Evaluation methodology
How individual ratings become public scores—and how to read them responsibly. Mouwaten presents structured civic feedback about agencies and public figures in their roles.
Last updated: November 8, 2026
Ratings run from 1 to 5. Each account has one current evaluation per target. Public summaries use medians, with no extra weight for identity verification.
Evaluate the right service or role
Agency evaluations concern the selected agency. Public-figure evaluations concern a specific role assignment: feedback about one role should not be treated as a score for every position a person has held. Read the target and criterion descriptions before rating, and use relevant experience rather than political affinity or personal dislike.
Rate every published criterion
An active account can submit ratings when evaluations are enabled. Each published criterion for the target type must receive exactly one whole-number score from 1 to 5. Higher values express a more favorable assessment. Criteria can differ between agencies and public figures. If you lack enough information to assess all criteria, wait rather than inventing a score.
One current evaluation, with revisions
Submitting again for the same agency or role assignment updates your existing evaluation; it does not add another vote. Previous scores and comment details are recorded as revisions. An evaluation removed or excluded from aggregation cannot be revised through the normal submission flow. Identity verification does not multiply the numerical weight of a rating.
How the median is calculated
For each criterion, eligible scores are sorted and the middle value is selected. With an even number of scores, the two middle values are averaged. For example, 1, 2, 4, 5 has a median of 3. The overall detail-page score is the median of the available criterion medians. If those medians are 2, 4 and 5, the overall score is 4. This is not an arithmetic average of all submitted scores. Your own evaluation summary is the median of your criterion ratings.
Which evaluations contribute
Criterion summaries use evaluations marked eligible for aggregation whose moderation status is approved and which have not been removed. The eligible-evaluation count additionally requires scores for every currently published criterion. If the published criteria change, older evaluations can contribute to existing criterion results without counting as complete evaluations. A criterion with no eligible scores has no median; missing data should not be read as a zero.
Read beyond the headline score
Compare the overall score with the criterion results and evaluation count. A score from a small group is less informative than the same score supported by many experiences. Participation is voluntary: the results are not a representative survey, an audit, or proof of wrongdoing. Differences in time, service context and criteria limit comparisons. The summary data includes per-criterion distributions and counts; the current detail screen emphasizes medians and the complete-evaluation count.
Comments and public anonymity
A written comment is limited to 2,000 characters. Identity approval is required for comments when the platform’s participation setting requires it; ratings without comments do not require identity verification. Every nonempty comment, including an edit, is screened before publication. A held comment and its ratings stay private and are excluded from public feeds and scores until approved. Pending comments cannot be edited while their review is open. Choosing an anonymous comment hides your author identifier and name publicly; the contribution remains linked to your account internally.
Useful feedback and score integrity
Describe the service, timeframe and observable experience. Separate facts from interpretation and use respectful language, including when expressing dissatisfaction. Avoid threats, private information and coordinated manipulation. Participation points and citizen levels recognize activity but do not increase a rating’s weight or grant permissions. Moderation can change which evaluations are included. A lower score alone is not evidence of misconduct, and identity verification does not establish that a comment is accurate.
Why a result can change
New evaluations, revisions, moderation decisions and changes to published criteria can change the results. The summary’s update timestamp reflects the most recent evaluation update, including evaluations that may not be eligible; it is not a certification date or a complete history of moderation. If a result appears inconsistent, contact the team with the target link, criterion and observed values.