How to Write a Fair Either-Or Question
Keep both options parallel, define the same time frame, remove hidden assumptions, and offer neither if the choice is not exhaustive.
Create balanced two-option questions for polls, research prompts, classrooms, and conversations.
Draft both options in parallel, then test whether context, tone, order, or a missing response changes what people think the question asks.
Use parallel wording
Name both options at the same level of detail. Compare working from home with working from an office, not freedom at home with the corporate office.
Keep grammar and detail parallel. If option A names a cost, time frame, and benefit, option B needs the same kinds of information. Remove praise and criticism from both labels. Read the pair aloud to catch rhythm that makes one choice sound natural and the other awkward. The goal is equal clarity, not equal attractiveness to every person.
Define the scenario
State the time frame, budget, audience, and relevant condition. Preferences can reverse when the context changes.
Include only context that could change how a reasonable person answers. State the audience, time period, budget, and key condition when they matter, but do not load the prompt with a case for one side. If the scenario needs several sentences to prevent misunderstanding, a two-option poll may be too compressed for the decision being studied.
For a low-stakes prompt, test parallel wording with the this-or-that generator, but do not treat the reaction as survey evidence.
Check whether two options are enough
Either-or questions are useful for constrained choices. If neither, both, or unsure are common legitimate answers, excluding them can distort the result.
List every response a participant might reasonably need before forcing a pair. Neither, both, unsure, not applicable, and it depends can signal different problems. Some are valid answers, while others show that the scenario is unclear. Decide whether to add an option, narrow the audience, or rewrite the condition. Never remove a common answer merely to create a cleaner chart.
Pilot the question before collecting results
Ask a small set of people to explain what each option means in their own words. If their interpretations differ from the intended scenario, revise the wording before launch. Also ask which answer they wanted to give but could not. Repeated missing answers reveal a false dilemma or an important condition.
Test the reversed option order. A simple public poll may not support randomized ordering, but the editor can still inspect whether the first version gives one option a more natural rhythm, a positive label, or the benefit of being mentioned first. Keep both versions in the review record so the wording decision can be explained later.
Keep the question focused on one tradeoff. If option A changes price, speed, quality, and provider while option B changes none of them, the vote cannot reveal which difference mattered. Hold shared conditions constant or state that the result concerns the complete bundles rather than one feature. Save the final wording with the published result.
After the wording is balanced, separate sample size from representativeness before making a claim about the people who did not answer.
Draft and pilot the question
- Shared scenario
Give both options the same audience, budget, and time frame.
- Parallel detail
Use the same number and type of descriptors for each option.
- Neutral terms
Remove praise, blame, fear, and assumed outcomes.
- Missing answers
Ask whether neither, both, other, or unsure is legitimate.
- Order test
Read the options in reverse order and check whether the question feels different.
Bag choice: repairing a loaded question
Instead of asking, 'Would you choose an expensive durable bag or a cheap bag?', ask, 'For a bag used on a daily commute for three years, would you choose A at $160 with repair service or B at $70 with a one-year warranty?'
Wording choices that tilt the response
- Pairing a benefit with a neutral label
- Forcing two answers when the choice is not exhaustive
- Changing multiple conditions between the options
When either-or is the wrong response format
A playful this-or-that prompt can be intentionally simple. Do not reuse its result as evidence for research, policy, or product demand.
Sources used for this guide
- AAPOR, Best Practices for Survey Research. Guidance on question objectives, response order, and survey transparency.
Browse all polls and preferences guides Review false dilemma in the decision glossary