0

First time asking a question. I have done my best to conduct my own searches and have looked at the manual. I hope I am not posting a repeat or an off-topic.

I have an output from using the pairwise.prop.test with a bonferroni correction using:

pairwise.prop.test(fishingsept13,p.adjust.method = 'bonferroni')

It came out just fine and ran smoothly.

My question is more on the interpretation of the p-values printed. I know that the Bonferroni correction for my data would be that anything less than 0.01 is significant. These are my results:

I assume that I would interpret them with my known Bonferonni and determine their significance from there. I'm paranoid that I am assuming wrong.

JayPeerachai
  • 3,499
  • 3
  • 14
  • 29
Cat
  • 1
  • 1

1 Answers1

1

Actually, the pairwise comparison functions in R already correct the p values for you. The p.adjust.method argument you used told R to perform a multiple comparisons correction by the Bonferroni method. Since you performed 3 separate comparisons, the p values you see have already been multiplied by 3: you don't have to adjust your significance threshold. To see for yourself, look what happens when you run

pairwise.prop.test(fishingsept13,p.adjust.method = 'none')

Matt
  • 954
  • 1
  • 9
  • 24
  • Oi. I wish I had thought of/caught on to that or taken a step back to breathe and approach it more logically. Thanks so much for your response! – Cat Jun 14 '17 at 17:47