Skip to content
English
Go back to qracorp.com
  • There are no suggestions because the search field is empty.

Transitioning to the New QVscribe Scoring

Tips for comparing results, setting a new baseline, and updating internal processes.

A quick mindset shift

When scoring changes, the most important step is not fixing anything. It is resetting your baseline and confirming that your configuration still reflects how you want requirements to be evaluated.

The updated scoring model applies starting with:

  • QVscribe Office: v8.0

  • QVscribe for QTS: v8.0

  • QVscribe for QRAcloud: v9.0

Any documents analyzed in these versions or later will reflect the new scoring behavior.

👉 If you have not already, start with Understanding the Updated QVscribe Requirements Scoring to learn what changed and why.

Step 1: Compare historical documents

Start with a document you already know well.

Recommended approach

  1. Choose one or two representative documents

  2. Record the existing scores from analyses run before the update

  3. Re-run QVscribe using QVscribe Office v8.0, QVscribe for QTS v8.0, or QVscribe for QRAcloud v9.0 (or later)

  4. Compare:

    • Average score

    • Score distribution

    • Common Problem Types

What you are looking for

  • Overall score trends may shift upward
  • Relative quality between requirements should mostly look familiar, though some requirements will shift places (for example, those with a single critical issue may move up)
  • Problem Types should remain consistent
  • Requirements with multiple instances of the same issue may now separate out where they previously scored the same. That’s expected, since the new scoring is built to show that difference.
  • Documents heavy in Excessive Continuances issues may trend lower, not higher

Focus on patterns and trends, not exact numbers.

Step 2: Review your configuration with the new scoring

After reviewing score changes, it is a good time to revisit your configuration settings.

Apart from the updated Excessive Continuances weighting, the scoring update does not change your configuration. However, because scores now behave more smoothly, some teams find it helpful to confirm that their rules still reflect current expectations. 

Only Configuration Authors can make changes, but all users can view the configuration details.

Step 3: Decide how you want to review configuration impact

There are two recommended ways to review your configuration with the new scoring.

Option 1: Review and update live in the Configuration window

Configuration Authors can open the Configuration window and review settings directly.

As you:

  • Turn Problem Types on or off

  • Add or remove problem phrases

  • Switch between Starting Configurations such as QV1 through QV4

You will see score changes update in real time.

This approach works well when you want to explore adjustments interactively. Just remember to save your configuration when you are done so changes are shared with your team.

👉 For help choosing between QV1 to QV4, see QVscribe Starting Configurations.

Option 2: Compare reports using old and new scoring

If you prefer a more documented approach, you can compare reports.

Recommended steps:

  1. Generate a report using the previous scoring before updating. If you’ve already updated, use an existing report from before the update.
  2. Generate a second report using the updated scoring.
  3. Include only the following analysis sections:
    1. Score Summary
    2. Analysis Breakdown
    3. Configuration Appendix

When reviewing the reports, compare:

  • Overall and average scores

  • Changes in score distribution

  • The list of Problem Types and associated phrases in the Configuration Appendix

This approach is especially useful for audits, governance reviews, or baseline documentation.

Step 4: Evaluate whether changes are needed

As you review results, ask:

  • Are any Problem Types too strict or too lenient?

  • Do certain phrases need to be added or removed?

  • Is the current Starting Configuration still the best fit?

  • Would switching between QV1, QV2, QV3, or QV4 better match this document type or project phase?

If adjustments are needed, Configuration Authors can update the configuration directly or create a refined version for specific projects.

👉 For detailed walkthroughs, see How to Further Customize Configurations.

Step 5: Set a new baseline

Once you are comfortable with both the scoring behavior and configuration:

  • Treat the first analysis run with the updated scoring as your new baseline

  • Document the new average score range

  • Capture a Snapshot to preserve this baseline

This helps prevent confusion when comparing future results.

Step 6: Update internal thresholds and guidance

If your organization uses scores for decisions, review and update:

  • Scoring thresholds and gates

  • Definitions of good enough

  • Review or release criteria

Avoid reusing old numeric thresholds without validation.

For example, under the previous scoring, a gate of “3 or higher” would exclude any requirement with no imperative or multiple imperatives. Under the new scoring, a requirement with a single one of these issues scores 3.0 and would pass. Review gates like this against your new baseline before relying on them.

Instead:

  • Look at how scores distribute under the new model

  • Align thresholds with observed clarity and consistency

  • Use trends and Problem Types to guide decisions

Remember that scores support judgment. They do not replace it.

Common mistakes to avoid

  • Assuming higher scores mean lower standards

  • Updating thresholds before reviewing configuration impact

  • Comparing old and new scores without documenting the baseline shift

  • Making configuration changes without testing on real requirements

Key takeaway

Transitioning to the new scoring is about alignment, not rework.

By comparing historical documents, reviewing configuration settings, and setting a clear new baseline, you can take full advantage of a more predictable and meaningful quality signal while keeping your process intact.