Trying to use the Bonferroni method to hypothesis test-CodePudding

I am trying to hypothesis test using the bonferroni method although I get an error message saying I can't pool together SD, does anyone know this problem and how to solve the code?

Code used:

with(final_data, pairwise.t.test, Concentration_of_PM2.5, Life_expectancy,
 p.adjust.method = 'bonferroni')

Error message;

function (x, g, p.adjust.method = p.adjust.methods, pool.sd = !paired, 
paired = FALSE, alternative = c("two.sided", "less", "greater"), 
...) 
{
     if (paired && pool.sd) 
        stop("pooling of SD is incompatible with paired tests")

Dataset snipet;

head(final_data, 10)
           Country             Continent Life_Expectancy Adult_Mortality Concentration_of_PM2.5       GDP GDP_Level
1          Afghanistan Eastern Mediterranean        62.68935       245.22490                  55.14  1896.993  Very Low
2              Albania                Europe        76.37373        96.40514                  18.07 11868.179    Medium
3              Algeria                Africa        76.36365        95.02545                  35.18 15036.364    Medium
4               Angola                Africa        62.63262       237.96940                  38.29  6756.935       Low
5  Antigua and Barbuda              Americas        74.99754       119.86570                  21.03 23670.302      High
6            Argentina              Americas        76.94621       111.42880                  12.58 20130.408      High
7              Armenia                Europe        74.83788       116.43580                  33.84  8808.573       Low
8            Australia       Western Pacific        82.90018        60.72528                   7.14 47305.880 Very High
9              Austria                Europe        81.87031        61.88845                  12.15 51809.514 Very High
10          Azerbaijan                Europe        73.07719       117.64890                  20.99 17417.087      High

CodePudding user response：

PM2.5 reduces life expectancy in deprived areas (low GDP), if the correlation between these two continuously scaled variables is negative and significant (p-value < 0.05):

library(tidyverse)

final_data %>%
  filter(GDP_Level %in% c("Very Low", "Low")) %>%
  cor.test(~ Concentration_of_PM2.5   Life_Expectancy,
     data = ., method = "pearson")

This looks for a linear relationship between these two variables. Use method = "spearman instead if non-linear but monotonous relationships should be studied.

However, this is just one test for one hypothesis, so there is no Bonferroni required.