Erik-Jan<p>Was going through old repositories; found this nice document on estimating treatment effect of a medicine for a single patient with seizures <a href="https://fosstodon.org/tags/rstats" class="mention hashtag" rel="tag">#<span>rstats</span></a> :rstats:</p><p>It goes from simple to complex:<br />- <a href="https://fosstodon.org/tags/Visualization" class="mention hashtag" rel="tag">#<span>Visualization</span></a> &amp; summary statistics<br />- Poisson regression (glm) with a treatment dummy<br />- Penalized poisson regression to deal with 0s<br />- Correcting for autocorrelation<br />- <a href="https://fosstodon.org/tags/Bayesian" class="mention hashtag" rel="tag">#<span>Bayesian</span></a> Poisson model with <a href="https://fosstodon.org/tags/stan" class="mention hashtag" rel="tag">#<span>stan</span></a> <a href="https://fosstodon.org/tags/rstanarm" class="mention hashtag" rel="tag">#<span>rstanarm</span></a> <br />- Correcting for autocorrelation and <a href="https://fosstodon.org/tags/BayesFactors" class="mention hashtag" rel="tag">#<span>BayesFactors</span></a> <a href="https://fosstodon.org/tags/bridgesampling" class="mention hashtag" rel="tag">#<span>bridgesampling</span></a></p><p><a href="https://github.com/vankesteren/single_subject" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="ellipsis">github.com/vankesteren/single_</span><span class="invisible">subject</span></a></p>