Graphl
LearnLibraryPlayPracticeTools
Go Pro
Go ProLearnLibraryPlayPracticeToolsAccount

Graphl

Economic models you can actually see move. Every curve is computed and every shift is verified. Built for students who want intuition, not just diagrams.

Explore

  • Model library
  • Courses: HSC to university
  • HSC Economics track
  • Graphdle (daily puzzle)
  • Graph builder
  • Compare models
  • Learn
  • Economics glossary
  • Live data

Tools

  • Exam mode
  • Which model do I use?
  • Your dashboard
  • Sign in
  • Pricing & Pro
  • Settings
  • Take the site tour

Company

  • About & contact
  • Terms of use
  • Privacy
  • Refunds

Educational purposes only, not financial advice. Graphl teaches which economic framework applies and how its mechanism works; it does not solve assessment problems for you.

  1. Library
  2. /
  3. Linear Regression & OLS
  4. /
  5. Scenario

Econometrics · intro

A data-entry error plants a wild outlier

What moves on the Linear Regression & OLS diagram

  • Fitted line ŷ = a + bXshifts up (small)

Squared loss makes OLS chase extreme points: a single high-leverage outlier can rotate the entire fitted line. Robust checks and plots exist for exactly this reason.

Watch it animateQuiz yourself on Linear Regression & OLS

Step by step

  1. 1

    OLS minimizes SQUARED errors, so big misses dominate the fit.

  2. 2

    One mistyped observation lands far from the cloud.

  3. 3

    The whole line tilts toward the outlier, dragged by its enormous squared residual.

  4. 4

    One bad point moved every prediction. Always plot your data before you trust your regression.

Where this shows up

Examiners ask this as: outlier, data error, influential observation, typo in data, leverage point.

Other scenarios on Linear Regression & OLS

Controlling for ability flattens the slope →