<- Back to laws

Statistical selection effect

Regression to the Mean

When two measurements are imperfectly correlated, units selected for an extreme first value tend to be less extreme on a subsequent measurement.

Scientific statusStatistical regularity
Predictive formConditional expectation under imperfect correlation
DomainRepeated measurement
EvidenceProbability + empirical studies
Key limitationRequires a stable comparison process
Common misuseCalling every reversal regression
INTERACTIVE MODEL

E[Z2 | Z1 = z] = r z when standardized in a linear model

The shrinkage factor is the correlation r, not a force pulling observations toward average. Real change, measurement error, and selection can coexist.

Extreme first scores send translucent paths toward expected retests; stronger reliability preserves more of the original extremity.

74.0Expected retest percentile index
()
0 %100 %
EXTREME-SCORE RETEST MACHINENoise selected on the first measurement partially disappears on the second.
Interactive visual model for Regression to the Mean.
VISIBLE PHASESTARTINGTAKEAWAYWATCH ONE FULL CYCLE

The animation runs automatically, pauses on the conclusion, and then repeats. The main control changes the scenario rather than scrubbing the timeline.

CHANGE
Test-retest reliability
WATCH
expected retest
MEANING
Extreme first scores send translucent paths toward expected retests; stronger reliability preserves more of the original extremity.
VISUAL MODEL

Select on an extreme, then watch imperfect correlation reveal itself.

A first-test versus retest scatterplot highlights the selected tail and its conditional average below the identity line.

selection thresholdidentity lineconditional expectation
01 / MEANING

What it actually says

Regression to the mean follows from conditional selection when repeated outcomes contain both persistent and nonpersistent components. It needs no compensating mechanism.

It is a central threat in before-and-after studies: patients, schools, athletes, or firms are often chosen precisely because the baseline was unusually bad or good.

Compact formE[Z2 | Z1 = z] = r z when standardized in a linear model
Best interpretationRepeated measurement evidence in statistics.
Important cautionRequires a stable comparison process.
"A useful law compresses a pattern. It does not erase the conditions that make the pattern true."
02 / ORIGIN

How the idea developed

The modern form emerged through observation, argument, and later refinement. The timeline separates the first insight from the version now used in textbooks and practice.[1]

18861886

Francis Galton reports regression in hereditary stature.

20th century20th century

The concept becomes foundational in correlation and experimental design.

TodayToday

Trials and causal studies use controls to separate treatment effects from regression.

Historical cautionEponymous laws often change after their first publication. Popular wording may be broader and cleaner than the original evidence.
03 / MECHANISM

How the pattern works

The relation becomes useful only when its mechanism, measurement process, and operating range are visible.

01Extreme selection

Cases enter because the first measure is unusual.

02Imperfect correlation

Transient components do not repeat fully.

03Conditional averaging

Expected follow-up is closer to the population mean.

MODELE[Z2 | Z1 = z] = r z when standardized in a linear model

The shrinkage factor is the correlation r, not a force pulling observations toward average. Real change, measurement error, and selection can coexist.

04 / APPLICATIONS

Where it earns its keep

Applications are strongest when the law changes a decision, measurement, model, or experiment rather than merely providing an analogy.

EXPERIMENTS

Use a comparison group

Application

Controls experience the same selection and time process.

PROFESSIONAL NOTE

Avoid naive before-after attribution.

OPERATIONS

Do not punish or praise noise

Application

Extreme performance often moderates naturally.

PROFESSIONAL NOTE

Estimate reliability.

05 / LIMITS & MISUSE

Where it stops working

Population drift, treatment, learning, seasonality, and heterogeneous subgroups can also change follow-up values and require separate modeling.

Misuse

"The mean causes the change"

Better: Regression is a conditional pattern, not a restoring force.
Misuse

"Every extreme must reverse"

Better: Individual outcomes can become more extreme; the statement concerns expectation.
07 / REFERENCES

Sources and further reading

Original publications and serious secondary scholarship are prioritized over summaries.

  1. Galton - Regression Towards Mediocrity in Hereditary StaturePrimary 1886 paper.https://galton.org/essays/1880-1889/galton-1886-jaigi-regression-stature.pdf
  2. Barnett, van der Pols, and Dobson - Regression to the MeanMethodological review.https://doi.org/10.1093/ije/dyh299
  3. BMJ - Statistics Notes: Regression Towards the MeanClinical research explanation.https://www.bmj.com/content/308/6942/1499
CONTINUE EXPLORING

Related laws, with the relationship made explicit.

These are editorial connections, not claims that the laws are mathematically equivalent.

CONTINUE READING

Place this law inside the collection.

LAW 088 / 100 PUBLISHED