From 2c063458615c41a340623b0112f1ddbc1b63d885 Mon Sep 17 00:00:00 2001 From: Joe Thorley Date: Mon, 9 Dec 2024 21:33:16 -0800 Subject: [PATCH] remove argument name without context. --- paper/paper.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/paper/paper.md b/paper/paper.md index 45b3bc45..04c49410 100644 --- a/paper/paper.md +++ b/paper/paper.md @@ -102,7 +102,7 @@ To estimate the values using the weighted arithmetic mean set `multi_est = FALSE `ssdtools` generates confidence intervals for $\text{HC}_x$ and $\text{HP}_u$ values via bootstrapping. By default all versions of `ssdtools` use parametric bootstrapping for non-censored data as it has better coverage than the equivalent non parametric approach used in other SSD modelling software such as `Burrlioz` [see @fox_methodologies_2022]. -The first two versions of `ssdtools` both calculated the model averaged CI from the weighted arithmetic mean of the CIs of the individual distributions (`weighted_arithmetic`). +The first two versions of `ssdtools` both calculated the model averaged CI from the weighted arithmetic mean of the CIs of the individual distributions. Unfortunately, this approach has recently been shown to have poor coverage [@fox_methodologies_2024] and is inconsistent with the *inversion principle*. Consequently, v2 also offers a parametric bootstrap method for non-censored data that uses the joint cdf to generate data before refitting the original distribution set and solving for the newly estimated joint cdf [see details in @fox_methodologies_2024].