Use of focus group research appears increasingly to be becoming more common in the gathering of empirical data in the field of social science. In a study of school achievements and student strategies at the upper secondary school level, we have utilised focus groups as a part of an ethnographic method. After a period of participatory observation in the classroom, students were then divided into groups and equipped with disposable cameras in order to take pictures of their schooldays - both good and bad experiences - thereafter discussing these pictures in focus groups (Aasebø 2006). Focus group research uses group interaction in order to generate data, a factor which implies that at an analytical level, it is the group's expression (as opposed to that of the individual) which is focused upon (Barbour and Kitzinger 1999). In this paper I wish to discuss the validity of focus group research. Validity is connected both with data's characteristics and the type of discussions related to regarding the data as being credible and transferable. A validity discussion is dependent upon the theoretical frame of reference with which the research group research is regarded. Two concepts that can be traced back to various theoretical suppositions are construction and representation. In a constructivist paradigm the focus group is regarded as being a context for meaning construction, in which meaning is derived from the interaction in the specific context, so that the data are appropriate for demonstrating how social processes lead to definite content-based interpretations. In this situation the research itself creates the social process. The concept of representation may be connected with another tradition, one which does not utilise the focus group concept but rather the delineation group discussion method. This tradition originates in Cultural Studies. The groups established are regarded as being representatives for a larger macro-social unit, comprised of discursive formations or interpretive communities, such as social class, generation, immigrant culture (Bohnsack 2004). In this paper I will discuss the credibility and transferability of data, or its validity, in light of the concepts of construction and representation, showing the implications this has in relation to a collection of data material. This also relates to research's generalisation and transferability. Methodological discussion is the topic of the paper, see 'Description of the paper'.Through evaluating concepts such as contextualisation, emphasising social processes instead of fixed attitudes and cultural discourses, on the one hand, and milieu-specific meaning and collective meaning patterns on the other, my intention is to demonstrate how these two traditions do not necessarily create contradictions in the discussion of the legitimacy of research's validity. On the contrary, I wish to demonstrate how one can utilise both through allowing them to complement one another.Aasebø, Turid Skarre (2006) School achievements: how social structures and cultural discourses are negotiated in students' relationship to school. In Tobias Werler and Christoph Wulf (Eds.). Hidden Dimensions of Education. Münster: Waxmann Barbour, Rosaline S and Kitzinger, Jenny (Eds.) (1999) Develpoing Focus Group Research. London: Sage publications.Bohnsack, Ralf. (2004) Group discussion and focus group. In Uwe Flick, Ernst von Kardhoff and Ines Steike (Eds.). A Companion to qualitative research. London: Sage.Guba, Egon G. and Lincoln, Yvonna S. (2005) Paradigmatic controversies, contradictions, and emerging confluences. In Norman K. Denzin and Yvonna S. Lincoln: Qualitative Research. Thousand Oaks: Sage Publications.