p_highed: High News Interest Oversample

In kuriwaki/svysim: Realistic Simulations of Biased Survey Sampling

Description

Propensity Score that oversamples high education and high-news interest people

Usage

p_highed(data)

samp_highed(data, n)

Arguments

data	A tibble with the columns race, educ, newsint, as coded by the CCES
n	Sample size to sample

Details

We model the true propensity score as being a separable function of race, education, news interest, and whether or not you are in an urban district. There is no explicit correlation between selection and the outcome (say, partisanship), so Meng's rho would be non-zero only so far as the covariates correlates with race, education, and news interest.

Value

A vector of propensity scores ranging from 0 to 1

Source

Inspired by Lauren Kennedy's code in rstanarm

Examples

p_highed(sample_n(pop_cces, 100))

kuriwaki/svysim documentation built on Jan. 27, 2021, 5:37 a.m.