Qualitative sampling in a multi-method survey

Because method triangulation is able to reduce disadvantages of one method by using another one, it has been demanded in social science for many years. Nevertheless, it is seldom applied in practical research. To show the importance of multi-method-research for improving the results of social science investigation, one possibility of method triangulation and its execution is presented. Out of a representative youth survey about sexual behavior and AIDS-prevention with 1500 German respondents a sample of 15–20 respondents for depth-interviews is extracted by combining clusteranalysis and random-sampling. The aim of that multi-method procedure is to get more detailed information about subgroup specific topics and problems. At first the paper shows the necessity of method triangulation and its application in our survey. After that the procedure and the results of the clusteranalysis are presented. In the last chapter the sampling and its results are discribed and discussed.