Proposed models

Followinh Hageman et al, I used keywords to try and identify candidate procedures for manual coding. Unlike previous studies, I used more terms for identifying the procedures. Thereafter, I ran a correspondence analysis (without normalization or adjustments).