I’m using Perplexity to research whether standing desks help office workers. The sources don’t seem to be answering one consistent question: some measure sitting time, some report discomfort, and others discuss longer-term health indicators.

A summary saying “the evidence is positive” would hide those differences. A study showing less sitting time does not, by itself, answer whether people experienced less back pain.

How would you frame a request for comparing study outcome measures so Perplexity keeps each finding attached to the outcome actually measured? I’m thinking of requesting the participant group, study duration, comparison group, and outcome for each paper, with missing details marked as unavailable.

Would you start with one narrowly defined outcome and search separately for the others? Or is there a reliable way to get an initial evidence map without letting several different questions collapse into a single verdict?