decide if there may be a large difference inside the quantity of testimonies that mention sweet, that point out fruit, or that mention neither candy nor fruit, among (a) the instances and the put up, (b) the times and the herald, and, (c) the publish and the usher in. However, is that a correct approach to make these three separate conclusions?

herraviuj

herraviuj

Answered question

2022-10-08

this is probably going to be a very dumb question, so my apologies in advance for that.
besides, i'm seeking to decide if there may be a large difference inside the quantity of testimonies that mention sweet, that point out fruit, or that mention neither candy nor fruit, among (a) the instances and the put up, (b) the times and the herald, and, (c) the publish and the usher in.
I ran this Chi-Square test with all three newspapers (from socscistatistics.com):
Chi-Square contingency table
However, is that a correct approach to make these three separate conclusions?
Or do I need to - instead - run three separate Chi-Square tests: (1) the Times and the Post (2) the Times and the Herald (3) the Post and the Herald

Answer & Explanation

bequejatz8d

bequejatz8d

Beginner2022-10-09Added 6 answers

The chi-squared test you showed in your link is an excellent first step. If you had rejected the null hypothesis that proportions of mentions of candy/fruit/neither are the same across all three newspapers, then it would be worthwhile to investigate between which pairs of newspapers the differences might lie. But you did not find evidence of such differences in the 3 × 3 table, so it seems best to stop there. I assume you did this test at the 5% level of significance.
However, you could run ( 3 2 ) = 3 comparisons between pairs of newspapers: T vs. P, T vs. H, P vs. H. However, to keep error probabilities within bounds for the whole project, you should do each of three tests at level α = .05 / 3 = .0167 = 1.67 % . This is called a 'Bonferroni multiple comparison procedure' (which you can google) because it is based on Bonferroni's (probability) Inequality.
Notes: (a) I ran these 3 tests. P-values as follows. TP (.767), TH (.195), PH (.671). Nothing there, not even close. (b) I'm wondering if articles that mentioned both fruit and candy are double-counted. (If there's much double counting, then assumptions for all of these analyses are violated, and conclusions possibly invalid.) Better then to go by the 'main' mention, or to have categories candy/fruit/both/neither.

Do you have a similar question?

Recalculate according to your conditions!

New Questions in High school probability

Ask your question.
Get an expert answer.

Let our experts help you. Answer in as fast as 15 minutes.

Didn't find what you were looking for?