Moved to https://www.metaculus.com/c/unjournal/42735/a-overall-wellby-vs-b-separate-measures-and-conversion-well_04/

Prediction market on metaculus. We've made this more readable in the [categorical question here](). Please make a prediction on that question instead. **Question Code: WELL\_04** This question is part of The Unjournal's Pivotal Questions project on measuring well-being and the reliability of the WELLBY metric. This is WELL\_04 in our Coda page for PQ evaluators. Consider general contexts where interventions may have impacts on both mental health, physical health, and consumption, such as a 'focal example' of allocating \$100,000 among a large set of charities. Do you agree: "In these contexts it is best to \[A] use a (potentially imperfect but single) overall WELLBY-based measure (as defined above) instead of \[B] directly measuring each dimension separately and then converting and combining these?" *Our notion of best (briefly):* Leads to the decisions that lead to the highest ‘true welfare/average welfare’ on average, perhaps taking into account the cost of doing the measurements. ... In the particular relevant domain (e.g., in comparing mental health and diseainterventions in Africa; see the [focal case](https://uj-wellbeing-workshop.netlify.app/linear-wellby-analysis)) *Note*: This question is somewhat secondary; strictly speaking, it's covered in the "focal question" (WELL\_01) but we repeat it here for emphasis and for the prediction interface. **Metaculus framing:** Suppose we chose an important context at random. With what probability would \[A] yield a better (even marginally better) choice than using \[B]?

Resolves: 1/11/2027.