Built site for gh-pages

.nojekyll

@@ -1 +1 @@
-cf89b8a1
+a34b9df3

index.html

@@ -225,7 +225,7 @@ <h2 data-anchor-id="some-statistical-investigations">Some Statistical Investigat
 </tr>
 </thead>
 <tbody class="list">
-<tr data-index="0" data-listing-date-sort="1733266800000" data-listing-file-modified-sort="1733320606858" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="3" data-listing-word-count-sort="592" data-listing-title-sort="Heteroskedastic Linear Mixed Models with
+<tr data-index="0" data-listing-date-sort="1733266800000" data-listing-file-modified-sort="1733320753650" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="3" data-listing-word-count-sort="574" data-listing-title-sort="Heteroskedastic Linear Mixed Models with
 `glmmTMB` and the argument `dispformula`" data-listing-filename-sort="index.qmd">
 <td>
 <span class="listing-date">Dec 4, 2024</span>

posts/lmm-heteroskedastic/index.html

@@ -222,7 +222,6 @@ <h1>Abstract</h1>
 <li><code>glmmTMB</code> can model heteroskedastic data via the <code>dispformula</code> argument (c.f. section <a href="#sec-sim-dat-vis-w-disp" class="quarto-xref">Section&nbsp;3.3</a>).</li>
 <li>Type I error rate is slightly inflated. Regardless if modeled on homo-/heteroskedastic data accounting/ignoring heteroskedasticity with lmer and glmmTMB</li>
 </ul>
-<p><code>glmmTMB</code> models dispersion as expected. <strong>BUT</strong> the Type I error rate is inflated for both (with and without <code>dispformula</code>)!</p>
 <p><strong>Example</strong>:</p>
 <pre><code>glmmTMB(y ~ trt + (1|id), data = D, dispformula = ~trt)</code></pre>
 <p><strong>Note</strong>: Before modelling heteroskedastic LMM try:</p>

search.json

@@ -284,7 +284,7 @@
     "href": "posts/lmm-heteroskedastic/index.html",
     "title": "Heteroskedastic Linear Mixed Models with glmmTMB and the argument dispformula",
     "section": "",
-    "text": "glmmTMB can model heteroskedastic data via the dispformula argument (c.f. section Section 3.3).\nType I error rate is slightly inflated. Regardless if modeled on homo-/heteroskedastic data accounting/ignoring heteroskedasticity with lmer and glmmTMB\n\nglmmTMB models dispersion as expected. BUT the Type I error rate is inflated for both (with and without dispformula)!\nExample:\nglmmTMB(y ~ trt + (1|id), data = D, dispformula = ~trt)\nNote: Before modelling heteroskedastic LMM try:\n\nfixing heteroskedasticity by transforming the response variable (log, sqrt, etc.)\nSimplify mixed model structure by aggregating data like done in this post"
+    "text": "glmmTMB can model heteroskedastic data via the dispformula argument (c.f. section Section 3.3).\nType I error rate is slightly inflated. Regardless if modeled on homo-/heteroskedastic data accounting/ignoring heteroskedasticity with lmer and glmmTMB\n\nExample:\nglmmTMB(y ~ trt + (1|id), data = D, dispformula = ~trt)\nNote: Before modelling heteroskedastic LMM try:\n\nfixing heteroskedasticity by transforming the response variable (log, sqrt, etc.)\nSimplify mixed model structure by aggregating data like done in this post"
   },
   {
     "objectID": "posts/lmm-heteroskedastic/index.html#raw-heteroscedastic-data",

sitemap.xml

@@ -54,7 +54,7 @@
   </url>
   <url>
     <loc>https://lgraz.com/posts/lmm-heteroskedastic/index.html</loc>
-    <lastmod>2024-12-04T13:56:46.858Z</lastmod>
+    <lastmod>2024-12-04T13:59:13.650Z</lastmod>
   </url>
   <url>
     <loc>https://lgraz.com/posts/lmm-slope-aggregate/index.html</loc>