Impact of weights

Analyzing Survey Data in R

Kelly McConville

Assistant Professor of Statistics

National Health and Nutrition Examination Survey (NHANES)

Conducted by the U.S. National Center for Health Statistics.
Goal: Understand the health of adults and children in the US.
It is collected using a 4 stage design.
Stage 0: The U.S. is stratified by geography and proportion of minority populations.
Stage 1: Within strata, counties are randomly selected.
Stage 2: Within counties, city blocks are randomly selected.
Stage 3: Within city blocks, households randomly selected.
Stage 4: Within households, people randomly selected.

NHANES

library(NHANES)
dim(NHANESraw)

20293    78

library(dplyr)
summarize(NHANESraw, N_hat = sum(WTMEC2YR))

# A tibble: 1 x 1 
      N_hat
      <dbl>
 1 608534400

NHANESraw <- mutate(NHANESraw, WTMEC4YR = WTMEC2YR / 2)

NHANES

NHANES_design <- svydesign(data = NHANESraw, 
                           strata = ~SDMVSTRA,
                           id = ~SDMVPSU, nest = TRUE, 
                           weights = ~WTMEC4YR)

distinct(NHANESraw, SDMVPSU)

# A tibble: 3 x 1
  SDMVPSU
    <int>
 1       1
 2       2
 3       3

Visualizing impact of weights

Survey-weighted and non-survey weighted bar plots of the distribution of race

Let's practice!

Analyzing Survey Data in R