Purpose
This measure is used to identify unusually high values within a dataset. Detecting such outliers helps highlight potential pricing anomalies, data quality issues, or exceptional business conditions that may require further analysis.
Business Context
Outliers can distort performance analysis and benchmarking results. This measure applies Tukey’s test, a widely recognized and statistically robust method, to determine whether values fall significantly above the expected range. The calculation is typically performed at an aggregated level and then used as a reference threshold to identify unusually high observations.
Definition
The measure calculates a threshold above which values are considered abnormally high. Using Tukey’s test, values are classified as too high if they exceed:
Q3 + 1.5 × IQR
Where:
-
Q3 = Third Quartile (75th percentile)
-
IQR = Interquartile Range (Q3 − Q1)
This can also be expressed as:
2.5 × Q3 − 1.5 × Q1
Business Formula
Threshold for Too High Values = 2.5 × Q3 − 1.5 × Q1
Agent Expression
2.5 * PERCENTILE_CONT(Metric, 0.75) - 1.5 * PERCENTILE_CONT(Metric, 0.25)