Thanks for writing this up and for highlighting this weakness in our prioritisation report (example 1).
Since the publication of this report (which was quite an early piece of research for me), I’ve built a lot more of these models and strongly agree that it’s important to not just blindly use a weighted average. (Didn’t change anything about our research outcomes in this case, but it could have important effects elsewhere.) Geometric mean is important. I also sometimes use completely different scoring tools (e.g., multiplication, more BOTEC style, as MichaelStJules has commented). It’s always helpful from my experience to experiment with different methods/​perspectives.
Thanks for writing this up and for highlighting this weakness in our prioritisation report (example 1).
Since the publication of this report (which was quite an early piece of research for me), I’ve built a lot more of these models and strongly agree that it’s important to not just blindly use a weighted average. (Didn’t change anything about our research outcomes in this case, but it could have important effects elsewhere.) Geometric mean is important. I also sometimes use completely different scoring tools (e.g., multiplication, more BOTEC style, as MichaelStJules has commented). It’s always helpful from my experience to experiment with different methods/​perspectives.