← Vault Index
Source: frameworks/kit-interview-scorecard-design/04-quality.md

Quality — Scorecard QC Checklists

Gate 1: Binary — all items must pass before building starts. No scoring. No partial credit. A single failure stops the build.

Gate 2: Weighted — 100 points total. Pass threshold: 90/100. Run after every build and after every revision.

Both gates must pass before any scorecard is deployed to an interview team.


Gate 1 — Pre-Build (Gap Protocol)

Run this before opening the build skill. Any failure stops the build. Produce a gap report and resolve before proceeding.

Routing Check — Run This First

Is an extraction interview available for this practitioner's scorecard methodology?

If the advisor confirms that a template-only build is intentional (e.g., producing a working draft for practitioner review) — document that decision, label the output clearly as a working draft, and proceed with the understanding that significant gaps will exist and the scorecard is not deployment-ready.

Upstream Inputs

Extraction Coverage

Focus Area Design

Scoring

Gap Report Status


Gate 2 — Post-Build (100 points, 90+ to pass)

Run in this order after every build and every revision.


Design Integrity (35 points)

#CheckPoints
1Every focus area traces to a must-have requirement or confirmed competency domain6
2Every critical must-have from the position profile is covered by at least one focus area6
3No focus area is duplicated across interviewers unless intentionally overlapped4
4Each focus area description defines what "good" looks like in observable terms5
5Each focus area contains 3-5 behavior-based questions (not hypothetical, not leading)5
6Questions within each focus area are distinct — no overlapping evaluation targets4
7Scoring scale levels are defined and distinguishable from each other5

Design integrity failures are blocking. A scorecard that evaluates the wrong things, misses critical requirements, or uses undefined scoring is not deployable regardless of total score.


Legal Defensibility (20 points)

#CheckPoints
8Same questions will be asked of every candidate for the same focus area5
9Scoring criteria are consistent across candidates — no per-candidate adjustments4
10Written justification is required (not optional) for every score4
11Justification prompt specifies fact-based evidence, not impressions or feelings4
12No questions that could elicit protected-class information (age, family status, religion, disability, national origin)3

Defensibility failures are blocking. A scorecard that enables inconsistent evaluation or includes legally problematic questions must be fixed before deployment.


Content Accuracy (15 points)

#CheckPoints
13Role title matches position profile exactly2
14Organization name matches reference data exactly2
15Interviewer names (if included) match reference data exactly3
16Must-have requirements referenced in focus areas match position profile language4
17No content drawn from the golden example as a source2
18No content from a prior client's scorecard carried into this build2

Usability (15 points)

#CheckPoints
19Scorecard can be completed within the interview time allocation (not too many questions per focus area)3
20Instructions are clear — interviewer knows what to do with each section without external explanation3
21Justification fields have enough space for meaningful responses (not a single-line text box)2
22Focus area descriptions are specific enough that two interviewers would evaluate the same things3
23Recommendation section is clearly separated from section-level scoring2
24Submission instructions are included (where to send, deadline, who receives it)2

Presentation Section (10 points — skip if no presentation in process)

#CheckPoints
25Presentation evaluation criteria are defined (not just "how they presented")3
26Time management is an explicit criterion (did they stay within limits?)2
27Q&A handling is an explicit criterion2
28Criteria are consistent for every candidate presenting3

If no presentation in the process, redistribute these 10 points: add 3 to Design Integrity, 3 to Defensibility, 2 to Usability, 2 to Content Accuracy.


Debrief Readiness (5 points)

#CheckPoints
29Scorecard structure allows facilitator to aggregate scores across interviewers2
30Recommendation scale is consistent across all interviewers' scorecards2
31Submission deadline and no-cross-visibility rule are documented in the scorecard or its instructions1

Scoring Summary

CategoryPoints
Design Integrity35
Legal Defensibility20
Content Accuracy15
Usability15
Presentation Section10
Debrief Readiness5
Total100

Pass threshold: 90/100

Blocking failures (must fix regardless of score):

After any revision: Return to Gate 2 and run all checks again. A fix may introduce a new issue.


Common Failure Modes

FailureWhat It Looks LikeRoot CauseFix
Generic focus areasScorecard evaluates "Leadership" and "Communication" without role-specific definitionFocus areas pulled from a generic competency model, not from the position profile and extractionRebuild focus areas from must-haves and extraction; generic competency names are starting points, not finished focus areas
Orphaned requirementsPosition profile lists "donor relationship management" as a must-have; no focus area evaluates itFocus area design didn't systematically trace back to must-havesCross-reference every must-have against focus area list; add coverage for any gap
Hypothetical questions"What would you do if you inherited a team with morale issues?"Questions not structured as behavior-based; interviewer gets hypothetical answers, not evidenceRewrite: "Tell me about a time you inherited a team with morale challenges. What did you find, what did you do, and what changed?"
Undefined scoring scaleScale says "1-5" but never defines what each number meansScoring methodology not captured during extraction; scale imported from a template without calibrationDefine each level in behavioral terms; confirm with practitioner
Duplicated focus areasTwo interviewers both evaluating "strategic thinking" with different questionsFocus area assignments not mapped during alignment; interviewers chose their own areasMap all focus areas to interviewers before deployment; identify intentional overlaps vs. accidental duplication
Opinion-based justificationInterviewer writes "Seemed like a great culture fit" with no supporting evidenceJustification prompt doesn't explicitly require fact-based evidenceAdd explicit prompt: "Reference specific candidate statements, behaviors, or demonstrated competencies"
Prior scorecard recycledScorecard uses focus areas from a different role at a different organizationTemplate-only build without extraction; prior scorecard treated as content sourceEvery scorecard is built from the current role's position profile and extraction, not from a prior engagement
Protected-class questions"Do you have children?" appears in a work-life balance focus areaQuestion bank not reviewed for legal complianceRemove immediately; review all questions for protected-class exposure; add legal review to QC