Why Hypothesis Testing Should Take a Cue from Hamlet
<p>If you’re a <a href="http://bit.ly/quaesita_scientists" rel="noopener ugc nofollow" target="_blank">scientist</a> or <a href="http://bit.ly/quaesita_universe" rel="noopener ugc nofollow" target="_blank">data professional</a>, chances are that your <a href="http://bit.ly/quaesita_donttrust" rel="noopener ugc nofollow" target="_blank">hypothesis testing procedure</a> lacks a crucial step that’s tragically — or tragicomically? — omitted from your typical coursework. Never fear, in this blog post I’ll show you the missing piece and why you’ll find the fix in the thespian’s playbook.</p>
<h1>Act One, Scene One</h1>
<p>The scene opens with you triumphantly scoring the budget to <a href="http://bit.ly/quaesita_srstrees2" rel="noopener ugc nofollow" target="_blank">go collect some actual data</a>. Perhaps it’ll all be digital; you’re off to tell your engineering team which <a href="http://bit.ly/quaesita_gistlist" rel="noopener ugc nofollow" target="_blank">variables</a> to start logging or which online <a href="http://bit.ly/quaesita_ab" rel="noopener ugc nofollow" target="_blank">experiments</a> to run. Or maybe you’re stepping out into the physical world to set some sensors, prep some pipettes, or whatever else it takes to get your hands on data. (Curious about the practical side of taking measurements from the real world? Check out my article about <a href="http://bit.ly/quaesita_srstrees1" rel="noopener ugc nofollow" target="_blank">sampling trees</a>.)</p>
<p><a href="https://medium.com/towards-data-science/why-hypothesis-testing-should-take-a-cue-from-hamlet-1e714733cec3"><strong>Read More</strong></a></p>