RS 4

Research Scenario 6.3: Mindfulness Training and Stress

library(readxl)
library(ggpubr)

## Loading required package: ggplot2

library(effectsize)
library(rstatix)

## 
## Attaching package: 'rstatix'

## The following objects are masked from 'package:effectsize':
## 
##     cohens_d, eta_squared

## The following object is masked from 'package:stats':
## 
##     filter

Dataset6.4 <- read_excel("/Users/asfia/Desktop/Dataset6.4.xlsx")

Pre_Test <- Dataset6.4$Stress_Pre
Post_Test <- Dataset6.4$Stress_Post

Differences <- Post_Test - Pre_Test
print(Differences)

##  [1]  -8.044361120  -3.529426078 -26.350753567  -6.276567859  -3.701374633
##  [6] -15.552084134  -4.356073741  -3.647770620  -0.008788193  -5.302543769
## [11]  -8.806996646  -3.200139192  -1.385585292 -19.179181634   3.899627158
## [16]  -6.612429249  -2.533382079  -1.317457905  -0.782853924 -32.441286182
## [21]  -4.132848699  -6.536275900  -6.896524983  -5.238701383 -36.697172771
## [26] -28.957287115  -9.817187881 -10.508880209 -19.386322836 -10.616491219
## [31] -13.700632408  -7.433337532 -20.087719095  -7.326665368 -15.099150128

STEP 4: DESCRIPTIVE STATISTICS

mean(Pre_Test, na.rm = TRUE)

## [1] 51.53601

median(Pre_Test, na.rm = TRUE)

## [1] 47.24008

sd(Pre_Test, na.rm = TRUE)

## [1] 17.21906

mean(Post_Test, na.rm = TRUE)

## [1] 41.4913

median(Post_Test, na.rm = TRUE)

## [1] 40.84836

sd(Post_Test, na.rm = TRUE)

## [1] 18.88901

STEP 5 & 6: VISUALIZATIONS (Difference Score)

hist(Differences,
     main = "Histogram of Difference Scores (Post - Pre)",
     xlab = "Change in Stress Levels",
     ylab = "Frequency",
     col = "purple",
     border = "darkblue",
     breaks = 15)

Skewness: Negatively skewed , Kurtosis: Bell Curve. So based on Reports we cannot use the Dependent T-test

Boxplot

boxplot(Differences,
        main = "Distribution of Score Differences (Post Stress - Pre Stress)",
        ylab = "Change in Stress Levels",
        col = "lightblue",
        border = "darkblue",
        horizontal = FALSE)

The box plot appears to be abnormal as two outliers are out of the whiskers, so based on Reports we can use the Wilcoxon-Sign Rank

Method 3: Shapiro-Wilk

shapiro.test(Differences)

## 
##  Shapiro-Wilk normality test
## 
## data:  Differences
## W = 0.87495, p-value = 0.0008963

The value of p < 0.05 (0.0008963), so the data is not normal, so based on Reports we use the Wilcoxon Sign Rank

Interpretation: After conducting all three normality tests, it is clear we must use a Wilcoxon Sign Rank.

Wilcoxon Sign Rank

wilcox.test(Pre_Test, Post_Test, paired = TRUE)

## 
##  Wilcoxon signed rank exact test
## 
## data:  Pre_Test and Post_Test
## V = 620, p-value = 2.503e-09
## alternative hypothesis: true location shift is not equal to 0

As the value of p < 0.05 (p = 0.000000002503), this means the results were SIGNIFICANT.

Calculate the Effect Size

df_long <- data.frame(
  id = rep(1:length(Pre_Test), 2),
  time = rep(c("Pre", "Post"), each = length(Pre_Test)),
  stress = c(Pre_Test, Post_Test)
)
  
effect_size_result <- wilcox_effsize(df_long, stress ~ time, paired = TRUE)
print(effect_size_result)

## # A tibble: 1 × 7
##   .y.    group1 group2 effsize    n1    n2 magnitude
## * <chr>  <chr>  <chr>    <dbl> <int> <int> <ord>    
## 1 stress Post   Pre      0.844    35    35 large

As the size of the effect is (0.84), this indiactes the effect is ‘Very Large’.

Report the Results

There was a significant difference in the stress between Pre-Stress Group (Mdn = 47.24) and Post-Stress (Mdn = 40.84), V = 620, p = .000000002503 The effect size was very large (r₍rb₎ = .84).