Acta Sociologica

About ten years ago, there was a debate on soft data within Swedish sociology. This article sums up the contents of the soft-data debate from an ideological historical point of view. According to both Erik Allardt and Gõran Therbom, this debate was unique for Sweden. The distinction between soft-data and hard-data sociology is based on formal aspects of data. In soft-data sociology qualitative data are used instead of quantitative data. According to the soft-data sociologist, precision is not the ultimate goal of social science. With this wide definition, there are different branches of soft-data sociology with respect to both philosophy of science and consequences for society. In this article, an effort will be made to specify these branches. Furthermore, there will be a discussion of the bases for scientific judgements in both hard-data and soft-data sociology. What are the reasons then for using soft data? The study of scientifically and politically important problems needs soft data. And it is also a way to an anti-elitism among social scientists. Moreover, soft-data sociology used in an un-mechanical research process promises to be empirically and theoretically richer than the positivistic hard-data sociology.

Studies of time-use are rather commonplace today. Central statistical bureaus in many countries have adopted the technique and are contemplating regular surveys. The idea that measurements of time and time-allocations can be used as general indicators of living conditions is taken up by many. Serious schemes have been developed that propose exchanging the well-known 'money-budget' for 'time-budgets'. Development programs for 'social indicators' all incorporate various measures based on time and time-use.

It is fair to say, however, that this field of inquiry is still in its infancy. It has proved possible to obtain data on people's daily use of time with sufficient reliability, but there is an alarming lack of reports on the methodological difficulties involved. The studies performed are typically empirical, fact-finding surveys with little attention given to the theoretical issues involved. The dominant instrument of data-gathering is the '24-hour time diary'. The various studies follow a general pattern which makes comparison possible, but there is a great need for further development of this research instrument. This is readily understood by those who have ventured to use it. The traditional time-use studies are very expensive and require great inputs of both time and money.

This paper will not deal with all the problems mentioned. First the concrete reality at hand will be dealt with and thereafter some major methodological problems illuminated. A framework for classifying activities is proposed and, finally, indication is given of how the 'diary approach' can possibly be adapted for inclusion in interview surveys.

Epistemological positions with regard to action research are presented for Empiricism, Logical Positivism, Structuralism, Pragmatism, and Dialectical Materialism. These schools are examined with regard to the following questions: How does man produce knowledge? How does man justify knowledge? What is the relation between theory and practice? What is the relation between values or ideology and science? and What are the implications of the foregoing for action research? The reasons for the rejection of action research by Empiricism, Logical Positivism, and Structuralism are presented, as well as the basis for its acceptance by Pragmatism and Dialectical Materialism.

This article reports a particular research experience in order to illustrate some general issues connected to the use of an inductive qualitative approach. When a research method is characterized by little structure, many false starts, and much possible influence of the scientist's personality and interests, can it really be called scientific? It can be argued that the inductive approach also follows a classic scientific model, but the answer to this question ultimately depends on the product and on the scientific community.