Experimental Study

The research design project entails gernerating the data for an experimental study with a control and test group where only the test group would receive treatment. The study will be conducted on clinically depressed patients who will receive newly tested treatment pills to see if symptoms can be alleviated. The control group will receive a placebo pill while the test group will receive the real pills. After a six month trial, the sample will be surveyed on a Likert Scale from 1 - 5 to see if their frequency of mood swing has been lowered and if their overall mental state has been improved.

As the experimental design consists of two independent sample groups – a controlled and experimental group, I will use a parametric test, an unpaired two-samples t-test to compare the means of two independent group’s Likert scale.

 

Generating Data

Code used to generate responses for control and test group on Likert scale for frequency of mood swings and overall mental state improvement.  

Frequency of Mood Swings

Likert Scale: 1 – not recently, 2 – once in the 3 months, 3 – once a month, 4 – once a week, 5 – almost every day  

Both groups contain 150 individuals each who ranked frequency of mood swings with predetermined distribution of Likert Scale shown in prob=c().

  • Experimental Group:
freq_moodswingsXG = sample(1:5, size=150, replace=TRUE, prob=c(.02,.367,.327,.167,.12))
    barplot(table(freq_moodswingsXG))

  • Control Group:
freq_moodswingsCG = sample(2:5, size=150, replace=TRUE, prob=c(.15,.19,.38,.28))
    barplot(table(freq_moodswingsCG))

Generating Dataframe for Frequency of Mood Swings

The code below generates a dataframe by creating a variable that labels which group each scale belongs to then merges the two tables together.

f_moodswingsXG <- cbind(group = "Experimental Group", scale = freq_moodswingsXG)
f_moodswingsCG <- cbind(group = "Control Group", scale = freq_moodswingsCG)
freq_moodswings <- rbind(f_moodswingsXG , f_moodswingsCG)

 

Mental Health State

Indivudals rate if their overall mental state has improved throughout the study on a Likert Scale: 1 – strongly disagree, 2 – disagree, 3 – neutral, 4 – agree, 5 – strongly agree.  

Both groups contain 150 individuals each who ranked frequency of mood swings with predetermined distribution of Likert Scale shown in prob=c().

  • Experimental Group:
men_stateXG = sample(1:5, size=150, replace=TRUE, prob=c(.01,.1,.16,.42,.31))
    barplot(table(men_stateXG))

  • Control Group:
men_stateCG = sample(1:5, size=150, replace=TRUE, prob=c(.27,.41,.19,.1,.03))
    barplot(table(men_stateCG))

Generating Dataframe for Mental Health State

The code below generates a dataframe by creating a variable that labels which group each scale belongs to then merges the two tables together.

m_stateXG <- cbind(group = "Experimental Group", scale = men_stateXG)
m_stateCG <- cbind(group = "Control Group", scale = men_stateCG)
men_state <- rbind(m_stateXG, m_stateCG)

 

Frequency of Usage

Likert Scale: 1 – not recently, 2 – once in the 3 months, 3 – once a month, 4 – once a week, 5 – almost every day  

Both groups contain 150 individuals each who ranked frequency of mood swings with predetermined distribution of Likert Scale shown in prob=c().

  • Experimental Group:
freq_usageXG = sample(3:5, size=150, replace=TRUE, prob=c(.205,.337,.46))
    barplot(table(freq_usageXG))

  • Control Group:
freq_usageCG = sample(2:5, size=150, replace=TRUE, prob=c(.053,.313,.253,.38))
    barplot(table(freq_usageCG))

Generating Dataframe for Frequency of Usage

The code below generates a dataframe by creating a variable that labels which group each scale belongs to then merges the two tables together.

freq_usageXG <- cbind(group = "Experimental Group", scale = freq_usageXG)
freq_usageCG <- cbind(group = "Control Group", scale = freq_usageCG)
freq_usage <- rbind(freq_usageXG, freq_usageCG)

 

Gender

The code below generates a dataframe to distinguish gender in the dataset that evenly distributes both genders into 150.

# Generating male & female data for experimental group & control group
  G_XG = sample(c('M', 'F'), 150, replace=TRUE, prob=c(0.5, 0.5))
  G_CG = sample(c('M', 'F'), 150, replace=TRUE, prob=c(0.5, 0.5))
# Labeling data for experimental group and control group
  G_XG <- cbind(group = "Experimental Group", gender = G_XG)
  G_CG <- cbind(group = "Control Group", gender = G_CG)
#Creating dataframe for gender with both groups
  gender <- rbind(G_XG, G_CG)

 

Converting Data to Excel File

data = read.csv("/Users/loanple/Desktop/Tidy_data.csv")
  • Data Structure:
str(data)
## 'data.frame':    300 obs. of  9 variables:
##  $ Respondent        : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ Group             : Factor w/ 2 levels "Control Group",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ Fmswing_Scale     : int  4 2 3 2 3 3 2 3 5 3 ...
##  $ Fmswing_Meaning   : Factor w/ 5 levels "Almost every day",..: 4 5 3 5 3 3 5 3 1 3 ...
##  $ Mstate_Scale      : int  5 4 4 5 4 4 4 5 4 3 ...
##  $ Mstate_Meanning   : Factor w/ 5 levels "Extremely good",..: 1 3 3 1 3 3 3 1 3 4 ...
##  $ Freq_usage        : int  5 5 5 5 3 5 4 5 5 5 ...
##  $ Freq_usage_Meaning: Factor w/ 4 levels "Every day","Every few days",..: 1 1 1 1 4 1 2 1 1 1 ...
##  $ Gender            : Factor w/ 2 levels "F","M": 1 1 2 2 2 1 2 1 2 2 ...

 

Reviewing Data Statistics

  • Mean for each group on the Mental State Scale:
##      Control Group Experimental Group 
##                  2                  4
  • Mean for each group on the Frequency of Mood Swing Scale:
##      Control Group Experimental Group 
##                  4                  3
  • Mean for each group on the Frequency of Usage:
##      Control Group Experimental Group 
##                  4                  4

 

  • Box Plot for Frequency Mood Swing by Group:

  • Box Plot for Mental State by Group:

 

Analysis Methodology - T Test

  • Packages used:
library("ggpubr")

 

  • T test for Frequency of Mood Swings:
t.test(Fmswing_Scale ~ Group, data=data, var.equal=TRUE)
## 
##  Two Sample t-test
## 
## data:  Fmswing_Scale by Group
## t = 8.1849, df = 298, p-value = 8.021e-15
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  0.7443717 1.2156283
## sample estimates:
##      mean in group Control Group mean in group Experimental Group 
##                         3.946667                         2.966667
  • T test for Overall Mental State:
t.test(Mstate_Scale ~ Group, data=data, var.equal = TRUE)
## 
##  Two Sample t-test
## 
## data:  Mstate_Scale by Group
## t = -13.951, df = 298, p-value < 2.2e-16
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.916985 -1.443015
## sample estimates:
##      mean in group Control Group mean in group Experimental Group 
##                         2.186667                         3.866667

 

Conclusion

  • Frequency of Mood Swings
    • The T-statistic is 8.18 while the df is 298 with a corresponding p-value of 8.021e-15. Considering the p-value is significantly less than our chosen alpha 0.05, the null hypothesis can be rejected, indicating there is a significant difference in the group taking marijuana and those who did not for frequency of mood swings.
  • Overall Mental State
    • The T-statistic is -13.95 while the df is 298 with a corresponding p-value of 2.2e-16. Considering the p-value is significantly less than our chosen alpha 0.05, the null hypothesis can be rejected and the alternate hypothesis is accepted. The alternate hypothesis indicates that there is significant difference for those who have actually been prescribed marijuana. This shows that marijuana consumption can improve overall mental state.

From the two t-test conducted on the two independent sample groups, there is evidence that shows medical marijuana consumed can positively effect symptoms in clinically depressed individuals by reducing frequency of mood swings, and increasing overall mental state.