local newspapers had taken place and only one main competitor was left. The method led to a typology, which made clear why people preferred one local newspaper to another and how, from a marketing point of view, the content of the newspaper could be adjusted to achieve a better penetration in some promising segments of the market. The study was based on a sample of 600 in Arnhem, a town with approximately 130.000 inhabitants, and took place in November 1973.
Perhaps one of the more fascinating aspects of taxonomy - of classification, that is, seen as a formation rather than as the identification of classes - consists in gradual research by "sifting", known as analysis of the latent structure, with the object of finding out what really lies below the surface of a given phenomenon. Classification, indeed, as we mean it, signifies grouping certain elements together on the basis of a characteristic, of an ensemble of characteristics or attributes particularly suited for grouping them together and, at the same time, distinguishing them from the others.