Mistake Master
Student view — seeing the site as a student does
Home› Unit 5 · Regression Analysis› 5.1·5.2·5.3·5.4·5.5

Linear Regression Models

▶︎  Watch it animatedinteractive step-through · ~3 min · optional

A linear regression model is $\hat{y} = b_0 + b_1 x$, where $\hat{y}$ is the predicted response. For the eight students, $\widehat{\text{score}} = 58.5 + 3.5(\text{hours})$. The slope is read as a predicted average change: for each additional hour of weekly study, the predicted score rises by 3.5 points, on average. The intercept is the predicted response at $x = 0$, meaningful only when that value is near the observed data. A prediction at 6 hours is $58.5 + 3.5(6) = 79.5$ points, and predictions hold only over the observed range of 1 to 8 hours.

The slope gets read deterministically and causally: each extra hour makes a student score 3.5 points higher, with no predicted and no on average, and sometimes with the units flipped. The intercept gets declared impossible and the model rejected, when a wolf of length zero weighing $-14$ kg is an extrapolation rather than a defect. And the line gets pushed past its data, returning 128.5 points at 20 hours of study from a data set covering 1 to 8, where the arithmetic works and nothing supports it.

score-hat = 58.5 + 3.5(hours) +1 hour +3.5 points 6 hours predicted 79.5 1 8 slope: PREDICTED score rises 3.5 points per extra hour, ON AVERAGE two students at 6 hours will not both score 79.5
The slope triangle is the interpretation drawn: one unit across, 3.5 up. The points scattered around the line are why the sentence needs both predicted and on average.
extrapolation: the arithmetic works, the prediction does not maximum score 100 data: 1 to 8 hours no data out here 128.5 at 20 hours the line returns 128.5 points, above any possible score
The line does not know where the data stopped. Reporting the observed range alongside the model is what tells a reader where the prediction rule applies.

The work

3 ways in · any order
Lesson
Linear Regression Models ›

Reads a fitted regression line: the slope as a predicted average change with units on both variables, the intercept as the predicted response at zero and when that is meaningless, predictions inside the observed range, and why extrapolation returns arithmetic without support.

Skill check · 10 scenarios
Open lesson →
Diagnostic
10-item topic check ›

Ten items on regression models: slopes read as causal and exact, on average dropped, units flipped, intercepts declared impossible, and predictions pushed far outside the data. Take it cold to find your habit, or after the lesson to check it is gone.

Not started · 10 items · ~15 min
Start diagnostic →
Targeted Practice
Drill a single misconception ›

Pick one of the failure modes you missed and drill it on its own. The round is adaptive: two correct in a row clears it for now and moves you to the next. Two in a row is a checkpoint, not proof: if the error resurfaces later, the misconception comes back.

Take the diagnostic to identify your misconceptions
Diagnostic first