How do I make an option comparison when some criteria cannot be measured precisely?

Use a matrix that separates concrete details from informed judgments. Give every option the same criteria, then label qualitative observations as judgments and explain what they are based on. You can still be specific: “The entrance is visible from the main path” is clearer than “welcoming,” while a note such as “I judged the arrival experience by visibility, signage, and how easy it seemed to find the room” explains your method.

For each criterion, record one of three things: a verified fact, a qualitative description, or an unknown. Don’t convert an uncertain impression into a decimal score just to make the matrix look decisive.

It can, if you define the labels before comparing the locations and include a short reason for each rating. For example, “Good” might mean the available information indicates that most participants can find the room without assistance; “Mixed” might mean some helpful features are present but a notable obstacle remains. Those are working definitions for your comparison, not universal standards.

If you cannot apply a label consistently, skip it and write a concise description instead. A rating without its basis can conceal differences in how you judged each option.

Here is a hypothetical example. The names and details are placeholders; replace them with your own observations. Notice that the evidence and uncertainty are visible alongside the judgment.

Criterion: Ease of finding the meeting room

1. Cedar Room — Judgment: Appears straightforward. Basis: On my visit, the entrance was visible from the main path and a sign named the room. Evidence: Personal observation during one visit. Limit: I did not observe someone unfamiliar with the venue arriving.
2. North Hall — Judgment: Unclear. Basis: My notes mention a shared entrance but do not describe signs along the route. Evidence: Existing notes only. Limit: I have not visited; this is not evidence that the room is hard to find.
3. Studio Annex — Judgment: May require extra guidance. Basis: The notes say the room is in a separate building. Evidence: Existing notes. Limit: I do not know whether directions or signs make the route easy.

You can add a separate line for practical facts, such as the stated room capacity, if you have a reliable source for them. Don’t let an unknown become a negative rating: mark it “not assessed” and say what would resolve it.

Tie the recommendation to the group’s stated priorities, then name the trade-off and the uncertainty. For example: “If easy arrival is the priority, Cedar Room is the best-supported choice from the information available because I observed a visible entrance and room sign during one visit. North Hall and Studio Annex are not assessed well enough on this point. I would confirm their signage before treating this as a final comparison.”

That wording distinguishes a provisional choice from a proven overall winner. If different people value the criteria differently, show the trade-offs rather than hiding them in one combined score. You can invite them to say which criteria matter most, or what evidence would change the recommendation.

Use this copyable structure for each criterion and option:

Criterion: [What matters] What I mean by it: [Plain-language definition] Option: [Name] Finding: [Fact, qualitative description, or “not assessed”] Basis: [Observation, note, document, or other evidence available] Judgment: [Optional label, using the shared definitions] Limit or uncertainty: [What this evidence does not establish]

After the entries, add: “Recommendation: [choice, if appropriate]. This depends on [priority or assumption]. The main trade-off is [trade-off]. I would revisit it if [new evidence or changed priority].” This keeps the comparison useful while making clear where evidence ends and judgment begins.