The Same Value Is an Outlier by One Rule and Not by Another
Gepubliceerd op 21-10-2026 · 4 min leestijd · Dagelijkse calculators
Daniel Okonkwo — Front-enddeveloper en techredacteur bij OneKitly
Webprestaties · Bestandsformaten
Getoetst aan 2 bronnen
Take the numbers 1 through 9 and add a 100. The interquartile rule puts the quartiles at 3.25 and 7.75, giving fences at −3.5 and 14.5, so the 100 is flagged without hesitation. The three-sigma rule disagrees: the mean is 14.5 and the standard deviation 30.15, so the 100 sits 2.84 deviations out — inside the usual threshold of three, and therefore not an outlier at all. Neither rule is wrong; they measure against different yardsticks, and the second one is being measured against a yardstick the suspect value itself stretched. Change the rule again to the MAD-based score and the same 100 comes out at 25.5, far beyond any threshold. An outlier is a decision, not a property of the number.
In 1 to 9 plus a 100, the IQR rule flags the 100 — the fences are −3.5 and 14.5. The three-sigma rule does not: its z-score is 2.84, because the 100 inflated the very deviation it is being measured against.
Masking: the suspect hides behind its own effect
The three-sigma rule fails here for a specific and predictable reason: the standard deviation it divides by was computed from data that includes the point under test. One extreme value inflates that denominator enough to make itself look ordinary, and two or three of them can hide each other completely — the phenomenon has a name, masking, and it is why a rule built on the mean and the standard deviation is a poor detector of the very thing it is meant to detect. Rules built on ranks do not have this weakness, because a single extreme value cannot inflate a quartile or a median.
Even the quartiles are not one thing
Statistical software offers half a dozen definitions of a quartile, and they disagree on small samples. This calculator interpolates linearly between order statistics, which puts the quartiles of a seven-value set at 11.5 and 12.5 where the older method of splitting at the median would put them at 11 and 13 — an interquartile range of 1 against 2, and fences twice as far apart. On the seven readings 10, 11, 12, 12, 12, 13 and 40 both conventions flag the 40, but on a borderline case they will not agree. Any pipeline that removes flagged points is therefore sensitive to a choice nobody wrote down, and the honest fix is to record which convention was used alongside the results.
| Rule | Score or fence | Verdict |
|---|---|---|
| 1.5 × IQR fences | −3.5 to 14.5 | Outlier |
| Three sigma | z = 2.84 | Not an outlier |
| Modified z (MAD) | 25.5 | Outlier |
Uitgewerkt met onze eigen calculator
Uitschieterdetector (IQR-methode)
Gegeven
- Getallen (door komma's gescheiden)
- 10, 12, 12, 13, 12, 11, 40
Resultaat
- Ondergrens
- 10
- Bovengrens
- 14
Deze cijfers komen uit de calculator hieronder, ze zijn niet met de hand ingetypt — ze worden opnieuw berekend telkens als de tool verandert.
Reken het door met je eigen cijfers →Veelgestelde vragen
- Should flagged points be deleted?
- Almost never automatically. A flag is a request to look, not a verdict: the value may be a transcription error, a genuine rare event, a different population accidentally included, or the most interesting thing in the dataset. Deleting on the strength of a rule removes real discoveries along with the mistakes, and it does so invisibly, because the deleted points leave no trace in the summary that follows. The defensible workflow is to keep the raw data, record which points were excluded and why, and be able to produce the analysis both ways when somebody asks.
- Why 1.5 and not some other multiple?
- Because it was chosen to be a useful compromise and then stuck. With normally distributed data the 1.5 × IQR fences leave roughly seven values in a thousand outside, which is rare enough to be worth investigating and common enough that the rule finds something in a real dataset. Three times the interquartile range is the conventional threshold for a far more extreme point. Neither number comes from a theorem, and both are adjustable inputs in serious work — a rule that flags a tenth of a large dataset is telling you about the distribution's shape rather than about individual points.
Artikelen die je misschien interesseren
Alle gidsen →Vergelijkbare tools
Bronnen
Een fout in dit artikel gezien?