You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
@@ -188,11 +184,11 @@ <h2 class="anchored" data-anchor-id="the-problem-with-parameters">The Problem wi
188
184
<p><spanclass="math inline">\(R^2\)</span> is a common metric that incentives naive scientists to keep adding variables. <spanclass="math inline">\(R^2\)</span> is defined as:</p>
<p><spanclass="math inline">\(R^2\)</span> always increases as more variables are added even when you just add random numbers which have no relation to the outcome.</p>
@@ -201,11 +197,11 @@ <h2 class="anchored" data-anchor-id="the-problem-with-parameters">The Problem wi
201
197
<sectionid="brain-body-vignette" class="level3">
202
198
<h3class="anchored" data-anchor-id="brain-body-vignette">Brain Body Vignette</h3>
203
199
<p>Plotted below is the relationship between the body size of several primate species and their corresponding brain size. There’s not a strong <em>a priori</em> reason to think body and brain size are perfectly linearly related. The true relationship between brain and body size could be any number of polynomial or log’d functions. Let’s go through a couple to see what’s gained and what’s lost with trying ever more complex functions to fit the data.</p>
0 commit comments