О выборке и генеральной совокупности
Feb. 2nd, 2016 01:30 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
savin_eВсе вычисляют уровень значимости, кто-то доверительные интервалы, некоторые рассчитывают коэффициенты Байеса. Это удивительно, если учесть, что в большинстве психологических исследований - в отличие от избирательных опросов или контроля качества, исследователи не делают случайные выборки из генеральной совокупности или же определяют ее как таковую. Gigerenzer G., Marewski J.N. Surrogate Science The Idol of a Universal Method for Scientific Inference // Journal of Management. 2014. V.41. №2.DOI: 10.1177/0149206314547522.
A majority compute p values all the time, a minority always compute confidence intervals, and a few calculate Bayes factors, irrespective of the problem at hand. These routines are surprising, given that in most psychological experiments—unlike in election polls and quality control—researchers do not draw random samples from a population or define a population in the first place. Thus, no one knows to what population an inference actually refers.
A majority compute p values all the time, a minority always compute confidence intervals, and a few calculate Bayes factors, irrespective of the problem at hand. These routines are surprising, given that in most psychological experiments—unlike in election polls and quality control—researchers do not draw random samples from a population or define a population in the first place. Thus, no one knows to what population an inference actually refers.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2016-11-07 08:49 am (UTC)no subject
Date: 2016-11-07 11:04 am (UTC)