P-Hacking

The term "p-hacking" refers to the process by which some researchers manipulate their data until they get a "statistically significant" result. A result is statistically significant if it achieves a p-value of less than 5%, meaning that fewer than 5% of the randomly permuted situations experienced the same outcome as that observed in the data.


The Situation

You have been hired to determine whether or not there is a link between how a state voted in the 2016 presidential election and some other outcomes related to hate crimes. You decide to run an A/B Test on your variable of interest to determine whether or not the two groups come from the same underlying distribution.

See if you can get a significant result below by toying with the test parameters.


App designed by Chris Pyles. Data from FiveThirtyEight on Github.