Fig. 1 | Scientific Reports

Fig. 1

From: Analyzing patient perspectives with large language models: a cross-sectional study of sentiment and thematic classification on exception from informed consent

Fig. 1

Large language models (LLM) polarity score across participating sites (n = 3,692 responses). Each row represents the mean LLM response polarity score to each question, color indicates assigned polarity. Each question is represented by question number and representative domain. All sites represent the mean LLM response polarity score from all nine sites, while the subsequent three columns provide a breakdown of the response polarity score at three sites, selected for detailed comparison to simplify review.

Back to article page