Skip to content

Use maximum variation to discover different mechanisms

Select contrasting cases deliberately when the goal is to uncover different explanations, not to estimate how common each explanation is.

Guides

How should exploratory sampling seek different experiences?

A study of public discussions about home composters may need to understand several failure mechanisms before designing follow-up questions. Selecting the most similar discussions would make the review easier but less informative. Maximum variation sampling deliberately seeks contrasting, observable conditions while making no claim that the resulting mix represents all owners.

Example research prompt
Compare supplied composter discussions across explicitly stated use conditions. Select cases that add contrasting mechanisms and leave unsupported matrix cells empty.
01 / Guides

Choose observable dimensions

Use stated equipment type, described setup stage and explicit indoor or outdoor use as candidate dimensions. Include only attributes visible in the relevant discussion. Exclude guessed income, household composition or location inferred from profiles. A useful contrast concerns the documented task, not a private characteristic assigned to the author.

02 / Guides

Build a contrast matrix

Create rows for the combinations most relevant to the decision, such as first assembly indoors versus routine maintenance outdoors. Mark combinations with no accessible evidence as gaps. Do not invent cases to complete the matrix or assume that a missing combination is impossible simply because the available material does not contain it.

03 / Guides

Select for explanatory difference

Review examples that add a new condition or challenge an existing mechanism. Keep a reason for each selection, including why a less popular discussion was useful. If one visible condition dominates the available evidence, report that limitation rather than repeatedly choosing similar cases and calling the result diverse.

04 / Guides

Deliver a mechanism map

Present the contrasting cases and what each contributes to understanding the problem. Separate shared obstacles from condition-specific ones and identify the next contrast worth investigating. Do not publish the proportion of selected cases in each cell as a market estimate; the sample was designed for variation, not representative frequency.

Keep the boundaries clear.

  • Variation must be based on observable conditions rather than inferred personal traits.
  • Platform badges describe research context, not guaranteed access to evidence on that platform.

Start with a read-only check: ask your agent to report actual source coverage before collecting records or exporting an audience.

Primary references

Documentation reviewed 2026-11-13. Product capabilities, client availability and plans can change. Third-party marks identify their products, not an affiliation.

Know what social media shows about you.
Start with a question.

Get started
montage