softlink

Author: ehsanx

_freeze/surveydata5/execute-results/html.json

@@ -1296,6 +1296,7 @@ <h2 id="toc-title">Table of contents</h2>
   <li><a href="#nhanes-predicting-cholesterol-level" id="toc-nhanes-predicting-cholesterol-level" class="nav-link" data-scroll-target="#nhanes-predicting-cholesterol-level">NHANES: Predicting cholesterol level</a></li>
   <li><a href="#nhanes-properly-subsetting-a-design-object" id="toc-nhanes-properly-subsetting-a-design-object" class="nav-link" data-scroll-target="#nhanes-properly-subsetting-a-design-object">NHANES: Properly subsetting a design object</a></li>
   <li><a href="#nhanes-reliability-standards" id="toc-nhanes-reliability-standards" class="nav-link" data-scroll-target="#nhanes-reliability-standards">NHANES: Reliability Standards</a></li>
+  <li><a href="#nhanes-model-performance" id="toc-nhanes-model-performance" class="nav-link" data-scroll-target="#nhanes-model-performance">NHANES: Model performance</a></li>
   </ul></li>
   </ul>
 </nav>
@@ -1380,6 +1381,10 @@ <h3 class="anchored" data-anchor-id="nhanes-properly-subsetting-a-design-object"
 <section id="nhanes-reliability-standards" class="level3">
 <h3 class="anchored" data-anchor-id="nhanes-reliability-standards"><a href="surveydata9.html">NHANES: Reliability Standards</a></h3>
 <p>This tutorial guides users in replicating key tables from a published article using NHANES data, emphasizing NCHS/CDC reliability standards for complex survey analyses. At first, it created a survey design object with weights, strata, and clusters. For Table 1, the <code>svytable1</code> package generates stratified descriptive summaries in “mixed mode” (unweighted N alongside weighted percentages), stratified by race/ethnicity for overall, male, and female subgroups, with automatic suppression (*) of unreliable estimates based on established rules, ensuring transparency on precision and design effects. Table 3 is reproduced via gender-stratified <code>svyglm</code> logistic regressions, incorporating diagnostics from <code>svyglmdiag()</code> (SE, p-values, CIs) and VIF checks for multicollinearity; results are formatted into a publication-ready table showing odds ratios and 95% CIs.</p>
+</section>
+<section id="nhanes-model-performance" class="level3">
+<h3 class="anchored" data-anchor-id="nhanes-model-performance"><a href="surveydata9c.html">NHANES: Model performance</a></h3>
+<p>This tutorial demonstrates how to evaluate a survey-weighted logistic regression model in R using NHANES data. It guides the user through preparing the data, fitting a design-based logistic regression model to predict obesity from demographic variables, and then assessing its performance. The evaluation uses two primary methods: calculating a design-correct Area Under the Curve (AUC) to measure predictive accuracy and running an Archer-Lemeshow test to assess goodness-of-fit.</p>
 <div class="callout callout-style-default callout-tip callout-titled">
 <div class="callout-header d-flex align-content-center">
 <div class="callout-icon-container">

docs/search.json

@@ -1324,6 +1324,19 @@ <h1 class="title">CCHS: Performance</h1>
 
 
 </header><p>The tutorial outlines the process for evaluating the performance of logistic regression models fitted to complex survey data using R. It focuses on two major aspects: creating Receiver Operating Characteristic (ROC) curves and conducting Archer and Lemeshow Goodness of Fit tests. Here AUC is a measure to evaluate the predictive accuracy of the model, and Archer and Lemeshow test is a statistical test to evaluate how well your model fits the observed data.</p>
+<div class="callout callout-style-default callout-important callout-titled">
+<div class="callout-header d-flex align-content-center">
+<div class="callout-icon-container">
+<i class="callout-icon"></i>
+</div>
+<div class="callout-title-container flex-fill">
+Important
+</div>
+</div>
+<div class="callout-body-container callout-body">
+<p>See an updated version of this tutorial <a href="surveydata9c.html">here</a> that uses <code>svyTable1</code> package on an NHANES data analysis.</p>
+</div>
+</div>
 <p>We start by importing the required R packages.</p>
 <div class="cell">
 <div class="sourceCode cell-code" id="cb1"><pre class="sourceCode numberSource r number-lines code-with-copy"><code class="sourceCode r"><span id="cb1-1"><a href="#cb1-1"></a><span class="co"># Load required packages</span></span>

docs/surveydata.html

@@ -50,7 +50,11 @@ The tutorial provides a comprehensive guide on how to handle and analyze a subse
 
 ### [NHANES: Reliability Standards](surveydata9.html)
 
-This tutorial guides users in replicating key tables from a published article using NHANES data, emphasizing NCHS/CDC reliability standards for complex survey analyses. At first, it created a survey design object with weights, strata, and clusters. For Table 1, the `svytable1` package generates stratified descriptive summaries in "mixed mode" (unweighted N alongside weighted percentages), stratified by race/ethnicity for overall, male, and female subgroups, with automatic suppression (*) of unreliable estimates based on established rules, ensuring transparency on precision and design effects. Table 3 is reproduced via gender-stratified `svyglm` logistic regressions, incorporating diagnostics from `svyglmdiag()` (SE, p-values, CIs) and VIF checks for multicollinearity; results are formatted into a publication-ready table showing odds ratios and 95% CIs. 
+This tutorial guides users in replicating key tables from a published article using NHANES data, emphasizing NCHS/CDC reliability standards for complex survey analyses. At first, it created a survey design object with weights, strata, and clusters. For Table 1, the `svytable1` package generates stratified descriptive summaries in "mixed mode" (unweighted N alongside weighted percentages), stratified by race/ethnicity for overall, male, and female subgroups, with automatic suppression (\*) of unreliable estimates based on established rules, ensuring transparency on precision and design effects. Table 3 is reproduced via gender-stratified `svyglm` logistic regressions, incorporating diagnostics from `svyglmdiag()` (SE, p-values, CIs) and VIF checks for multicollinearity; results are formatted into a publication-ready table showing odds ratios and 95% CIs.
+
+### [NHANES: Model performance](surveydata9c.html)
+
+This tutorial demonstrates how to evaluate a survey-weighted logistic regression model in R using NHANES data. It guides the user through preparing the data, fitting a design-based logistic regression model to predict obesity from demographic variables, and then assessing its performance. The evaluation uses two primary methods: calculating a design-correct Area Under the Curve (AUC) to measure predictive accuracy and running an Archer-Lemeshow test to assess goodness-of-fit. 
 
 ::: callout-tip
 **Optional Content**:

docs/surveydata5.html

@@ -2,6 +2,10 @@
 
 The tutorial outlines the process for evaluating the performance of logistic regression models fitted to complex survey data using R. It focuses on two major aspects: creating Receiver Operating Characteristic (ROC) curves and conducting Archer and Lemeshow Goodness of Fit tests. Here AUC is a measure to evaluate the predictive accuracy of the model, and Archer and Lemeshow test is a statistical test to evaluate how well your model fits the observed data.
 
+::: callout-important
+See an updated version of this tutorial [here](surveydata9c.html) that uses `svyTable1` package on an NHANES data analysis.
+:::
+
 We start by importing the required R packages.
 
 ```{r setup, warning=FALSE, message=FALSE, cache=TRUE}