News from Computational Statistics
http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-0-0-0
The latest news from Computational StatisticsSun, 20 May 2018 18:28:58 GMT2018-05-20T18:28:58ZSpringer Computational Statisticshttp://images.springer.com/cda/content/designimage/cda_displaydesignimage.gif?SGWID=0-0-17-901483-0
http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-0-0-0
The odds are against ESP
http://www.springer.com/about+springer/media/springer+select?SGWID=0-11001-6-1153922-0
New statistical approach doesn’t support claims that extra-sensory perception exists<br /><img align="right" src="https://images.springer.com/sgw/journals/medium/13423.jpg" /><div>Can people truly feel the future? Researchers remain skeptical, according to a new study by Jeffrey Rouder and Richard Morey from the University of Missouri in the US, and the University of Groningen in the Netherlands, respectively. Their work (1) appears online in the <em>Psychonomic Bulletin & Review</em> (2), published by Springer.<br /><br />Although extra-sensory perception (ESP) seems impossible given our current scientific knowledge, and certainly runs counter to our everyday experience, a leading psychologist, Daryl Bem of Cornell University, is claiming evidence for ESP. Rouder and Morey look at the strength of the evidence in Dr. Bem's experiments.<br /><br />Their application of a relatively new statistical method that quantifies how beliefs should change in light of data, suggests that there is only modest evidence behind Dr. Bem's findings (that people can feel, or sense, salient events in the future that could not otherwise be anticipated, and cannot be explained by chance alone), certainly not enough to sway the beliefs of a skeptic.<br /><br />They highlight the limitations of conventional statistical significance testing (p values), and apply a new technique (meta-analytical Bayes factor) to Dr. Bem's data, which overcomes some of these limitations. According to Rouder and Morey, in order to accurately assess the total evidence in Bem's data, it is necessary to combine the evidence across several of his experiments, not look at each one in isolation, which is what researchers have done up till now. They find there is some evidence for ESP – people should update their beliefs by a factor of 40. <br /><br />In other words, beliefs are odds. For example, a skeptic might hold odds that ESP is a long shot at a million-to-one, while a believer might believe it is as possible as not (one-to-one odds). Whatever one's beliefs, Rouder and Morey show that Bem's experiments indicate they should change by a factor of 40 in favor of ESP. The believer should now be 40-to-1 sure of ESP, while the skeptic should be 25000-to-1 sure against it.<br /><br />Rouder and Morey conclude that the skeptics odds are appropriate: “We remain unconvinced of the viability of ESP. There is no plausible mechanism for it, and it seems contradicted by well-substantiated theories in both physics and biology. Against this background, a change in odds of 40 is negligible." <br /><br /><strong>Reference</strong><br />1. Rouder JN & Morey RD (2011). A Bayes factor meta-analysis of Bem's ESP claim. <em>Psychonomic Bulletin & Review</em>. DOI 10.3758/s13423-011-0088-7<br />2. <em>Psychonomic Bulletin & Review</em> is the official journal of the Psychonomic Society.<br /><br /><strong>The full-text article is available to journalists on request.</strong><br /></div><br /><h2>More information about Psychonomic Bulletin & Review:</h2><a href="http://www.springer.com/psychology/cognitive+psychology/journal/13423">Psychonomic Bulletin & Review</a><br /><br /><h2>Abstract of the Study:</h2><a href="http://link.springer.com/10.3758%2Fs13423-011-0088-7">A Bayes factor meta-analysis of Bem’s ESP claim</a><br /><br /><h2>Contact:</h2><a href="http://www.springer.com/about+springer/media/contact?SGWID=0-40582-19-69370-0">Renate Bayaz</a><br /><br />New York / HeidelbergTue, 17 May 2011 22:00:00 GMThttp://www.springer.com/about+springer/media/springer+select?SGWID=0-11001-6-1153922-02011-05-17T22:00:00ZThe R Software
http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-6-1465245-0
<div><script type="text/javascript" src="http://www.edge-cdn.net/videojs_652952?playerskin=37016"></script><br /></div><br />http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-6-1465245-0Call For Papers: Special Feature—Perspectives on Data Science for Advanced Statistics
http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-6-1574257-0
<h3>Editor: Makoto Aoshima (Institute of Mathematics, University of Tsukuba, Japan)</h3><div>Data Science is certainly a hot term right now. Some PhD students mentioned that to get a job, you should list "Data Science" on your PhD whether you know anything about it or not! It seems to be becoming increasingly known that there are a lot of unqualified "data scientists" out there.<br />What is "Data Science"? I took the following snippet from the Wikipedia page on "Data Science": ‘to extract knowledge or insights from data’. For me, it is really interesting that this is so similar to the definition of statistics that I offer, and have been offering for the past 25 years, when I teach elementary statistic courses.<br />That being said, is there any value added from this new terminology? I feel that there is, and the reason is that by renaming statistics it somehow seems to free up folks to start thinking about statistical matters, especially those who would probably not do so without that naming. In particular, lots of very talented people from fields such as math, CS and various areas of engineering want to call themselves "Data Scientists", and they are in the process of bringing lots of exciting new ideas, approaches and ways of thinking to statistics.<br />The aim of this special feature is to gain perspectives on data science for advanced statistics. We welcome original research articles, reviews, theoretical articles and methodological articles. Experimental and theoretical contributions are also welcomed.<br /></div><br /><h3>Possible topics could include, but are not limited to, the following</h3><div>- Interplay between machine learning, signal processing, and statistics for Big Data<br />- Integration of complex and diverse data types<br />- Model selection for high-dimensional data<br />- Natural language processing<br />- Data heterogeneity<br />- Stochastic processes and statistical learning theory toward modeling of dependencies<br />- Estimation pitfalls when the noise is not i.i.d.<br />- History and statistical inferences of intrinsic stationary random fields<br />- Required skills for data scientists and curriculum of data science<br /><br /><br /></div><br /><h3>When you submit</h3><div>Please select <br />Special Feature: Perspectives on Data Science for Advanced Statistics <br />at the “Additional Information” stage.</div><br /><h3>Submission deadline:</h3><div>December 31, 2017<br /></div><br />http://www.springer.com/statistics/computational+statistics?SGWID=0-10130-6-1574257-0