Talk:Statistical significance

Wiki Education Foundation-supported course assignment

This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Cokusiak. Peer reviewers: Cokusiak.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 10:10, 17 January 2022 (UTC)

Comment on Dubious

The citation (USEPA December 1992) contains numerous statistical tests, some presented as p-values and some as confidence intervals. Figures 5-1 through 5-4 show some of the test statistics used in the citation. In two of the figures the statistics are for individual studies and can be assumed to be prospective. In the other two, the statistics are for pooled studies and can be assumed to be retrospective. Table 5-9 includes the results of the test of the hypothesis RR = 1 versus RR > 1 for individual studies and for pooled studies. In the cited report, no distinction between prospective tests and retrospective tests was made. This is a departure from the traditional scientific method, which makes a strict distinction between predictions of the future and explanations of the past. Gjsis (talk)

Statistical significance in Psychology

In the late 80s-early 90's, and perhaps afterwards, Clark University taught that in Psychology, as opposed to just about all other fields of science, P-values under 0.1 were accepted, that was the norm, rather than the standard .05. This was because results often weren't as robust for a variety of reasons, namely that human subjects were needed for the experiments, it was more expensive to get participation (compared to experiments on animals, for example), and humans have rights that require a lot of work to document and follow. I don't think it's a coincidence that it is psychology journals that bring up the issues of statistical significance. It would be great to see additional information on statistical significance in psychology in particular, because that's where its most controversial. Tumacama (talk) 00:53, 9 May 2020 (UTC)

I reverted your recent edit because the problem And concern of overuse is not unique to psychology unless the sources explicitly say otherwise, which I don’t believe they do. danielkueh (talk) 01:29, 9 May 2020 (UTC)

Actually the onus is on you to provide examples. You can't just revert my claim because you "think" overuse is a problem elsewhere. It has been a systematic issue in psychology for at least the 30 years I was taught it at university. Issue because it's not a few bad eggs, it's a common and accepted practice. If you think it is a systematic problem in other fields, just give a citation. Sometimes I wish Wikipedia didn't work that way, but once I made the claim, you HAVE to provide evidence of the contrary to prove your point. It's a reasonable claim (although I don't think any fields of study welcome this difference, and even "parapsychology" and other fields don't even recognize P-values Tumacama (talk) 23:12, 4 August 2020 (UTC)

In Wikipedia, we have a process called BRD (see WP:BRD) whereby if you were to make a bold edit, and someone else reverts it, the next step would be to discuss it on the article's talk page and seek consensus (see WP:consensus). Also, we stick with the sources (see WP:V and WP:RS). Other than your own anecdotal experience at your university, you have not provided any reliable sources to support the claim that this issue is a perennial problem in psychology and not in other fields. In fact your assertion is not supported by the sources that have been cited in this article. None of the sources listed claimed that this is a problem unique to psychology. So unless the sources say otherwise, there is no reason to change that heading. And as far as burden of proofs are concerned, you have it backwards. I recommend learning the fallacy of "appealing to ignorance" to better understand why. Since you made the claim, the burden of proof is on you to substantiate it by providing reliable sources that anyone can independently verify. This is consistent with WP's policy (see WP:BURDEN in WP:V). danielkueh (talk) 23:39,

BUT ITS NOT A BOLD EDIT IT IS A SIMPLE ADDITION THAT DOESN"T CONTRADICT ANYTHING!!!!

4 August 2020 (UTC) Failing Grade: 89% of Introduction-to-Psychology Textbooks That Define or Explain Statistical Significance Do So Incorrectly

Show all authors

Scott A. Cassidy, Ralitza Dimova, Benjamin Giguère, ...

First Published June 27, 2019 Research Article

https://doi.org/10.1177/2515245919858072

Please indent your comments appropriately so that it is easy for other editors to distinguish them from mine. In WP, a bold edit just means "Go for it." The source that you just gave does not say that there is an overuse of statistical significance in psychology. Rather, it is saying that it is "commonly used" and that it is often defined or explained incorrectly, particularly in introductory psychology textbooks. Also, please do not re-revert the page without achieving consensus. By doing so, you're disregarding the views of other editors and WP's policies. Please be patient and continue the discussion process. Thanks. danielkueh (talk) 00:00, 5 August 2020 (UTC)

And since you wanted evidence, I encourage you to read this recent 2019 article [1] from Nature magazine, which is a very well-established scientific journal. If you were to scroll down, you will see a passage that states that the authors invited experts to comment on a draft to abandon the "entire concept of statistical significance," which included "statisticians, clinical and medical researchers, biologists and psychologists from more than 50 countries and across all continents except Antarctica." Again, not an issue that is unique to psychology. danielkueh (talk) 00:10, 5 August 2020 (UTC)

"Official statement" from the ASA actually not official

> In 2019, over 800 statisticians and scientists signed a message calling for the abandonment of the term "statistical significance" in science, and the ASA published a further official statement declaring (page 2): [...]

If you click the link, the very first thing the editorial says is this:

> EDITORIAL: The editorial was written by the three editors acting as individuals and reflects their scientific views not an endorsed position of the American Statistical Association. 60.115.154.227 (talk) 14:12, 23 January 2023 (UTC)