Talk:Hosmer–Lemeshow test

I quote 'It is used frequently in risk prediction models, especially in cancer epidemiology.' Where do you get such comments? I know it's widely used in goodness fit of logistic model in epidemiology, but to say 'especially in cancer epidemiology', it's too liberal! Yg17 (talk) 02:03, 10 January 2011 (UTC)

I agree, it was ambigious. I only referenced a specific example where I've encountered this test in application. It may be better suited for an example section, but I think higher priority would be a "motivation" section where we display a plot of the distribution of aggregate event rates and risks based on deciles. Saffloped (talk) 19:58, 9 September 2011 (UTC)

Recent long (and possibly tedious) additions to this article

Recently, Michaelg2015 introduced some very long and, I suggest, tedious material to the article: [1]. I propose that this material be removed. Isambard Kingdom (talk) 17:04, 14 January 2017 (UTC)

---

In response to the proposed deletion:

Please notice the suggestions in the Talk page above for this article:

• "higher priority would be a "motivation" section

• "a plot of the distribution of aggregate event rates"

• "an example section"

The material that Mr. Kingdom proposes to delete provides exactly these: a motivation section, plots of event rates showing lack of fit, and an example section. If this material is deleted, there will be no motivation section, no plots, and no example section.

Please notice also the stated goal for Wikipedia articles:

https://en.wikipedia.org/wiki/Help:Editing

"The goal of a Wikipedia article is to create a comprehensive and neutrally written summary of existing mainstream knowledge about a topic."

The version of the article that Mr. Kingdom prefers lacked several items that are necessary for comprehensive coverage. In addition to lacking a motivation section, lacking any examples, and lacking any plots to illustrate lack of fit, it lacked the following.

• Example of calculation of the Hosmer-Lemeshow statistic
• Relationship to the Pearson chi-squared statistic, and why the Pearson statistic cannot be applied
• Limitations of the Hosmer-Lemeshow test
• Alternatives to the Hosmer-Lemeshow test

The material that Mr. Kingdom proposes to delete provides exactly these.

Mr. Kingdom indicates that the article is "possibly tedious". I would point out the frequent criticism that Wikipedia statistics articles are not understandable for non-statisticians. Here are a few quotes from the Talk pages of frequently-read Wikipedia statistics pages:

• "Would it be possible to write an introductory section that gives just a conceptual description ... before we enter the maths?"

• "This article is quite technical. It would be nice to have a simpler layman's description too."

• "Please, somebody, take pity on those of us who need more fundamental understanding, and write an introduction to this subject that would be useful and graspable by anybody with the basic interest to look it up. That's how to make Wikipedia better; make it useful."

There is clearly demand for introductory material directed to the non-statistician reader.

That said, there are readers with more advanced knowledge of statistics, such as Kingdom, who find the introductory material too detailed and tedious. To address this issue, I have moved the material to a section with the title "Introduction to the Hosmer-Lemeshow test for the novice".

Readers who have advanced knowledge of statistics and wish for a brief, mathematical, highly technical explanation need not read the introductory material. Readers who wish for a more comprehensible lay-oriented explanation can find it in the Introduction Section. 75.80.25.234 (talk) 20:01, 15 January 2017 (UTC)

I agree with Isambard Kingdom. The article shows a specific example study: like if somebody had to use the HL test and decided to put his stuff on Wikipedia, almost as is (only the R code is missing !). An example should not constitute 100% of an article. There are too many images (tables should not be just copy/pasted as images from R - as images!!! Didn't even bother to use Wiki tables). The "limitations" section is too short and is also something copy-pasted, from the web. There are links in the text (didn't even bother to use "ref" tags). No efforts shown here, someone just threw things and the community has to cope with it. Deleting is a way to cope with poor-quality material. Kiwipidae (talk) 07:22, 13 March 2018 (UTC)