Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Parallel analysis
open-in-new
<p>Parallel analysis, also known as Horn's parallel analysis, is a statistical method used to determine the number of components to keep in a <a href="/facts/Principal_component_analysis/pCbVTqdR">principal component analysis</a> or factors to keep in an <a href="/facts/Exploratory_factor_analysis/8kzsWtPf">exploratory factor analysis</a>. It is named after psychologist <a href="/facts/John_L._Horn/ML2Hdokb">John L. Horn</a>, who created the method, publishing it in the journal <i><a href="/facts/Psychometrika/ES7QkLzI">Psychometrika</a></i> in 1965. The method compares the <a href="/facts/Eigenvalues/8TjEoT8u">eigenvalues</a> generated from the data matrix to the eigenvalues generated from a <a href="/facts/Monte-Carlo_simulation/AkHjY7jc">Monte-Carlo simulated</a> matrix created from random data of the same size. </p>