Sociolinguistic research based on spoken language data usually involves creating a written record of the audio material. This process involves making a number of choices that in themselves may have sociolinguistic consequences, in addition to touching upon issues of theory of science when dealing with data. This issue is explored in an examination of 20 sociolinguistic master’s theses in the field of Nordic/Norwegian studies. The selected sample studies have been submitted at Norwegian universities in the period 2014–2024 and have been analysed with particular attention to two issues: How do the students account for their tran­scription strategies in terms of methodological deliberation, and how do they proceed in terms of actual transcription practice? The findings from the qualitative analysis of the material are discussed and problematized in light of key principles of research ethics and sociolinguistic research and theory related to the indexica­lity of written language and language ideologies. Based on this, we argue that there is a need for increased awareness and reflection on transcription as an essential part of the research process. Choices regarding transcription have conse­quences for the scientific quality of the research, but they are also funda­mentally constitutive of reality; as reflections of theoretical assumptions and authoritative language ideologies, they can also contribute to (re)producing marginalized identity positions.