Talk:Goodman and Kruskal's lambda

Statistics Mid‑importance

	This article is within the scope of WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.StatisticsWikipedia:WikiProject StatisticsTemplate:WikiProject StatisticsStatistics articles
Mid	This article has been rated as Mid-importance on the importance scale.

Mathematics Low‑priority

	Mathematics portal This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.MathematicsWikipedia:WikiProject MathematicsTemplate:WikiProject Mathematicsmathematics articles
Low	This article has been rated as Low-priority on the project's priority scale.

The example for the weakness of the lambda test seems to be wrong because it does not take the largest dependent values of each category. 163.1.208.200 (talk) 17:17, 8 May 2010 (UTC)[reply]

I second that - I also noticed the example for weakness seems to be incorrect.
Using Goodman and Kruskal's orginal formula I got: (222+218-222-200)/(2*350-222-200) = 18/278 ~ 0.065
If someone double-checks my calculation and it is correct I'll post an edit to article page revising the example but I don't know enough about the topic to directly speak to the weaknesses.142.76.1.62 (talk) 22:33, 11 December 2011 (UTC)[reply]

The point is that both ideas are correct. Lambda can be calculated in three ways: symmetric (L = 0.065), asymmetric LB (B explains A) = 0.00 and LA (A explains B) = 0.12. Only the asymmetric ones are usually seen in the textbooks (e.g. Siegel & Castellan, 1988). Now the example tells only one of these options. However, it seems that the critique is correct: there really seems to be challenge in Lambda because there IS a connection between the blood pressure and marital status. However, the example is NOT good for the problem in case because, naturally, the blood pressure would NOT explain the marital status and, hence, LB = 0.00 is quite a PROPER judgment of the situation! If changed the names of the variables in the hypothetical data, the basic challenge would be more obvious. JMe 37.136.85.180 (talk) 18:22, 11 October 2016 (UTC)[reply]

I added some explanations and removed the "Challenged" block. Hopefully, the explanations are somewhat better now ... --User:Haraldmmueller 11:09, 3 December 2018 (UTC)[reply]

I think that the article still lacks an explanation about the symmetric calculation, which makes it quite incomplete (e.g. it's the default in R: https://rdrr.io/cran/DescTools/man/Lambda.html). I didn't know about the GK Lambda until a few days ago, and when reading this wikipedia article I found it quite surprising that no researcher came up with a symmetric way of calculating it. I had to search on other websites to find out that there was such a thing and to learn more details about the calculations. I was about to leave a message on this discussion page about this issue. Is there some kind of warning we could put at the beginning of the article? To be honest, I find it a bit misleading by omission. 77.206.169.95 (talk) 11:23, 28 June 2020 (UTC)[reply]