8.1 Introduction
What are the chances that this modern artist is Gen X or even younger, i.e., born in 1965 or later?
Speculating assumptions:
major modern art museums disproportionately display artists born before 1965.
\(n=100\) artists
\(\pi\) is the proportion of artists represented in major U.S. modern art museums that are Gen X or younger, it most likely falls below 0.5.
\(Beta(4,6)\) is the prior model for \(\pi\)
library(bayesrules)
library(tidyverse)
We use a dataset in the {bayesrules} package which contains data made available by MoMA.
data("moma_sample")
%>%head moma_sample
artist country birth death alive genx gender count
1 Ad Gerritsen dutch 1940 2015 FALSE FALSE male 1
2 Kirstine Roepstorff danish 1972 <NA> TRUE TRUE female 3
3 Lisa Baumgardner american 1958 2015 FALSE FALSE female 2
4 David Bates american 1952 <NA> TRUE FALSE male 1
5 Simon Levy american 1946 <NA> TRUE FALSE male 1
6 Pierre Mercure canadian 1927 1966 FALSE FALSE male 8
year_acquired_min year_acquired_max
1 1981 1981
2 2005 2005
3 2016 2016
4 2001 2001
5 2012 2012
6 2008 2008
genx
are artists 14 or younger:
%>%
moma_sample count(genx)%>%
mutate(prop=n/sum(n))
## genx n prop
## 1 FALSE 86 0.86
## 2 TRUE 14 0.14
We use a Beta-Binomial framework as \(Y\) (genx) follows a Binomial model.
\[Y|\pi \sim Bin(100,\pi)\] \[\pi \sim Beta(4,6)\]
The posterior distribution is obtained:
\(\alpha=4\) set to this value \(\beta=6\)
\[\pi|(Y=y) \sim Beta(\alpha + \beta, \beta +n-y)\]
\[\pi|(Y=14) \sim Beta(18,92)\]
Tasks in posterior analysis:
- estimation
- hypothesis testing
- prediction.