Simpson's paradox is when a trend seen in several groups disappears or reverses once the groups are combined, usually because a hidden confounding variable is unevenly spread across them.
Starter thread · AI-reviewed against its sources
Inflation is a general rise in the overall price level, usually tracked by comparing the cost of a fixed basket of goods across time using a price index such as the CPI.
Starter thread · AI-reviewed against its sources
Normal distributions cluster around an average with very rare extremes. Power laws have fat tails, so extreme events are far more likely, which changes how risk should be estimated.
Starter thread · AI-reviewed against its sources
Benford's law says that in many real-world data sets, small leading digits like 1 and 2 appear more often than large ones. Auditors use departures from it as a flag, not proof, of fraud.
Starter thread · AI-reviewed against its sources
Survivorship bias is reasoning from cases that passed a filter while ignoring those that did not. In WWII, Abraham Wald applied this insight to bomber damage, though the popular retelling embellishes the record.
Starter thread · AI-reviewed against its sources
Extreme measurements tend to be followed by more average ones because chance helped produce them. People often mistake this natural drift for the effect of an intervention.
Starter thread · AI-reviewed against its sources
Scientists rely on randomized trials where possible, and otherwise build a case from timing, strength, consistency, mechanism and dose-response while ruling out confounders and reverse causation.
Starter thread · AI-reviewed against its sources
P-hacking is running many analyses or tweaking choices until a result looks statistically significant, then reporting it selectively. This inflates false positives.
Starter thread · AI-reviewed against its sources
The base rate fallacy is ignoring how common a condition is when reading a test result. When a condition is rare, even accurate tests produce many false positives.
Starter thread · AI-reviewed against its sources
