CPG Demo · Beer Concept Test (n=803)

Non-Response Stress Test

Stress test on the published headline

Not everyone invited responds to a survey. Could the people who didn't respond have changed the headline number? Adjust the sliders to find out.

Specify a nonresponse pattern and see whether the 54.7% top-2 purchase intent would survive it. This is a vulnerability check on the released score, separate from the explanatory model.

Based on the beer concept-test consumer survey of 803 respondents.

Published model: All-or-Nothing Simulator · Fine-Tuning Simulator | Inspect & challenge: Model Builder · Survey Explorer

How to Use This Tool

Set the sliders and click Run Simulation. Expand for details.

1. The Headline

The headline being stress-tested is top-2 purchase intent — the share of respondents who rated their likelihood to buy as Extremely or Very likely after seeing the package concept. The reported figure is 54.7%. The question is: could nonresponse have changed that number?

2. Set the Sliders

Two sliders define your scenario. How many non-respondents to assume? sets how many of those invited you imagine chose not to participate. What would they have said? sets what share of those people would have given top-2 purchase intent. A live preview updates as you drag.

3. Run the Simulation

Click Run Simulation. The tool recalculates the headline as if your hypothetical non-respondents had actually answered, and shows an uncertainty range around the new figure — so you can see both the central estimate and how much it could realistically vary.

4. Read the Verdict

The results section asks: could non-response have changed the headline? It compares the shift in your scenario to the survey's margin of error, rates the threat level, and flags the most serious outcome — a majority flip where the conclusion reverses entirely (e.g., majority interest becomes minority interest).

More options

Save and compare

Each run is saved automatically. Expand the Pinned Scenarios drawer at the bottom to compare previous runs side by side.

Reset and iterate

Click Reset to restore defaults. Try fixing one slider and sweeping the other to map where the headline would change.

Base

—% (± —)

What the poll reported

Simulated

—% (± —)

If those non-respondents had answered

You replaced — respondents (—% of sample)

Could non-response have changed the headline?

How big is the shift?

Non-response effect —× the poll's margin of error

Direction of change —

Magnitude of change — points

Verdict

Risk level —

The headline is — under this scenario.

The difference between the headline number and your scenario —.

Detailed Analysis

Typical headline: the outcome at —% (±—)

Simulated estimate: —% (95% CI: —–—%; MoE ±—%)

You replaced —% of the sample with a group that had —% approval (gap: — pts).

Why margin of error? We benchmark against the poll's reported margin of error (MoE) because that's the yardstick the public sees in headlines. A "misleading" verdict here means your scenario's shift exceeds the reported MoE.

Base vs. simulated rate

How certain is this result? (10,000 simulated outcomes)

Base Simulated Overlap

The green curve shows 10,000 possible baseline rates; the blue curve shows 10,000 possible rates under your scenario. When the curves barely overlap, your scenario shift is a real signal. When they overlap heavily, the difference could be due to chance.