Abstract:Highlights
Confidence Sets (CSs) extend the idea of confidence intervals to fMRI maps.
For a Cohen’s
threshold
upper CS asserts where
lower CS where
.
We demonstrate the CSs method on HCP subject-level Cohen’s
d
data.
We compare the CSs with results from standard statistical voxelwise inference.
Unlike tra… Show more
“…Our proposed methods still rely on hypothesis testing using SEI so are not completely free of the limitations of PVT. For imaging, modern approaches that construct confidence sets using effect size thresholding approaches or Bayesian inference procedures hold promise as true alternatives to PVT-based inference for neuroimaging (Bowring et al, 2019(Bowring et al, , 2021Chen et al, 2019;Sommerfeld, Sain, & Schwartzman, 2018). Our suggestions here of EST demonstrates the advantage of considering alternatives to classical PVT.…”
Section: Discussionmentioning
confidence: 89%
“…In this article, we argue for using an effect size-based CFT, instead of a threshold based on a p-value. Our suggestion is motivated by increasing criticism of PVT in the context of hypothesis testing and an increased interest in potential alternatives (Bowring et al, 2019;Bowring, Telschow, Schwartzman, & Nichols, 2021;Chen et al, 2019;Chen, Taylor, & Cox, 2017;Wasserstein & Lazar, 2016;Wasserstein, Schirm, & Lazar, 2019). This approach resolves the limitations of PVT stated above: (a) Using effect size thresholding (EST), the set of voxels that are identified as activated in an analysis does not depend on the sample size, so it improves the consistency of the target regions of activation across studies.…”
The classical approach for testing statistical images using spatial extent inference (SEI) thresholds the statistical image based on the p-value. This approach has an unfortunate consequence on the replicability of neuroimaging findings because the targeted brain regions are affected by the sample size-larger studies have more power to detect smaller effects. Here, we use simulations based on the preprocessed Autism Brain Imaging Data Exchange (ABIDE) to show that thresholding statistical images by effect sizes has more consistent estimates of activated regions across studies than thresholding by p-values. Using a constant effect size threshold means that the p-value threshold naturally scales with the sample size to ensure that the target set is similar across repetitions of the study that use different sample sizes. As a consequence of thresholding by the effect size, the type 1 and type 2 error rates go to zero as the sample size gets larger. We use a newly proposed robust effect size index that is defined for an arbitrary statistical image so that effect size thresholding can be used regardless of the test statistic or model.
“…Our proposed methods still rely on hypothesis testing using SEI so are not completely free of the limitations of PVT. For imaging, modern approaches that construct confidence sets using effect size thresholding approaches or Bayesian inference procedures hold promise as true alternatives to PVT-based inference for neuroimaging (Bowring et al, 2019(Bowring et al, , 2021Chen et al, 2019;Sommerfeld, Sain, & Schwartzman, 2018). Our suggestions here of EST demonstrates the advantage of considering alternatives to classical PVT.…”
Section: Discussionmentioning
confidence: 89%
“…In this article, we argue for using an effect size-based CFT, instead of a threshold based on a p-value. Our suggestion is motivated by increasing criticism of PVT in the context of hypothesis testing and an increased interest in potential alternatives (Bowring et al, 2019;Bowring, Telschow, Schwartzman, & Nichols, 2021;Chen et al, 2019;Chen, Taylor, & Cox, 2017;Wasserstein & Lazar, 2016;Wasserstein, Schirm, & Lazar, 2019). This approach resolves the limitations of PVT stated above: (a) Using effect size thresholding (EST), the set of voxels that are identified as activated in an analysis does not depend on the sample size, so it improves the consistency of the target regions of activation across studies.…”
The classical approach for testing statistical images using spatial extent inference (SEI) thresholds the statistical image based on the p-value. This approach has an unfortunate consequence on the replicability of neuroimaging findings because the targeted brain regions are affected by the sample size-larger studies have more power to detect smaller effects. Here, we use simulations based on the preprocessed Autism Brain Imaging Data Exchange (ABIDE) to show that thresholding statistical images by effect sizes has more consistent estimates of activated regions across studies than thresholding by p-values. Using a constant effect size threshold means that the p-value threshold naturally scales with the sample size to ensure that the target set is similar across repetitions of the study that use different sample sizes. As a consequence of thresholding by the effect size, the type 1 and type 2 error rates go to zero as the sample size gets larger. We use a newly proposed robust effect size index that is defined for an arbitrary statistical image so that effect size thresholding can be used regardless of the test statistic or model.
“…See also Bowring et al (2021). Please note that for conventional one-sample t-tests √ ′( ′ ) − = √ 1 and DoF=n-1, and for conventional two-sample t-tests √ ′( ′…”
Section: Standardized Effect Sizes For T-tests In Fmrimentioning
confidence: 99%
“…In contrast, with a very high number of participants an analysis will in most cases deliver a large number of significant voxels. In most cases it is also conceptually more interesting to know whether the activation of a brain region is more or less strongly associated with a specific behavior or intervention (see for example Bowring et al, 2019;Bowring et al, 2021).…”
Section: Introductionmentioning
confidence: 99%
“…Such approaches have so far rarely been applied to MRI data (see Bowring et al, 2021;Pardoe et al, 2016;Reggev et al, 2020 for examples). With this paper we want to contribute to the use of this important statistical method in fMRI research.…”
Null hypothesis significance testing is the major statistical procedure in the field of fMRI, but provides only a rather limited picture of the effects in a data set. When sample size and power is low relying only on strict significance testing may lead to a host of false negative findings. In contrast, with very large data sets virtually every voxel might become significant.It is thus desirable to complement significance testing with procedures like inferiority and equivalence tests that allow to formally compare effect sizes within and between data sets and offer novel approaches to obtain insight into fMRI data. The major component of these tests are estimates of standardized effect sizes and their confidence intervals. Here we show how Hedge’s g, the bias corrected version of Cohen’s d, and its confidence interval can be obtained from SPM t maps. We then demonstrate how these values can be used to evaluate whether non-significant effects are really statistically smaller than significant effects to obtain “regions of undecidability” within a data set, and to test for the replicability and lateralization of effects.This method allows the analysis of fMRI data beyond point estimates enabling researchers to take measurement uncertainty into account when interpreting their findings.
Null hypothesis significance testing is the major statistical procedure in fMRI, but provides only a rather limited picture of the effects in a data set. When sample size and power is low relying only on strict significance testing may lead to a host of false negative findings. In contrast, with very large data sets virtually every voxel might become significant. It is thus desirable to complement significance testing with procedures like
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.