How is have to be a dataset to perform an ANOVA test in R?

I have three columns, one per group, with numeric values. I want to analyze them using an Anova test, but I found applications when you have the different groups in a column and the respective values in the second column. I wonder if it is necessary to reorder the data like that, or if there is a method that I can use for the columns that I currently have. Here I attached a capture:

Thanks!

Solution

You can convert a wide table having many columns into another table having only two columns for key (group) and value (response) by pivoting the data:

library(tidyverse)

# create example data
set.seed(1337)
data <- tibble(
  VIH = runif(100),
  VIH2 = runif(100),
  VIH3 = runif(100)
)
data
#> # A tibble: 100 × 3
#>       VIH   VIH2   VIH3
#>     <dbl>  <dbl>  <dbl>
#>  1 0.576  0.485  0.583 
#>  2 0.565  0.495  0.108 
#>  3 0.0740 0.868  0.350 
#>  4 0.454  0.833  0.324 
#>  5 0.373  0.242  0.915 
#>  6 0.331  0.0694 0.0790
#>  7 0.948  0.130  0.563 
#>  8 0.281  0.122  0.287 
#>  9 0.245  0.270  0.419 
#> 10 0.146  0.488  0.838 
#> # … with 90 more rows

data %>%
  pivot_longer(everything()) %>%
  aov(value ~ name, data = .)
#> Call:
#>    aov(formula = value ~ name, data = .)
#> 
#> Terms:
#>                      name Residuals
#> Sum of Squares   0.124558 25.171730
#> Deg. of Freedom         2       297
#> 
#> Residual standard error: 0.2911242
#> Estimated effects may be unbalanced

^{Created on 2022-05-10 by the reprex package (v2.0.0)}