在 svydesign 中使用 BRFSS 数据嵌套年份

Question

我正在努力将多年的国家 BRFSS 数据合并为一组，并将适当的复杂调查设计与 survey 包结合起来，以便能够计算不确定性。我看过几个关于如何在一年内执行此操作的示例，并且知道在执行多年时我需要嵌套 year，但我不太确定如何包含它。这是我使用变量 finalwt:

重新加权后的数据

> glimpse(df)
Rows: 1,756,594
Columns: 15
$ year        <dbl> 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016, 2016~
$ age         <dbl> 43, 59, 80, 70, 18, 65, 74, 76, 43, 56, 75, 61, 57, 58, 70, 62, 65, 52, 37, 80, 36, 34, 80, 66, 72, ~
$ sex         <fct> Male, Female, Female, Male, Male, Female, Female, Female, Female, Male, Female, Female, Male, Female~
$ race        <fct> White, White, White, White, White, White, White, White, White, White, White, White, White, Other, Wh~
$ insured     <fct> Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, Yes, ~
$ met_mam     <dbl> NA, 1, NA, NA, NA, 1, 1, NA, NA, NA, NA, 0, NA, 1, 0, 1, 1, NA, NA, NA, NA, NA, NA, NA, NA, NA, 1, N~
$ met_pap     <dbl> NA, 1, NA, NA, NA, 0, NA, NA, 1, NA, NA, 0, NA, 1, NA, 1, 1, 1, 0, NA, NA, 1, NA, NA, NA, NA, 0, 1, ~
$ met_crc     <dbl> NA, 1, NA, 1, NA, 1, 1, NA, NA, 1, 1, 0, 1, 1, 1, 1, 1, 0, NA, NA, NA, NA, NA, 1, 1, NA, 1, NA, 1, 1~
$ met_lcs     <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, ~
$ psu         <dbl> 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2.016e+09, 2~
$ ststr       <dbl> 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11011, 11~
$ llcpwt      <dbl> 767.8446, 329.6599, 290.7493, 211.0392, 1582.5398, 540.0474, 323.2903, 522.1566, 973.1335, 205.5001,~
$ finalwt1618 <dbl> 395.39989, 169.75764, 149.72072, 108.67418, 814.92545, 278.09623, 166.47765, 268.88341, 501.11298, 1~
$ finalwt1719 <dbl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, ~
$ finalwt     <dbl> 203.69399, 87.45225, 77.13005, 55.98453, 419.81655, 143.26390, 85.76254, 138.51783, 258.15309, 54.51~

这是我的 svydesign() 代码：

#build in complex survey design
options(survey.lonely.psu = "adjust")
des <- svydesign(ids= ~1, strata= ~ststr, weights= ~finalwt, nest=T, data = df)

请指教。谢谢。

Answer 1

您想将年份视为额外的层次。它们是分层的，因为每年抽样的 PSU 数量是预先固定的。所以，正如@Anthony 在评论中所说

brfss_design <- svydesign( id = ~ psu , strata = ~ interaction( ststr , year ) , data = df , weight = ~ finalwt , nest = TRUE )

其中finalwt需要原始权重除以年数，所以它加起来是美国人口的一份（而不是年数的份数）

在 svydesign 中使用 BRFSS 数据嵌套年份

Nesting for Year with BRFSS Data in svydesign

r

survey