Question

我想对性别（男性，女性）之间的年龄差异进行t检验，但需要额外的分组变量grp（a，b）。数据在数据帧（df）中。

这让整个样本按性别划分：

with(df, t.test(age~sex))

这让我按照整个样本分组：

with(df, t.test(age~grp))

我希望按性别和群体年龄划分，即b中女性与女性比较，b中男性与男性比较。

Answer 1

使用原生R：

lapply(split(df,df$sex),function(x)with(x, t.test(age~grp)))

$f

    Welch Two Sample t-test

data:  age by grp
t = 1.3978, df = 42.029, p-value = 0.1695
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -3.241762 17.854665
sample estimates:
mean in group 1 mean in group 2 
       56.50000        49.19355 


$m

    Welch Two Sample t-test

data:  age by grp
t = 0.33265, df = 36.741, p-value = 0.7413
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
 -7.457013 10.385584
sample estimates:
mean in group 1 mean in group 2 
       54.00000        52.53571

Answer 2

df <- data.frame(
  age = sample(x = 20:80, 100, TRUE),
  sex = sample(c("m", "f"), 100, TRUE),
  grp = sample(1:2, 100, TRUE)
)

library(plyr)

# Split df by "sex" and apply function to each subset of df. Returns a list of the return values.
dlply(.data = df, .variables = "sex", .fun = function(x) {
  with(x, t.test(age~grp))
})

带有两个分组变量的T检验

2 个答案: