False Discovery Rate and k-FWER are error measures for situations where multiple null hypotheses are tested simultaneously. The False Discovery Rate limits the expected proportion of falsely rejected hypotheses among all rejections. In contrast, the k-FWER limits the probability of making at least k false rejections. Both allow for different error objectives than the family-wise error rate, which controls for even a single false positive within the entire family of hypotheses.<\/p>\n
The False Discovery Rate relates to the composition of positive findings. If ten hypotheses are rejected, it describes the expected proportion of those that are false positives; if no rejections are made, the associated proportion is treated as zero by definition. It is therefore particularly focused on a list of detected signals and not on the overall probability of an error in a single study.<\/p>\n
With k-FWER, the number of false rejections that are still tolerated before an error event occurs is determined in advance. For k equal to one, it corresponds to the conventional FWER. For a larger k, the procedure can detect more true effects because not every single false rejection needs to be excluded. The value of k is therefore a professional determination for handling multiple potential false positives and not an optimization to be derived from the data.<\/p>\n
In exploratory biomarker analyses or extensive safety evaluations, large families of questions often arise, the results of which are intended for further investigation as signals. An FDR or k-FWER control can structure the relationship between signal detection and false positives in such cases, without treating every single finding as definitive proof of efficacy. Which hypotheses belong to the family, what dependencies exist, and which procedure is used must be defined before the analysis.<\/p>\n
These procedures do not permit selectively inferring clinical efficacy from a long list of statistical findings. Biological plausibility, measurement quality, pre-definition, and external confirmation remain relevant for a biomarker. For safety data, it is also important to distinguish whether the analysis aims to detect an unexpected pattern or whether a specific risk hypothesis is being tested with its own confirmatory statement.<\/p>\n
Dependencies between tests are essential for selecting a specific procedure. Some FDR methods assume certain dependency structures, while the k-FWER approaches described by Lehmann and Romano were constructed without assumptions about the dependency of individual p-values. The sample data itself does not prove that the structure required for a chosen correction method is present.<\/p>\n
False Discovery Rate and k-FWER are less stringent than the classical family-wise error rate, provided that more than one error is tolerated. Therefore, they are not suitable for the confirmatory primary endpoint without special justification, as a positive statement for such an endpoint typically should not allow for any avoidable false positives. The EMA guideline on multiplicity emphasizes that inadequate control of multiple testing can lead to unsubstantiated efficacy claims.<\/p>\n
Furthermore, the two measures are not interchangeable. A controlled FDR does not directly state how likely at least two false findings are; conversely, a k-FWER does not describe the expected false proportion within the reported hits. The chosen measure must therefore match the objective: prioritizing a list of hits requires a different error description than limiting a specific number of incorrect decisions.<\/p>\n
In a study protocol, hypothesis families for genomic markers, numerous laboratory parameters, or predefined safety signals should be listed separately. The SAP must specify whether FDR or k-FWER is controlled, how p-values are sorted or processed step-wise, and whether the results remain exploratory. Evaluations must not be made more favorable retrospectively by omitting undesirable hypotheses from the family.<\/p>\n
Tables for results presentation should name the total number of hypotheses tested, the rejected hypotheses, and the error measure applied. This prevents a selection of positive markers from appearing as a complete evaluation, even though it only reflects the predefined discovery process.<\/p>\n
Full-service CROs like Mediconomics support FDR and k-FWER analyses by documenting hypothesis families, selecting and programming the pre-agreed multiple testing procedure, reviewing the resulting lists, and labeling exploratory results in the table, listing, and figure plan. This ensures that it remains clear whether a positive finding is a signal for further investigation or part of a confirmatory proof.<\/p>\n
Is the FDR the probability that a single positive finding is false?<\/strong><\/p>\n No. It is an expected value over repeated applications of the procedure and refers to the proportion of false rejections within the entire set of rejections. It is not a posterior error probability for a specific biomarker.<\/p>\n Why is there a \u201ck\u201d before k-FWER?<\/strong><\/p>\n The k denotes the number of false rejections from which an error event is counted. Controlling the 2-FWER thus limits the probability of at least two false rejections and is less restrictive than controlling for even a single false positive.<\/p>\n Can FDR and k-FWER occur in the same study?<\/strong><\/p>\n Yes, provided they are used for separate, pre-described hypothesis families or analysis objectives. The SAP must transparently indicate which measure belongs to which evaluation; a single result must not be assessed with varying error measures depending on the desired statement.<\/p>\n False Discovery Rate and k-FWER are error measures for situations where multiple null hypotheses are tested simultaneously. The False Discovery Rate limits the expected proportion of falsely rejected hypotheses among all rejections. In contrast, the k-FWER limits the probability of making at least k false rejections. Both allow for different error objectives than the family-wise […]<\/p>\n","protected":false},"author":10,"featured_media":0,"parent":0,"template":"","meta":{"_acf_changed":false,"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"default","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"glossary-cat":[22],"class_list":["post-8075","glossary","type-glossary","status-publish","hentry","glossary-cat-biostatistik-methodik"],"acf":[],"related_terms":"","external_url":"","internal_reference_id":"","_links":{"self":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/8075","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary"}],"about":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/types\/glossary"}],"author":[{"embeddable":true,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/users\/10"}],"version-history":[{"count":0,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary\/8075\/revisions"}],"wp:attachment":[{"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/media?parent=8075"}],"wp:term":[{"taxonomy":"glossary-cat","embeddable":true,"href":"https:\/\/mediconomics.com\/en\/wp-json\/wp\/v2\/glossary-cat?post=8075"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}Regulatory References<\/h2>\n
\n