Talk:Log-logistic distribution

Burr distribution
I've just noticed that Talk:Shape parameter says it's a generalization of the log-logistic. I've never heard of it though. Anyone who knows any more or feels qualified to relate the two please do. --Qwfp (talk) 22:45, 9 February 2008 (UTC)
 * Done it myself now. Qwfp (talk) 14:16, 14 February 2008 (UTC)

Formula for cdf is wrong
It must be 1-[The formula in the article] —Preceding unsigned comment added by 142.205.241.254 (talk) 22:24, 9 February 2009 (UTC)

no, it is correct
you are misinterpreting it. — Preceding unsigned comment added by Stephen Robertson (talk • contribs) 13:51, 19 February 2013 (UTC)

For AFT, vary shape or scale?
The Applications section cross-refers to AFT (accelerated failure time) models, and says that you can allow beta (shape) to vary but fix alpha (scale). But the AFT article referred to talks about varying a scale parameter. Does this (log-logistic) article have it the wrong way round?

Stephen Robertson (talk) 13:57, 19 February 2013 (UTC)

Mistake in article
I see the similar problem as Stephen Robertson. The fifth property says: "If X has a log-logistic distribution with scale parameter α and shape parameter β then Y = log(X) has a logistic distribution with location parameter log ⁡ ( α ) and scale parameter 1 / β ." It should be: "If X has a log-logistic distribution with location parameter α and shape parameter β then Y = log(X) has a logistic distribution with location parameter log ⁡ ( α ) and scale parameter 1 / β ." Please fix it. — Preceding unsigned comment added by 31.183.223.48 (talk) 20:59, 28 April 2019 (UTC)

Missing Context
The article repeatedly states that the log-logistic is used to model one application or the other. What is missing is whether the log-logistic is purely heuristic, or if it has a mathematical derivation underpinning its use. I get the impression that it is purely heuristic and replaces more difficult-to-manipulate distributions on account of having a closed-form CDF. 137.79.239.217 (talk) 20:35, 27 October 2023 (UTC)