Workshop 6: Statistical analysis in R

Fursham Hamid

Overview

Duration: 2 hours

Objectives

1. Differentiate continuous and discrete variables
2. Test variables for normality
3. Perform parametric tests on continuous variables
4. Perform nonparametric tests on discrete variables
5. Test for correlation between variables
   

Lesson preparations

In this workshop, we will perform various statistical analyses using R and interpret the outputs together. Please ensure that you have BOTH R and RStudio installed on your computer. Click on the links below to download these software.

• Latest version of R
  • For Windows
  • For Mac
  • For Linux
• Latest version of RStudio

Most of the functions used in this lesson will be from the stats package, which is pre-built with R and RStudio. In addition, we will be manipulating and plotting data using the Tidyverse package. Please install Tidyverse (if you have not) on your computer by typing the following command in RStudio console.

install.packages("tidyverse")

# to check if installation is successful,
# run this and you should not see an error message
library(tidyverse)

During the lesson, we will be using etherpad, an online collaborative platform for our discussions and activities. You may bookmark this etherpad page before the lesson commence!

Lesson schedule

10:00	Class check-in
10:05	1. Overview on data types	What are the different data types?
10:10	2. Generate basic descriptive statistics	How do I obtain mean, median and standard deviation for my dataset? How do I breakdown these statistics for each of my samples?
10:20	3. Testing data for normality	How can I tell if my data is normally distributed?
10:35	4. One-sample tests of means	What is null and alternate hypothesis? How do I test if the mean of my normally distributed data is significantly different from the null hypothesis? What do I do if my data is not normally distributed?
11:00	Break
11:10	5. Two-sample tests of means	How do I compare the means of two of my samples?
11:20	6. More than 2 sample tests of means	How do I compare the means of more than 2 samples? What assumptions do these tests hold?
11:35	7. Testing categorical variables for independence	How do I perform statistical tests on categorical variables?
11:50	8. Test variables for correlation	How do I test for correlation for two of my variables?
12:00	Finish