Hvordan studere generiske nyhetsrammer? – Forslag til metodiske forbedringer

In this article, I propose a new strategy for measuring generic news frames, extending the work of Semetko and Valkenburg (2000). I argue that frames should be measured through concrete variables that together form a frame (several items per frame, cf. ibid. and de Vreese (2003)). Secondly, I argue that these variables should be measured using graded scale variables as opposed to dichotomous variables. This is seen to pose important challenges, and advantages, to inter-coder reliability as well as validity issues.