# Instructor guide: Correlated uncertainty investigation

Course: Measurement, uncertainty, and evidence

Suggested time: 35–45 minutes

## Learning target

Learners distinguish independent quadrature from correlated uncertainty and defend a comparison using controlled variables.

## Prepare

- Open the uncertainty composer on every device.
- Choose one familiar unit such as centimeters.
- Prepare one deliberately flawed comparison for critique.

## Facilitation moves

- Ask which variable changed before discussing the number.
- Have pairs reproduce one another's recorded setting.
- Require units in every spoken conclusion.

## Misconception checks

- **More measurements always remove uncertainty.** Repeated independent noise may shrink; shared bias and correlation do not disappear by repetition.
- **Negative correlation means bad data.** It describes opposing co-movement and can reduce a particular combined uncertainty.

## Accessibility and participation

- Read every Greek symbol aloud and pair it with its plain-language name.
- Allow keyboard arrow control of every slider.
- Offer the observation table before learners begin changing values.

## Evidence of learning

- A controlled two-run comparison
- Correct use of units
- A conclusion that names the role of correlation

## Extension

Build a three-input budget and identify which covariance term dominates the output.

## Evidence boundary

Assess the learner's reasoning only within the declared model and recorded observations. Do not upgrade a simulation result into a claim about an unmodeled physical system.