Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Misuse of p-values
open-in-new
<p>Misuse of <i>p</i>-values is common in <a href="/facts/Scientific_research/182j8RWE">scientific research</a> and <a href="/facts/Scientific_education/lxLmFxcq">scientific education</a>. <a href="/facts/P-value/InnYd1bz"><i>p</i>-values</a> are often used or interpreted incorrectly; the <a href="/facts/American_Statistical_Association/QzF1h4Dx">American Statistical Association</a> states that <i>p</i>-values can indicate how incompatible the data are with a specified statistical model. From a <a href="/facts/Foundations_of_statistics/oX0dUhfP">Neyman–Pearson hypothesis testing approach</a> to statistical inferences, the data obtained by comparing the <i>p</i>-value to a <a href="/facts/Significance_level/PMxjdlAu">significance level</a> will yield one of two results: either the <a href="/facts/Null_hypothesis/C8NwSE9J">null hypothesis</a> is rejected (which however does not prove that the null hypothesis is <i>false</i>), or the null hypothesis <i>cannot</i> be rejected at that significance level (which however does not prove that the null hypothesis is <i>true</i>). From a <a href="/facts/Foundations_of_statistics/oX0dUhfP">Fisherian statistical testing approach</a> to statistical inferences, a low <i>p</i>-value means <i>either</i> that the null hypothesis is true and a highly improbable event has occurred <i>or</i> that the null hypothesis is false. </p>