Att utveckla sociala indikatorer : en surveyansats belyst med exemplet arbetsmiljö

Sammanfattning: This study shall be seen as part of a general effort striving for new and better social measures. The focus is on collecting data through surveys with interviews/questionnaires. The survey method has obvious limitations that stem from its nature of being an indirect assessment of the conditions it is supposed to measure. The final results are dependent on how well the respondents fulfil their task as reporters. The results are also dependent on the means of communication that is used, i. e. the common every day language. The language can leave an imprint on the results, stemming from its ambiguity and many unprecise concepts. In this study there is first of all a stress on these latter semantic problems.Many survey results seem difficult to interpret because of this unsatisfactory lack of precision in the formulations used. An effort is made in this study to develop ways of constructing more precise question formulations. Many examples of such precise operationalizations are also given. There are however obvious limitations regarding how far the precision process can reach. One has to try to strike a balance between what is most desirable and what is practical. Empirical testing is a necessity to reach satisfactory results. Many different validity tests have also been done in this study. It could with the help of these tests, be clearly demonstrated that the measurement error could be decreased a great deal, if the precision in the formulation used is increased.At the same time it should be emphasized that there also are some risks involved in trying to use very precise formulations. Precise formulations could be easier to develop in some contexts compared to others. An important part of the developmental work must always be to test the relevance of the formulations used. Such tests have also been done and have given support to many of the formulations used.