Comments - The Death of the Statistical Tests of Hypotheses - Data Science Central2019-03-24T21:13:33Zhttps://www.datasciencecentral.com/profiles/comment/feed?attachedTo=6448529%3ABlogPost%3A398821&xn_auth=noHi, it's late, sorry
For some…tag:www.datasciencecentral.com,2017-08-16:6448529:Comment:6076242017-08-16T15:31:41.442ZChristian Cantoshttps://www.datasciencecentral.com/profile/ChristianCantos
<p>Hi, it's late, sorry</p>
<p>For some reason I hit the problem of p-value aka significance in Statistical testing. I wrote this on my blog. </p>
<p>Yet another contribution to the P-value discussion : Probabilities are maths, not logic.</p>
<p>Please would you kindly have a look at</p>
<p><a href="https://chcantos.blogspot.fr/" target="_blank">https://chcantos.blogspot.fr/</a></p>
<p>Thanks</p>
<p>Hi, it's late, sorry</p>
<p>For some reason I hit the problem of p-value aka significance in Statistical testing. I wrote this on my blog. </p>
<p>Yet another contribution to the P-value discussion : Probabilities are maths, not logic.</p>
<p>Please would you kindly have a look at</p>
<p><a href="https://chcantos.blogspot.fr/" target="_blank">https://chcantos.blogspot.fr/</a></p>
<p>Thanks</p> Another simple approach that…tag:www.datasciencecentral.com,2016-07-07:6448529:Comment:4459612016-07-07T20:43:16.875ZChristie Haskellhttps://www.datasciencecentral.com/profile/ChristieHaskell
<p>Another simple approach that I'm using to communicate the results of A/B tests to Designers with no statistics background is to calculate the Cohen's d between variants and then convert this to the probability that the winning variant is the superior variant. If the data isn't normally distributed I'm using a Box-Cox transformation first. Probability is easy for people to understand and it has thus far simplified communicating the results of A/B tests.</p>
<p>Another simple approach that I'm using to communicate the results of A/B tests to Designers with no statistics background is to calculate the Cohen's d between variants and then convert this to the probability that the winning variant is the superior variant. If the data isn't normally distributed I'm using a Box-Cox transformation first. Probability is easy for people to understand and it has thus far simplified communicating the results of A/B tests.</p> Thanks...again. The key see…tag:www.datasciencecentral.com,2016-05-25:6448529:Comment:4290222016-05-25T03:10:17.165ZMichael Claytonhttps://www.datasciencecentral.com/profile/MichaelClayton
<p>Thanks...again. The key seems to be visualization first then structured statements that meet legal requirements in court appearances, and academic communications standards. For example. I was once told by a legal eagle that any statistical statement should start with <strong>"It seems to me"</strong> and then <strong>"that there is sufficient statistical evidence to assume"</strong> and then <strong>"that A is > B" with 99% certainty."</strong> I am sure its worse now days, but…</p>
<p>Thanks...again. The key seems to be visualization first then structured statements that meet legal requirements in court appearances, and academic communications standards. For example. I was once told by a legal eagle that any statistical statement should start with <strong>"It seems to me"</strong> and then <strong>"that there is sufficient statistical evidence to assume"</strong> and then <strong>"that A is > B" with 99% certainty."</strong> I am sure its worse now days, but GRAPHICS makes all the difference. </p> Hello Vincent. Your article i…tag:www.datasciencecentral.com,2016-03-17:6448529:Comment:4021802016-03-17T19:45:31.965ZRobert Lemayhttps://www.datasciencecentral.com/profile/RobertLEmay
<p>Hello Vincent. Your article is 100% top level and you are 100% right. The problem is not only with traditional statisticians but with traditional mathematicians. The difficult point is that a mathematician is only interested talking to another mathematician. Explaining a simple "thumb" rule to a non specialist is loosing time for him. But on the other side, we utterly need them...we need them a week per year. The 51 weeks left, we need a mathematician that builds "thumbs" rules from…</p>
<p>Hello Vincent. Your article is 100% top level and you are 100% right. The problem is not only with traditional statisticians but with traditional mathematicians. The difficult point is that a mathematician is only interested talking to another mathematician. Explaining a simple "thumb" rule to a non specialist is loosing time for him. But on the other side, we utterly need them...we need them a week per year. The 51 weeks left, we need a mathematician that builds "thumbs" rules from complicated theories. But on my 28 year on the job, I never met such a profile. The next Field's price should be given to such work. Another point of this problem is that a Fortune 500 company would never pay a mathematician for "thumb" rules production... </p>
<p>Myself, I am not a mathematician nor statistician but a simple engineer using mathematics as a tool. So not finding any such profile I build myself a very simple toolbox (but very useful) using "truth tables", Venn diagram (up to 8 fields) and basic linear algebra. My mathematicians's friends (beleive me, I have some!) call it the "the calculation cooking book". </p>
<p>As a consultant I work in Service Desk, user/Customer relations and IT ressource management and I am using numbers, relations, categorisation, proof... every day. Where I loose ground, is where I have 3 or 4 simple relations about one event and to transforme them into one unique relation. If you have ideas about this, ready to talk about. Regards.Robert</p> Dalila, this is not point est…tag:www.datasciencecentral.com,2016-03-11:6448529:Comment:4001612016-03-11T20:10:45.433ZVincent Granvillehttps://www.datasciencecentral.com/profile/VincentGranville
<p>Dalila, this is not point estimation, unless you consider computing the lower and upper bounds of a confidence intervals to be point estimation - in which case every number is a point estimation - including the power of a test, or the type I error.</p>
<p>Dalila, this is not point estimation, unless you consider computing the lower and upper bounds of a confidence intervals to be point estimation - in which case every number is a point estimation - including the power of a test, or the type I error.</p> By the way, when analyzing a…tag:www.datasciencecentral.com,2016-03-11:6448529:Comment:4003052016-03-11T14:36:02.505ZDalila Benachenhouhttps://www.datasciencecentral.com/profile/DalilaBenachenhou
<p>By the way, when analyzing a confusion matrix, you may need to talk about Type I and Type II error, which are call False Positive of False Negative, by data scientists. By the way, False Negative is type II error, and False Positive is type I error.</p>
<p>By the way, when analyzing a confusion matrix, you may need to talk about Type I and Type II error, which are call False Positive of False Negative, by data scientists. By the way, False Negative is type II error, and False Positive is type I error.</p> There are 2 categories for pa…tag:www.datasciencecentral.com,2016-03-11:6448529:Comment:4003022016-03-11T14:14:56.271ZDalila Benachenhouhttps://www.datasciencecentral.com/profile/DalilaBenachenhou
<p>There are 2 categories for parameters inference: estimation, and statistical test. What you presented in "<strong>Statistical tests of hypotheses revisited" </strong>is called point estimation.</p>
<p>There are 2 categories for parameters inference: estimation, and statistical test. What you presented in "<strong>Statistical tests of hypotheses revisited" </strong>is called point estimation.</p> Very insightful, enough of th…tag:www.datasciencecentral.com,2016-03-10:6448529:Comment:3999252016-03-10T17:58:30.173Zakindaini bolarinwahttps://www.datasciencecentral.com/profile/akindainibolarinwa
<p>Very insightful, enough of this pessimistic approach to hypothesis testing called 'P-value'. This term is only understood by academicians and hard to explain to professionals who actually use the result of data analysis.</p>
<p>Very insightful, enough of this pessimistic approach to hypothesis testing called 'P-value'. This term is only understood by academicians and hard to explain to professionals who actually use the result of data analysis.</p> This was indeed a much-requir…tag:www.datasciencecentral.com,2016-03-09:6448529:Comment:3992182016-03-09T06:44:45.212ZHaardik Sharmahttps://www.datasciencecentral.com/profile/HaardikSharma
<p>This was indeed a much-required change from the ASA. These changes should have been done almost a decade ago, still it is better late than never. <span style="color: #444444; font-family: Tahoma; font-size: 14px; line-height: 18px;">p-values are often misleading </span><span style="color: #444444; font-family: Tahoma; font-size: 14px; line-height: 18px;">at times, especially when we are working with huge datasets. With the onset of Big Data Analytics, there is actually no need to worrying…</span></p>
<p>This was indeed a much-required change from the ASA. These changes should have been done almost a decade ago, still it is better late than never. <span style="color: #444444; font-family: Tahoma; font-size: 14px; line-height: 18px;">p-values are often misleading </span><span style="color: #444444; font-family: Tahoma; font-size: 14px; line-height: 18px;">at times, especially when we are working with huge datasets. With the onset of Big Data Analytics, there is actually no need to worrying about analyzing a sample when we have computation powers to analyze entire population!. Great job by Dr. Vincent in proposing this alternative set of strategy/framework to fill the tech gap.</span></p>
<p><span style="color: #444444; font-family: Tahoma; font-size: 14px; line-height: 18px;">p-values are </span><a href="http://heather.cs.ucdavis.edu/~matloff/132/PLN/ProbStatBook.pdf" target="_blank" style="outline: none; color: #205b87; font-family: Tahoma; font-size: 14px; line-height: 18px;">at best underinformative and often misleading</a></p>