Replication
Repeating a test on fresh data to check that the first result wasn't a fluke.
In plain English
One good result can be luck. If the effect is real, it should show up again on data the test has never seen. If it disappears, it probably was luck.
Uncle Gary explains
Kevin bowled a strike on his very first throw and announced a professional career. Second throw: gutter. Third throw: gutter. Replication ended the career.
Why it matters for your money
Our AI news test looked promising on the first 299 stories and passed a standard statistical test. On 300 fresh stories, the effect vanished completely.
Where we tested it
- Can AI Predict Stock Prices From the News? We Asked It 599 Times
- Does Volume Confirm the Trend? Sometimes It Contradicts It
Related terms
All glossary terms. Educational content, not financial advice.