Association between two paired nominal variables

A notebook made by Paul Amat, amat-design.com

Starting session

library(readr)
data <- read_csv("data.csv", show_col_types = FALSE)

Context and objectives

Research hypothesis

Statistical hypothesis

H0:

H1:

Data preparation and exploration

Contingency table

t <- table(data$mood, data$mood.after)
t

       
        Happy Sad
  Happy    65  47
  Sad      60  28

Parametric test

Conditions

Independent observations

The independence of observations is guaranteed by the recruitment process.

N>30 in each group

t[2,1]+t[1,2] > 30

[1] TRUE

For each combination whose two modalities are different, the sum of this combination must be greater than 30.

Test

mcnemar.test(data$mood, data$mood.after, correct = F) #ok jamovi

    McNemar's Chi-squared test

data:  data$mood and data$mood.after
McNemar's chi-squared = 1.5794, df = 1, p-value = 0.2088

Effect size

Non-parametric test

mcnemar.test(data$mood, data$mood.after) #ok jamovi

    McNemar's Chi-squared test with continuity correction

data:  data$mood and data$mood.after
McNemar's chi-squared = 1.3458, df = 1, p-value = 0.246

The p-value of X2 continuity correction (Yates continuity) is to read if effective below 30 for discordant pairs.

Results and recommendations