Pinned post
When is an 'area' of valid truth, rather than a single point, the more appropriate assumption for annotating texts? In our new preprint,led by @AhrabhiKat with @jana_bhard and H.Boomgaarden,we study the presence & sources of coder bias in annotating constructs of marginalization.
Preprint alert! 🚨Happy to share our work on a key challenge in content analysis: coder bias – the systematic variation in annotations driven by individual characteristics when coding with human or LLMs. Let’s break it down… osf.io/preprints/soca… #Bias #Research #Methods #AIView quoted post on X