There is an ongoing discussion about what content that should be taught in science education and there are different views among teachers about what represent good science content. However, teachers are not isolated individuals making their own interpretations, but are part of institutionalised systems building on patterns in the selection of teaching goals and content. Earlier research shows that teachers teach in alignment with different selective traditions, which can be understood as well-developed teaching habits. Individual teachers seem to develop their personal habits on the basis of the contextual situations created by earlier generations of teachers. In order to find out which content teachers find representative for science education, we asked nine teachers to take part in group interviews to talk about what they value as "good" science content. The participants were grouped according to their selective traditions expressed in earlier studies. The method was used to dynamically explore, challenge and highlight teachers' views. The starting point for the group discussions is national tests in science. In Sweden, national tests in biology, physics and chemistry were introduced in secondary school science (year 9) in 2009. One overarching aim of these tests is to support the implementation of the science curricula and to include for example knowledge about socio-scientific issues (SSI). The content of the tests can consequently be seen as important for teachers to consider. The findings show that 'resistance' to including SSI is not just an issue for individual teachers. As individuals teachers can create many kinds of obstacles, but still be interested in integrating SSI in their science teaching. However, in group discussions the teachers tend to collectively adopt the scientific rational discourse. This discourse is what joins them and creates their common identity as science teachers. In turn, they seek to free scientific knowledge from social knowledge and thereby make assessment easier.