Scoring Methodology

How responses are evaluated.

InterviewRigor scores a structured behavioral interview the way a trained assessor would — using behaviorally anchored rating scales (BARS) and weighted, transparent criteria. Each criterion is scored 1–10; the criteria are split into a Content score (what you said) and a Delivery score (how you said it), and your overall performance is summarized as a readiness band. All composites are computed in code from the criterion scores, so the numbers always reconcile. This page documents exactly how each piece works.

View an example report →
Step 1 — Every question is classified

Not every interview question asks for a STAR story, so each question is first classified by what it actually asks for, and then scored on a rubric built for that type. STAR is used only where a specific example is requested.

TypeWhat it asks forUses STAR?
BehavioralA specific past example or story — "tell me about a time…", "describe a situation…", or "…provide a specific example." Hybrids that ask for a view and an example count here.Yes
MotivationWhy you want the role/company, or what about its mission or values appeals to you (e.g. "what aspects of the Credo inspired you?").No
OpinionYour approach, philosophy, or view — "how do you ensure…", "what does integrity mean to you?" — with no example required.No
Background"Tell me about yourself" / walk through your experience.No
Step 2 — Each type has its own weighted rubric

Behavioral

STAR + Outcome45%
Relevance25%
Conciseness15%
Filler words10%
I vs. We5%

Motivation

Company knowledge30%
Relevance25%
Conciseness20%
Real examples15%
Filler words10%

Opinion

Answering the question35%
Substance30%
Real examples15%
Conciseness10%
Filler words10%

Background

Conciseness40%
Relevance40%
Filler words20%
The BARS method

A Behaviorally Anchored Rating Scale ties every point on a scale to a concrete, observable description of a response at that level — instead of vague adjectives. This makes scoring consistent and the feedback specific. Two parts of the evaluation use BARS: the four STAR components and Answering the Question, each on a 1–5 scale.

STAR components behavioral only

Each component is rated 1–5 against the anchors below. The STAR criterion (1–10) = the average of the four × 2, reduced by 1–2 points if the Result is missing or non-specific (the result is the most important element).

Situation

5Specific context (when/where/stakes) in 1–2 sentences
4Clear but a little vague or long
3Present but vague or overlong
2Barely sketched
1Missing or hypothetical

Task

5Crisply states the candidate's OWN responsibility and why it mattered
4Clear ownership, light on stakes
3Implied or blended with the situation, ownership fuzzy
2Barely identifiable
1No defined task

Action

5Several concrete, sequenced steps they personally took, with rationale, consistent "I"
4Specific actions, mostly first-person, limited rationale
3General actions, limited detail, mixed I/we
2Vague or mostly team-attributed
1"Dealt with it" / no personal actions

Result

5Specific, quantified outcome tied to the task, plus learning/impact
4Clear outcome, partly measurable
3Stated but generic or unquantified
2Vague gesture at an outcome
1No result, or unrelated to the task
STAR (1–10) = average(Situation, Task, Action, Result) × 2 − penalty if Result weak

Answering the Question all question types

How directly and completely the answer responds to the exact question asked. This BARS drives the Relevance criterion (and is the dominant factor in it).

5Directly and fully answers the exact question; every part of a multi-part question addressed; no drift
4Answers well and stays on point; one sub-part light or a brief tangent
3Partially answers — misses part of the question, answers a near-miss version, or drifts then recovers
2Largely non-responsive — a loosely related story, or answers a different question
1Does not answer — off-topic, evasive, or a generic "I always…"
Experience-level calibration

Before starting, you pick a level: Individual Contributor, Manager, or Senior Leader. The competencies and questions stay the same — what changes is the anchor: the same competency requires greater scope, complexity, and impact to earn a high score at higher levels. This mirrors how professional leveled competency frameworks work.

Example — what a strong (5) Drive for Results answer looks like at each level:

LevelA "5" requires…
ICExcellent personal execution — an owned goal, concrete personal actions, a measurable result
ManagerDelivering through a team — balancing priorities, removing blockers, developing people; team-scope impact
Senior LeaderSetting direction and delivering through other leaders — organizational scope, ambiguity, senior-stakeholder alignment

An IC is never penalized for lacking organizational scope — the anchors calibrate down as well as up. Manager and Senior Leader levels also unlock three senior competencies in the default bank: Strategic Thinking, Leading Change & Transformation, and Building Organizational Capability. Each question's competency is tagged in your report.

Every criterion, explained

STAR + Outcome 45% on behavioral

Built from the four STAR-component BARS above. Rewards a clear situation and task, concrete first-person actions, and — most heavily — a specific, measurable result.

Relevance 25% behavioral · 35% opinion · others vary

Driven by the Answering the Question BARS. On behavioral questions it blends two sub-scores: how well you answered the question (80%) and how well the example fits the specific role per the job description (20%).

With a job description: Relevance = (0.8 × AnswersQuestion + 0.2 × JobFit) × 2 No job description: Relevance = AnswersQuestion × 2 (job fit can't be assessed) Opinion questions: Relevance = AnswersQuestion × 2

Both sub-scores are 1–5; the composite is computed in code (not by the model) so the weighting is exact.

Conciseness 10–40% by type

Three signals, in priority order:

  • Rambling & repetition (most important) — does the answer stay on point, or wander, repeat itself, and over-explain?
  • Time — measured from your response time (delivering the answer), with 3–5 minutes treated as ideal; ~6 min a mild penalty, ~7 min+ a larger one. Thinking time before you start is separate: up to ~60 seconds is free, beyond that a small penalty.
  • Word count — a supporting check (~450–750 words ≈ the 3–5 minute ideal).

Filler Words 10–20% by type

Counts verbal disfluencies (um, uh, er, hmm, and tic-words like "you know") and phrase fillers ("basically", "at the end of the day", "to be honest"). The report shows an explicit um/uh count and a per-token breakdown.

Note: when answering by voice, browsers often strip "um/uh" from the transcript automatically, so the count is most accurate for typed answers.

I vs. We 5% on behavioral

Rewards claiming personal ownership with "I" and flags "we / our team" phrasing that dilutes individual credit.

Company Knowledge 30% on motivation

Does the answer demonstrate specific knowledge of the role, company, industry, culture, or mission — naming real values or details — rather than generic "I want to grow" statements?

Substance 30% on opinion

Quality of reasoning on a philosophy or approach question — is the view specific, well-reasoned, and insightful, with clear principles, or vague and generic?

Real Examples 15% on motivation & opinion

Non-behavioral questions don't require a STAR story, but answers grounded in a concrete, real experience are stronger. This criterion gives partial credit for genuine examples — without penalizing heavily when none is given.

Content & Delivery

The criteria measure two different things, so we report them separately rather than blending them into one number — it's both more accurate and more useful, because it tells you what to work on.

ScoreWhat it covers
ContentSTAR structure, relevance, substance, real examples, company knowledge — what you said and how strongly it answers the question.
DeliveryConciseness, filler words, and personal ownership (I vs. We) — how you said it.

Each question shows its own Content and Delivery scores; your report rolls them up into two headline scores. Both are whole numbers (1–10) — we don't imply false precision with decimals.

Readiness band

Your overall performance is summarized as a readiness band — a developmental signal, not a prediction of whether you'd be hired. (We deliberately avoid a fabricated "offer likelihood" percentage, which would imply a predictive accuracy we haven't validated.)

BandReading
Interview-ReadyStrong, well-evidenced answers across the board
On TrackSolid foundation — a few areas to sharpen
DevelopingReal material to build on; focus on the flagged gaps
Needs WorkRebuild answers around specific examples and clear results

The band is derived from your scores and is meant to guide practice, not to judge you.

← Home