Language quality

Measuring comprehension.

Measure task success and correct understanding first, then use readability measures as supporting diagnostics.

Clarify the task and decision

This guide turns measuring comprehension into a reviewable operating workflow. It connects domain decisions, ownership, evidence, and acceptance so the result continues to work in production.

Measure task success and correct understanding first, then use readability measures as supporting diagnostics.

Practical workflow

  1. 1

    Define the audience with evidence, including reading experience, context, device, and support available.

  2. 2

    Select one real task and state what a successful reader must understand or do.

  3. 3

    Create the language version while preserving every decision-relevant fact, condition, date, and contact path.

  4. 4

    Run specialist and editorial review with recorded acceptance criteria.

  5. 5

    Test with members of the intended audience, record findings, revise, and approve the release.

Worked example or tool

A monthly scorecard combines comprehension, completion, confidence, and support demand. In the tool, also record the baseline, owner, decision, evidence, open issue, and approval date. Use a real page or transaction so the team sees dependencies, exceptions, and the maintenance work that follows release.

Decision pointRecordAcceptance criterion
BaselineObserved current stateSource and date recorded
DecisionSelected option and rationaleRisk and audience considered
EvidenceTest, document, or measureReviewable and version-specific
ApprovalName, role, and dateAll mandatory criteria met

Measure understanding as task performance

Define comprehension as a reader reaching the correct interpretation and using it to make the intended decision or action. Measure four dimensions: factual understanding, task completion, confidence, and assistance required. Keep readability as a diagnostic measure of surface complexity. It can identify long sentences or unusual words, but it cannot show whether a deadline was understood, an exception was applied correctly, or the next action was completed.

Write the measurement plan before revising the content. State the audience, scenario, critical facts, correct outcome, unacceptable misunderstanding, and collection method. For a benefits page, critical facts may include eligibility, required documents, deadline, responsible office, and submission route. A successful participant must explain these correctly and start the right application path without being directed by the facilitator.

DimensionMeasureExample target
UnderstandingCorrect answers to scenario-based questionsNo critical fact below 90 percent
Task successIndependent completion of the next actionAt least 85 percent
ConfidenceSelf-rating after the answerIncorrect high-confidence answers below 5 percent
Support demandHelp requests per completed serviceReduction against a fixed baseline

Combine controlled tests with service data

Use moderated or unmoderated task tests before release to observe understanding directly. After release, combine analytics events, form completion, validation errors, search refinement, support contacts, and short intercept questions. Do not infer comprehension from time on page alone. A short visit can mean success or immediate abandonment, while a long visit can mean careful reading or confusion.

Create a stable content-version identifier and include it in the test record and relevant analytics events. Segment results by device, language mode, referral path, and audience characteristics that can be collected lawfully and ethically. Compare equivalent periods and account for service changes, campaigns, and seasonal volume. Keep raw observations separate from derived scores so calculations can be reproduced.

  1. 1

    Establish a four-week baseline using the current service and definitions.

  2. 2

    Run scenario-based comprehension tasks on the proposed revision.

  3. 3

    Release with versioned analytics and support categories.

  4. 4

    Review the first two weeks for critical failures, then report monthly.

  5. 5

    Retest after material content, process, audience, or interface changes.

Build a scorecard that preserves the underlying evidence

Report each dimension separately before presenting any composite score. A practical monthly scorecard lists the number of participants or service sessions, correct understanding by critical fact, independent task completion, median confidence, high-confidence errors, assistance rate, form validation errors, and support contacts per 1,000 starts. Show the baseline, current result, target, direction of change, and data owner.

If leadership needs one index, publish its formula and weights. For example, combine 40 percent critical-fact comprehension, 30 percent task completion, 15 percent calibrated confidence, and 15 percent inverse support demand. Never let a high readability or satisfaction value compensate for a critical misunderstanding. Add a release gate that operates on individual critical measures before the composite result is calculated.

  • Every measure has a definition, numerator, denominator, source, and owner.

  • Critical facts have individual release thresholds.

  • The tested and released content versions are identifiable.

  • Sample size and missing data are visible beside percentages.

  • Composite weights and exclusions are documented.

Turn measurement into an improvement decision

When a measure misses its target, review recordings, responses, and service evidence to locate the cause. A misunderstood condition may need a clearer example. A low completion rate with correct understanding may indicate a broken form or missing document-upload instruction. A rise in support contacts may follow a policy change rather than a writing defect. Assign the action to content, design, engineering, policy, or operations and state the expected measurable effect.

Approve the next revision only when the team can explain which evidence it addresses. Keep an experiment log with hypothesis, change, affected audience, start date, owner, and result. Close the loop by adding confirmed terminology and patterns to the editorial standard. Quarterly, retire measures that no longer support a decision and add measures for new service risks.

Roles, evidence, and approval

Language quality is an operating responsibility. Give the source owner responsibility for factual accuracy, an editor responsibility for clarity, a specialist responsibility for accessibility rules, and a product owner responsibility for publication. The release record should identify the source version, language mode, reviewers, test evidence, approval date, and the event that triggers a new review.

Operations and maintenance

The work does not end at publication. Link the language version or configuration to its source, monitor quality and service measures, and define concrete review triggers. Triggers include source changes, legal changes, new audience needs, recurring support questions, technical changes, and incidents. A named owner evaluates the trigger, opens a new revision when needed, and records renewed approval.

Release checklist

  • The audience and task are explicit.

  • All legal and operational facts match the source.

  • Dates, amounts, conditions, and exceptions are complete.

  • Headings describe reader questions or actions.

  • Links and controls have meaningful labels.

  • Specialist terminology is explained consistently.

  • The intended audience has tested the critical journey.

  • Ownership and the next review trigger are recorded.

Authoritative sources

  1. Federal Centre of Expertise for Accessibility - Easy Language
  2. DIN ISO 24495-1 Plain language - Governing principles and guidelines
  3. W3C - Making Content Usable for People with Cognitive and Learning Disabilities

Put the guide into practice

Test Simple8 with representative content and use the checklist to plan a controlled production workflow.

Test your own text