Ermias Diro (276580), Severine Blesson (296582), Tansy Edwards (122527) et al. · Figshare · 2019
<p>The vertical axis, (<i>Z</i>), is the number of observed treatment successes minus the number of expected successes. The higher the value represented on that vertical axis, the more favorable (better efficacy observed). The horizontal axis <i>(V)</i> is proportional to the number of patients who, over time, have been evaluated at day 29. Hence, over time, each arm’s data points extend to the right. Each arm starts inside the triangle (green area) and recruitment is stopped on crossing either the upper (blue area) or the lower boundary (pink area). Recruitment to the AmBisome monotherapy arm was stopped after the first interim analysis (I1) and to the combination arm after the second interim analysis (I2). Also shown are the final results for each arm, including those who had not reached the primary endpoint (day29) when the interim analysis was performed and the recruitment was stopped. These final results take the trajectory of each arm well away from the triangular boundary, which is known as ‘over-run'. The two numbers after each label are the numbers of patients included in that analysis and the number of these with treatment success at day 29.</p>
Cited by 0Open access
William Fisher, Robin Milhausen, Taylor Kohut et al. · OSF Preprints (OSF Preprints) · 2022
Cited by 0
Rhonda Nicole Balzarini, William Fisher, Lorne Campbell et al. · OSF Preprints (OSF Preprints) · 2022
Cited by 0
James O. Berger · The New Palgrave Dictionary of Economics · 2008
Data often arrives sequentially, rather than as a collection. When this occurs – or when the experiment can be designed so that this occurs – there can be a considerable advantage in using statistical methods, called sequential analysis, that are tailored to such situations. Classical frequentist statistical methods require analysis with a pre-specified collection of data, and hence cannot be used in such sequential settings. Interestingly, Bayesian methods can be directly used in sequential settings.
Cited by 0
G. A. Barnard, Robert Ferber, Abraham Wald · Journal of the American Statistical Association · 1947
Cited by 36
Aug Nishizaka, Kaoru Hayano · 2025
All human activities unfold temporally, and all actions as organisational constituents of a distinct activity are formed and ascribed as such in the sequential context in which they are performed. One general issue is how sequential-contextual features contribute to the formation and ascription of a particular action-type, while another is what, if any, basic sequential organisations look like that generate the orderliness of each temporally unfolding activity. This chapter first explains three main sequential organisations of talk-in-interaction that conversation analysis has explicated, that is, turn-taking, sequence, and repair organisations, and then illustrates how actions are formed and become ascribable in manners sensitive to their sequential positions within these organisations. All actions in talk-in-interaction are implemented within a turn-at-talk, which, in turn, is positioned within a sequence of turns. For example, an answerer to another’s question has to address at least two practical issues: (a) how the answer should be tied to the implementation of the question within the preceding turn, and (b) how it should be sequentially implemented within the current turn. In conclusion, potentialities of such analysis will be discussed, including how other interactional resources than talk can be sequentially analysed.
Cited by 0
B. K. Ghosh · Encyclopedia of Statistical Sciences · 2004
Cited by 0
Encyclopedia of Epidemiology · 2008
Cited by 0
James O. Berger · Palgrave Macmillan UK eBooks · 1990
Statistical experiments are of either fixed sample or sequential design. A fixed sample size experiment is one in which the sample size taken for experimentation is predetermined, while a sequential experiment involves monitoring incoming data to help determine an appropriate time to stop experimentation.
Cited by 0
Tze Leung Lai · Encyclopedia of Biostatistics · 2005
Abstract This article gives an overview of sequential analysis and its applications to data and safety monitoring of clinical trials, determination of median effective dose and maximum tolerated dose, fixed‐width confidence intervals, and adaptive design of biomedical studies.
Cited by 3
The SAGE Encyclopedia of Social Science Research Methods · 2004
Cited by 0
James O. Berger · The New Palgrave Dictionary of Economics · 2008
Cited by 0
Stallard, N. · CentAUR (University of Reading) · 2005
Cited by 0
P. K. Sen · Encyclopedia of Statistical Sciences · 2005
Cited by 0
Kohut, Taylor, Fisher, William, Milhausen, Robin et al. · Open Science Framework · 2016
Cited by 0Open access
Wald, Abraham · 1947
In 1943, while in charge of Columbia University's Statistical Research Group, Abraham Wald devised Sequential Design, an innovative statistical inference system. Because the decision to terminate an experiment is not predetermined, sequential analysis can arrive at a decision much sooner and with substantially fewer observations than equally reliable test procedures based on a predetermined number of observations. The system's immense value was immediately recognized, and its use was restricted to wartime research and procedures. In 1945, it was released to the public and has since revolutio
Cited by 446
Michael D, Michael D. Basil · 1985
Cited by 0
Lorne Campbell, William Fisher, Robin Milhausen et al. · OSF Preprints (OSF Preprints) · 2016
Cited by 0
Van Nostrand's Encyclopedia of Chemistry · 2005
Cited by 0
Roger Bakeman · The Encyclopedia of Clinical Psychology · 2015
Abstract Sequential analysis attempts to detect patterns and temporal associations among observed behaviors and is more a data analytic toolbox than one particular technique. It requires the sort of sequential categorical data that result from behavioral observation when trained observers, often working from video records, assign predefined codes to either events or successive time intervals. Using a code‐unit grid to represent sequential data facilitates sequential analysis; columns are untimed events, intervals, or time units reflecting the precision used to record time. Characterizing observer agreement with kappa (Cohen's or variants) may require application of a sequence‐alignment algorithm. Basic contingency indexes are the odds ratio and the related Yule's Q . When insufficient tallies require pooling individual cases, log‐linear analysis may be needed to characterize group differences. Computing kappa, basic descriptive statistics, and contingency indexes is facilitated by the code‐unit grid, whose columns serve as computational units.
Cited by 0