Thread by @StatStas

By how much do response rate differ between the different population groups? Today, we will explore this with the @BRFSS data (cdc.gov/brfss/annual_d…) 🧵👇 1/11

The data set makes it possible to study these differentials as it is one of the relatively rare data sets with both the design weights and the calibrated weights... in our case, _WT2RAKE and _LLCPTWT. 2/11

(It's not me who likes to YELL, it is CDC. If you don't like the variable names, janitor them the way do like) 3/11

# libraries
library(tidyverse)
library(zip)
library(haven)
library(magrittr)
library(sjlabelled)
# 4/11

# get the data -- dumped in the current directory getwd()
temp_zip <- tempfile()
download.file("cdc.gov/brfss/annual_d…",temp_zip)
unzip(temp_zip)
read_xpt("LLCP2020.XPT") -> brfss2020
unlink(temp_zip)
# 5/11

# compute the raking ratio
brfss2020 %<>% mutate(raking_ratio = `_LLCPWT` / `_WT2RAKE`)
# and yes, I do like magrittr::`%<>% two-way pipe
# 6/11

# label values of education
brfss2020 %<>% mutate(EDUCA=labelled(EDUCA, label="Highest level of education",
labels=c("No school"=1, "Grade 1-8"=2, "Grade 9-11"=3,
"HS/GED"=4, "Some college"=5, "College or more"=6, "REF"=9)))
# 7/11

... waiting for another five or so years until #rstats supports labeled integers as a data structure properly... ah well...
8/11

# summary by education
brfss2020 %>% group_by(EDUCA) %>% summarize(mean_rr=mean(1/raking_ratio))

# 9/11

# difference between the extremes -- college-educated vs. HS dropouts
brfss2020 %>% group_by(EDUCA) %>% summarize(mean_rr=mean(1/raking_ratio)) %>%
ungroup() %>% summarize(diff=max(mean_rr)/min(mean_rr))

# 10/11

So the differences in response rates may be more than twofold. /fin

@threadreaderapp unroll or compile svp