Talk:Significance level

A recent edit that I made has been challenged. Following WP:BRD, this is for a discussion of that.

Your claim of a "different concept" is false. Indeed, the term "significance level" is defined in the first paragraph of statistical significance. On the other hand, the Type I and type II errors article is not specifically about this topic, and does not have nearly as good a treatment. BetterMath (talk)

User:Danielkueh, your point about self-linking is again invalid. The article statistical significance is specifically on the topic. BetterMath (talk) 19:59, 7 May 2019 (UTC)[reply]

It is not false. They are DIFFERENT concepts. Do you have any experience using inferential statistics? Do you know the definitions of these terms??!? Do you know that type I error is sometimes used synonymously with significance level? Do you know the issue of “self-linking?” Other than making statements like “invalid,” do you actually have secondary sources to back up your claim? Are you aware of WP policies such as WP:V and WP:consensus? danielkueh (talk) 20:07, 7 May 2019 (UTC)[reply]

Thank you for discussing here. I am irrelevant for this discussion, but I have worked professionally as a statistical analyst in finance (which generally has better quality than academia). The term "significance level" refers specifically to statistical significance, and that is stated, and described in the latter article. Where is the self-linking? BetterMath (talk) 20:13, 7 May 2019 (UTC)[reply]

”Better quality than academia.” Please... what a baseless assertion. So do you even know the difference between statistical significance and significance level? Because if you do, you would know that significance level does not refer to statistical significance even through the latter does require a significance level to be set. Hence, you’re confusing the two concepts, which is analogous to confusing “wining a race” and a “finish line.” They are related but different. Your competence on this subject is therefore especially relevant, considering that you don’t know the nuance or difference between the two concepts and you don’t even know the WP policies but have the gall to tell others to follow them, which makes interacting with you on this WP project very irritating. The statistical significance page has a wiki-link to the type I and II error page, which actually explains significance level. But thanks to you, readers will no longer have access to the definition as they will be taken back to the statistical significance page itself. DO YOU NOT UNDERSTAND THAT?!?!?!? Sigh. danielkueh (talk) 20:25, 7 May 2019 (UTC)[reply]

(In finance, there is an objective measure of performance.) The article statistical significance currently has two wikilinks to Type I and type II errors#Type I error and one wikilink to Type I and type II errors; I have not edited the latter article.

The latter article states this: "The type I error rate or significance level is the probability of rejecting the null hypothesis given that it is true". The article statistical significance states this: "a study's defined significance level, denoted

α

, is the probability of the study rejecting the null hypothesis, given that it were true". Thus, they are the same.

The article statistical significance then gives the history of

α

(i.e. significance level), and an extended discussion of interpretation and related concepts. Thus, the article statistical significance contains substantial information about alpha levels that is not in Type I and type II errors#Type I error. What information about alpha levels does Type I and type II errors#Type I error give that is not in statistical significance? BetterMath (talk) 20:40, 7 May 2019 (UTC)[reply]

A quote from Online Statistics Education, about Type I and Type II Errors, is below.

It might seem that α is the probability of a Type I error. However, this is not correct. Instead, α is the probability of a Type I error given that the null hypothesis is true. If the null hypothesis is false, then it is impossible to make a Type I error.

BetterMath (talk) 20:59, 7 May 2019 (UTC)[reply]

UTC)

So finance has an ”objective measure of performance” and academia doesn’t?!?! *eye roll* That’s the dumbest thing I have ever read.

I feel like a broken record but here goes. Type I error rate, which you now know is what the alpha or significance level is based on, is explored in much greater detail in the type I and type II article. Readers not only learn about type I errors but also about type II errors and how they are related to inferential statistics. And finally, this is a simple case of logic and common sense. Since type I error rate is just another name for the alpha or significance level, it naturally follows that this page should redirect to an alternative and very closely-related concept. In fact, just go read WP:redirect. danielkueh (talk) 21:17, 7 May 2019 (UTC)[reply]

We agree that we are going around in circles. Especially since you ignored the points that I raised. Perhaps we should attempt WP:dispute resolution. BetterMath (talk) 21:24, 7 May 2019 (UTC)[reply]

What the hell are you talking about?!?!?! I have addressed all your damn points!! You are either being ~~a recalcitrant~~intransigent or a complete idiot or maybe both. Whatever. Do what you want. I have already wasted enough time and energy on this fruitless discussion. Sheesh. danielkueh (talk) 21:47, 7 May 2019 (UTC)[reply]

Please select which form of WP:dispute resolution you would like us to follow. My suggestion is to begin with Third Opinion. BetterMath (talk) 21:55, 7 May 2019 (UTC)[reply]