Statistics 📈 Relationships & Regression

Regression Pitfalls

A correlation coefficient and a fitted line can both look convincing and still mislead you. Here are the traps — and the one habit that catches them all.

Intro StatisticsAP Statistics level
💡
The big idea: A single number like r, or a tidy line through a scatter, compresses a whole cloud of data into something that fits on a slide. That compression is the point — and the danger. The same r can come from wildly different pictures, a line trusted past the edge of the data can say anything, one stray point can swing the whole fit, and two things moving together need not be cause and effect. The fix for almost all of it is the cheapest habit in statistics: plot the data and look.
🎯 By the end, you'll be able to
  • Explain why a correlation between two variables does not establish that one causes the other, and name the role of a lurking (confounding) variable
  • Recognize extrapolation — using a fitted line outside the range of the data — and why it is unreliable
  • Describe how outliers and high-leverage points can distort a correlation or regression line
  • Understand that very different datasets can share the same r, so a scatterplot is essential
  • State the ecological fallacy: group-level associations need not hold for individuals
📎 You should already know
  • Correlation and the coefficient r
  • Least-squares regression lines

A number that hides as much as it shows

Correlation and regression are summaries. You feed in dozens or thousands of points and get back a single coefficient r, or the slope and intercept of one line. That is enormously useful — but a summary throws information away on purpose, and what it throws away is exactly where the traps hide.

This lesson collects four ways a perfectly real, correctly computed number can still lead you somewhere false: mistaking correlation for causation, trusting a line past the edge of the data, letting a single point steer the fit, and forgetting that r alone never tells you what the cloud actually looks like. Every one of them is caught by the same reflex — plot the data first.

🔑 Four traps, one habit

Watch for all four:

  • Correlation is not causation. A lurking variable can drive both.
  • Extrapolation. A line is only trustworthy inside the range of x you actually saw.
  • Outliers and leverage. One extreme point can move r and the whole line.
  • The same r, different pictures. A coefficient never describes the shape — only a plot does.

The habit that defends against all four: always plot the data before you trust a number.

Trap 1: correlation is not causation

When two variables rise and fall together, the tempting story is that one makes the other happen. But a correlation is perfectly consistent with three different worlds: x causes y, y causes x, or some third variable — a lurking variable (also called a confounder) — drives both while x and y do nothing to each other.

The classic case: across a summer, ice-cream sales and drowning deaths are strongly correlated. Ice cream does not cause drowning. Hot weather independently pushes both up — more cones sold, more people swimming. The correlation is real; the causal story is wrong. Correlation can be a clue that a causal link exists, but on its own it never proves one.

\[ r = \frac{\sum (x_i - \bar{x})(y_i - \bar{y})}{\sqrt{\sum (x_i - \bar{x})^2}\;\sqrt{\sum (y_i - \bar{y})^2}} \]
Pearson's correlation coefficient r measures the strength and direction of a LINEAR relationship only, on a scale from -1 to +1. It says nothing about causation, and nothing about non-linear patterns.

Trap 2 and 3: reaching past the data, and points that pull

Extrapolation is using a fitted line to predict outside the range of x values you actually collected. Inside that range you have evidence the linear pattern holds; outside it you are guessing that the same straight line keeps going — and real relationships bend, level off, or reverse. A line fit to the heights of children aged 2 to 12 would, extended blindly, predict absurd heights at age 40.

Individual points matter too. An outlier sits far from the overall pattern. A high-leverage point sits far from the other x-values — out at the left or right edge — and because the least-squares line pivots to reduce its residual, such a point can tilt the entire line and inflate or deflate r. One well-placed stray point can manufacture a correlation that the rest of the data does not support, or hide one that does.

🎮 Interactive: how loosely r summarizes a cloud LIVE
Dial a target r and resample. Notice that at moderate r (say 0.4 to 0.6) the cloud is a broad, shapeless smear — yet it collapses to a single number. Hit Resample repeatedly at a fixed target: each fresh cloud looks different, but the reported r barely moves. That gap between what r reports and what the data looks like is exactly why you must plot.
✨ The same r, very different data — so always plot

A famous demonstration (Anscombe's quartet) shows four datasets with identical means, identical standard deviations, and the same correlation r near 0.82 — yet one is a clean straight line, one is a smooth curve, one is a perfect line dragged off by a single outlier, and one is a vertical stack fooled by one far-off point. The summary numbers are indistinguishable; the pictures are nothing alike.

The lesson is blunt: r and the regression line describe a linear pattern only, and never reveal shape. A curve, a cluster, or a lone leverage point can all hide behind a respectable-looking coefficient. Plotting takes seconds and catches what the numbers cannot.

Trap 4 (briefly): the ecological fallacy

One more trap worth naming. A correlation measured on groups — states, schools, countries — need not hold for the individuals inside them. Concluding an individual-level relationship from group averages is the ecological fallacy. Regions with higher average income might show higher average spending on some good, yet within any single region richer people might spend less on it. Group-level and individual-level associations can even point in opposite directions, so be careful which one your data actually measures.

⚠️ Extrapolation quietly turns evidence into guesswork

A regression line carries no warning label at the edge of the data. It will happily return a prediction for any x you plug in, including values far beyond anything you observed. Those predictions look just as precise as the trustworthy ones — same equation, same decimals — but they rest on the untested assumption that the straight line continues. Treat any prediction outside your observed range of x with deep suspicion.

📝 Worked example: A researcher finds a strong positive correlation (r = 0.78) between the number of firefighters sent to a fire and the amount of property damage. She concludes that sending fewer firefighters would reduce damage. What has gone wrong, and how would you check?
  1. The correlation is almost certainly real: bigger fires do coincide with more property damage. The error is the causal leap from that correlation.
  2. Look for a lurking variable that drives both. The size of the fire is the obvious one: large fires cause more damage AND prompt dispatchers to send more firefighters. Firefighters are a response to fire size, not the cause of the damage.
  3. Reversing the logic exposes the absurdity: sending one firefighter to a warehouse blaze would not make the building burn less.
  4. How to check honestly: hold fire size roughly constant (compare fires of similar size) and see whether more firefighters still track with more damage. Controlling for the confounder typically makes the apparent effect shrink or flip.
✓ The r = 0.78 is genuine but does not mean firefighters cause damage. A lurking variable — fire size — drives both, so the causal conclusion is invalid. Correlation flagged a relationship worth investigating; only comparing like-sized fires (controlling for the confounder) can address the causal question.

Check your understanding

1. Across a summer, ice-cream sales and the number of drowning deaths are strongly positively correlated. What is the best explanation?
Hot weather is a confounder: it pushes up both ice-cream sales and swimming (hence drownings), while the two have no direct causal link. The correlation is real but not causal.
2. A line is fit to data on adults aged 20 to 40 and then used to predict a value for someone aged 75. Using the line this far outside the observed range of ages is called…
Predicting outside the range of x values you actually collected is extrapolation. The linear pattern was only ever verified inside the observed range, so predictions beyond it are unreliable.
3. Two datasets are found to have exactly the same correlation, r = 0.82. What can you safely conclude?
An identical r reflects only the same linear-association summary. As Anscombe's quartet shows, a line, a curve, and an outlier-driven cloud can all share the same r — which is why plotting is essential.
4. A single point sits far to the right of all the others in the x-direction and, on its own, tilts the least-squares line noticeably toward itself. This point is best described as…
A point far from the other x-values has high leverage: the fitted line pivots to reduce its residual, so one such point can swing the whole line and change r.
✅ Key takeaways
  • Correlation is not causation: two variables can move together because a lurking (confounding) variable drives both. Correlation is a clue, never a proof, of cause.
  • r measures the strength and direction of a LINEAR relationship only, on a -1 to +1 scale — it says nothing about curves, clusters, or cause.
  • Extrapolation — trusting a fitted line outside the observed range of x — turns evidence into guesswork; real relationships bend and reverse beyond the data.
  • Outliers and high-leverage points can single-handedly tilt a regression line and distort r.
  • The same r can come from completely different-looking data (Anscombe's quartet), so ALWAYS plot the data before trusting a number.
  • The ecological fallacy: an association seen across groups need not hold for the individuals within them.