One Last Gripe

From Task 2:

Use a for loop to compute Fisher’s Exact Test for each country in the dataset and extract the p-value from the object returned by function fisher.test

Is there any evidence of an association between GMO use and disease incidence in each country?

  • 🤯
  • If you have multiple p-values, you need to correct them (e.g., using Holm-Bonferroni)
  • If you have a set of 20 countries, by the arbitrary p = 0.05 standard you'd expect a significant result even if the null hypothesis is true!
    • also known as the "every food both causes and cures cancer" problem

One Last Gripe

From Task 2:

Use a for loop to compute Fisher’s Exact Test for each country in the dataset and extract the p-value from the object returned by function fisher.test

Is there any evidence of an association between GMO use and disease incidence in each country?

  • 🤯
  • If you have multiple p-values, you need to correct them (e.g., using Holm-Bonferroni)
  • If you have a set of 20 countries, by the arbitrary p = 0.05 standard you'd expect a significant result even if the null hypothesis is true!
    • also known as the "every food both causes and cures cancer" problem