diff --git a/paper/rsabsence.Rnw b/paper/rsabsence.Rnw index 556f1496953a3d9b72c5dfb811752e1f8879be00..9944c1379690f5e7effbcc876ba470ef4dc6930b 100755 --- a/paper/rsabsence.Rnw +++ b/paper/rsabsence.Rnw @@ -8,10 +8,14 @@ \usepackage{doi} \usetikzlibrary{decorations.pathreplacing,calligraphy} % for tikz curly braces \usepackage{todonotes} +\usepackage{nameref} +\usepackage{caption} \definecolor{darkblue2}{HTML}{273B81} \definecolor{darkred2}{HTML}{D92102} +\fboxsep=20pt % for Box + \title{Replication of ``null results'' -- Absence of evidence or evidence of absence?} @@ -861,68 +865,71 @@ or absence of an effect. It seems logically questionable to declare an inconclusive replication of an inconclusive original study as a replication success. While it is important to replicate original studies with null results, our analysis highlights that they should be analyzed and interpreted -appropriately. Table~\ref{tab:recommendations} summarizes our recommendations. +appropriately. Box~\hyperref[box:recommendations]{1} summarizes our +recommendations. -\begin{table}[!htb] +\begin{table}[!h] \centering - \caption{Recommendations for the analysis of replication studies of original - null results. Calculations are based on effect estimates $\hat{\theta}_{i}$ - with standard errors $\sigma_{i}$ for $i \in \{o, r\}$ from an original - study (subscript $o$) and its replication (subscript $r$). Both effect - estimates are assumed to be normally distributed around the true effect size - $\theta$ with known variance $\sigma^{2}$. The effect size $\theta_{n}$ - represents the value of no effect, typically $\theta_{n} = 0$.} - \label{tab:recommendations} - \begin{tabular}{p{0.95\textwidth}} - \toprule - \textbf{Equivalence test} - \begin{enumerate} - \item Specify a margin $\Delta > 0$ that defines an equivalence range - $[\theta_{n} - \Delta, \theta_{n} + \Delta]$ in which effects are - considered absent for practical purposes. - \item Compute the TOST $p$-values for original and replication data - $$p_{\text{TOST}i} - = \max\left\{\Phi\left(\frac{\hat{\theta}_{i} - \theta_{n} - \Delta}{\sigma_{i}}\right), + \caption*{Box 1: Recommendations for the analysis of replication studies of + original null results. Calculations are based on effect estimates + $\hat{\theta}_{i}$ with standard errors $\sigma_{i}$ for $i \in \{o, r\}$ + from an original study (subscript $o$) and its replication (subscript $r$). + Both effect estimates are assumed to be normally distributed around the true + effect size $\theta$ with known variance $\sigma^{2}$. The effect size + $\theta_{n}$ represents the value of no effect, typically $\theta_{n} = 0$.} + \label{box:recommendations} + \fbox{ + \begin{tabular}{p{0.875\textwidth}} + % \toprule + \textbf{Equivalence test} + \begin{enumerate} + \item Specify a margin $\Delta > 0$ that defines an equivalence range + $[\theta_{n} - \Delta, \theta_{n} + \Delta]$ in which effects are + considered absent for practical purposes. + \item Compute the TOST $p$-values for original and replication data + $$p_{\text{TOST}i} + = \max\left\{\Phi\left(\frac{\hat{\theta}_{i} - \theta_{n} - \Delta}{\sigma_{i}}\right), 1 - \Phi\left(\frac{\hat{\theta}_{i} - \theta_{n} + \Delta}{\sigma_{i}}\right)\right\}, - ~ i \in \{o, r\}$$ - with $\Phi(\cdot)$ the cumulative distribution function of the - standard normal distribution. - \item Declare replication success at level $\alpha$ if - $p_{\text{TOST}o} \leq \alpha$ and $p_{\text{TOST}r} \leq \alpha$, - conventionally $\alpha = 0.05$. - \item Perform a sensitivity analysis with respect to the margin $\Delta$. - For example, visualize the TOST $p$-values for different margins to - assess the robustness of the conclusions. - \end{enumerate} \\ - \midrule \textbf{Bayes factor} - \begin{enumerate} - \item Specify a prior distribution for the effect size $\theta$ that - represents plausible values under the alternative hypothesis that - there is an effect ($H_{1}\colon \theta \neq \theta_{n})$. For - example, specify the mean $m$ and variance $v$ of a normal - distribution $\theta \given H_{1} \sim \Nor(m ,v)$. - \item Compute the Bayes factors contrasting - $H_{0} \colon \theta = \theta_{n}$ to - $H_{1} \colon \theta \neq \theta_{n}$ for original and replication - data. Assuming a normal prior distribution, - % $\theta \given H_{1} \sim \Nor(m ,v)$, - the Bayes factor is - $$\BF_{01i} - = \sqrt{1 + \frac{v}{\sigma^{2}_{i}}} \, \exp\left[-\frac{1}{2} \left\{\frac{(\hat{\theta}_{i} - - \theta_{n})^{2}}{\sigma^{2}_{i}} - \frac{(\hat{\theta}_{i} - m)^{2}}{\sigma^{2}_{i} + v} + ~ i \in \{o, r\}$$ + with $\Phi(\cdot)$ the cumulative distribution function of the + standard normal distribution. + \item Declare replication success at level $\alpha$ if + $p_{\text{TOST}o} \leq \alpha$ and $p_{\text{TOST}r} \leq \alpha$, + conventionally $\alpha = 0.05$. + \item Perform a sensitivity analysis with respect to the margin $\Delta$. + For example, visualize the TOST $p$-values for different margins to + assess the robustness of the conclusions. + \end{enumerate} \\ + % \midrule + \textbf{Bayes factor} + \begin{enumerate} + \item Specify a prior distribution for the effect size $\theta$ that + represents plausible values under the alternative hypothesis that + there is an effect ($H_{1}\colon \theta \neq \theta_{n})$. For + example, specify the mean $m$ and variance $v$ of a normal + distribution $\theta \given H_{1} \sim \Nor(m ,v)$. + \item Compute the Bayes factors contrasting + $H_{0} \colon \theta = \theta_{n}$ to + $H_{1} \colon \theta \neq \theta_{n}$ for original and replication + data. Assuming a normal prior distribution, + % $\theta \given H_{1} \sim \Nor(m ,v)$, + the Bayes factor is + $$\BF_{01i} + = \sqrt{1 + \frac{v}{\sigma^{2}_{i}}} \, \exp\left[-\frac{1}{2} \left\{\frac{(\hat{\theta}_{i} - + \theta_{n})^{2}}{\sigma^{2}_{i}} - \frac{(\hat{\theta}_{i} - m)^{2}}{\sigma^{2}_{i} + v} \right\}\right], ~ i \in \{o, r\}.$$ - \item Declare replication success at level $\gamma > 1$ if - $\BF_{01o} \geq \gamma$ and $\BF_{01r} \geq \gamma$, conventionally - $\gamma = 3$ (substantial evidence) or $\gamma = 10$ (strong - evidence). - \item Perform a sensitivity analysis with respect to the prior - distribution. For example, visualize the Bayes factors for different - prior standard deviations to assess the robustness of the - conclusions. - \end{enumerate} - \\ - \bottomrule - \end{tabular} + \item Declare replication success at level $\gamma > 1$ if + $\BF_{01o} \geq \gamma$ and $\BF_{01r} \geq \gamma$, conventionally + $\gamma = 3$ (substantial evidence) or $\gamma = 10$ (strong + evidence). + \item Perform a sensitivity analysis with respect to the prior + distribution. For example, visualize the Bayes factors for different + prior standard deviations to assess the robustness of the + conclusions. + \end{enumerate} + % \\ \bottomrule + \end{tabular} + } \end{table} For both the equivalence testing and the Bayes factor approach, it is critical diff --git a/rsabsence.pdf b/rsabsence.pdf old mode 100644 new mode 100755 index 7a00c3ca91f3a19ed11cce801c92c925de2c2fc9..0d149fea0aef339acf6cfdd5c92f5dd5ed99dfde Binary files a/rsabsence.pdf and b/rsabsence.pdf differ