MonashHonoursStatistics

Home

Acknowledgements: This MonashHonoursStatistics git repo is forked from Josh Wiley’s original repo. Dr. Wiley created this course and some of the material (i.e., your “readings”) remains his original work, although I have been updating and changing it since 2021.

The vision of this unit is to train you in applied statistics and data science for Psychology. These skills are useful not only for quantitative research but also are highly in demand industry/job skills. With industry and modern, agile teams in mind, we will be using several tools that are common tools used in industry. Some of these you may have used, some may be new.

1. Microsoft Teams – team communication

Teams is an instant messaging / chat platform. It allows easier, real time communication, sharing code, files, and images than email. A channel that sits under the “PSY Honours 2022” Team called “PSY4210 - Data Science Forum” has been created, and if you are enrolled in the unit, you will automatically be added to the channel.

I will place all official announcements on both Teams and Moodle, but most of the day to day communication will be on Teams. You will not be able to stay up to date in the unit (or any of your other Honours units) without Teams. It provides real time interactions that are closer to how real teams work and operate. Please feel free to chat openly, ask and answer questions with each other in this space. If you need specific help from one of the teachers you can use “@name” to mention them specifically and tag them to look at a particular message.

A note on screenshots: It can be very useful to provide screenshots of errors that you need help troubleshooting on Teams. Please be aware, that small errors and debugging of your lab reports are fine to share. However, you should avoid posting entire sections of your lab report code and writeup to Teams. Lab reports are not collaborative assignments and you must work on them independendently. If you are unsure, please use the Chat function on Teams to send a DM to myself and/or the tutors. Relatedly, while we can help you debug your code and understand the errors you might be getting, we cannot completely solve your coding issues, since that is your independent work. We can give you hints as to why your code did not run - we cannot re-write it for you.

On Teams, as with all communication, the rule of 48 business hours still applies. On Teams, your message is more likely to be answered by myself, the unit coordinator, a tutor, or a fellow student much more quickly than emailing everyone separately. There are also times when I am available at my computer to answer messages quickly - however, I cannot guarantee when I will be “live” on Teams. I will do my best to answer messages when I can, but please do not expect answers from myself or the other instructors before a 48 business hour period (i.e., not overnight or on the weekends).

2. GitHub – collaborative code

GitHub is an amazing platform and tool for building software, storing and managing code with over 40 million people actively using it. GitHub performs version control which just means keeping a history of all changes to a file, and lets you store and share files. By keeping a history, multiple people can work on files and then merge or combine their changes so it is common to use it for software development. Although its very powerful, in this unit, we will just introduce it and use it in a simple way. The lecture content and other support resources (worksheets, etc.) will be made available through GitHub. To use GitHub for this unit, follow these steps:

  1. Create an account here, or if you already have an account, just sign in.

  2. Once you are signed up, download the desktop app here.

  3. Go to the repository (a collection of files and folders in GitHub) for PSY4210, located here. Then click on “Clone or download” to get all the files synced to your own computer.

  4. Make sure to regularly “pull” (download) the latest version to your computer. We will update files regularly, for example adding the latest information for later lectures and updating or clarifying any materials as you all ask questions or find any errors. That is the whole beauty of GitHub: it lets files be edited and updated keeping a history but also letting you get the very latest versions.

3. Install and Setup R

R is primarily a language. To have a nicer way to interact with and use R we will use RStudio. RStudio is known as an Integrated Development Environment or IDE. It is a program that helps to keep your code, output, graphs, and files all together in one place. Please follow these steps on your own laptop/computer so that you can work on the unit and later in your studies or work from anywhere and do not depend on lab computers or a good internet connection.

As an alternative, if you are having trouble or cannot install R, as a Monash member, you can access it through Monash MoVE here. When you login, choose “RStudio” and you are ready to go.

Note, with many of the instructions, especially on a Mac, a lot of output code will be generated. You do not need to be able to read or understand this and it is normal.

Windows OS

  1. Download R by going here and clicking on “Download R for Windows” then click on “install R for the first time” which should let you “Download R 4.0.4 for Windows”. If you are having difficulty finding it, this link may help get you straight there.

  2. Once you have downloaded R double click the file to install it on your own computer. Follow the step by step installation guide that pops up.

  3. Download RStudio this is an interface to R and makes it much nicer to work with. To download it, just go here and then under the “Installers” pick your operating system (Windows 7+).

  4. Once you have downloaded RStudio, double click to install it and follow the step by step installation guide.

  5. Once you think you’ve followed all the installation steps, move on to the next section to Try R.

macOS

  1. The first step in getting R running on macOS is to make sure that you have the latest version of macOS. This is important as the installation process differs across versions. At the time of writing, the latest version is macOS Mojave 10.14.6. You can get help with upgrading from Apple here.

  2. Install X Code from App Store. After installation, open it to accept terms otherwise it may not work. You can see a video guide here.

  3. Once you have X Code installed, you also need to install command line tools. Open Terminal (if you cannot find it, try spotlight search) and type in: xcode-select --install then hit enter to run. See a video guide here. If you run into any access issues, you may need to enable root. You can do that by type in Terminal the following: dsenableroot and hit enter/return to enable root user. Note that at the terminal, if you are asked to enter a password, type the password you use to login to your Mac and press enter/return. When typing your password in the terminal, no characters will appear, but it is still being entered.

  4. Install XQuartz/X11. Go here, download and run the file and follow any on screen instructions.

  5. Install gfortran: go here and download gfortran 8.2 for Mojave. Download and install following any instructions.

  6. Install homebrew for macOS. Go here and follow the “Install Homebrew” steps. If you run into any access issues, you may need to enable root. You can do that by type in Terminal the following: dsenableroot and hit enter/return to enable root user. Note that at the terminal, if you are asked to enter a password, type the password you use to login to your Mac and press enter/return. When typing your password in the terminal, no characters will appear, but it is still being entered.

  7. Install openssl which will allow R to securely download files and packages from the internet. Do this by opening the terminal (you can search for “terminal” or look in the launchpad) and type this code once the terminal opens and press enter: brew install openssl

  8. Install libgit2 which is needed for one of our graphing packages. Do this by opening the terminal (you can search for “terminal” or look in the launchpad) and type this code once the terminal opens and press enter: brew install libgit2

  9. Download R by going here and clicking on “Download R for (Mac) OS X” then download the download version 3.6.3. Once you have downloaded R make sure to install it to your apps.

  10. Download RStudio this is an interface to R and makes it much nicer to work with. To download it, just go here and then under the “Installers” pick your operating system. Once you have downloaded RStudio, make sure to install it to your apps.

  11. Make sure that R.app and RStudio.app are able to access disk resources needed. Follow this guide to give them the necessary permissions.

  12. Install clang: go to R page here From there, download and install clang-7.0.0.pkg. do not install clang-8 as that is for the beta version of R.

  13. Once you think you’ve followed all the installation steps, move on to the next section to Try R.

4. Try R

To make sure that R and RStudio are installed and working correctly, follow these steps:

  1. Find and open RStudio on your computer. On Windows OS this should be under all of your programs/applications (on Windows 10 open the start menu and search for RStudio). On macOS you can try spotlight search, but if it installed but does not show on spotlight search, try looking in the launchpad. Once you have RStudio opened, there should be several “windows”.

  2. Go to the window named “Console” and click at the prompt. The prompt is just this symbol: “>”.

  3. At the console prompt, type this code and then press “enter”/”return”:


3 + 1

If everything worked, R should return the answer, “[1] 4”. Note: the [1] just means that the answer, 4, is in position 1. If not, double check the installation steps. If you are stuck, post on the Moodle Discussion Forum promptly so we can help you figure things out.

Assuming everything is working, try to run this code at the console to install a few packages we will be using this semester. R may generate a lot of output and strange code. Most of that is OK. However, you should keep notes on any Error messages (just copy and paste from the console) and raise these on the class discussion boards if you are stuck. Note that you will not be using any of these right now. Just try to run it so everything is installed and ready to go.

Please Post problems and errors to Teams, also will help us troubleshoot so that in later weeks when we present analyses, you have a fully functioning R on your system.

5. Learn more about RStudio

To find out a bit more about RStudio, you can watch this <5 minute video.

6. DataCamp: Learn about R

To begin learning R we are going to be using the online platform, DataCamp. Follow the instructions on Moodle (under Assessments) for getting your account setup and use the Invite Link there. You should be added to the Organization called “Statistics and Data Science for Psychology 2022”. Under “My assignments” you can see a list of chapters. Follow the schedule in the Table on Moodle.

Note the first 1-2 weeks will be fairly intense as to follow the class, you need to get up to a basic ability to read and follow R code.

The DataCamp activities are official homework and are assessed, based on the number completed (and you must have minimum 10,000 XP before you lose marks - see Moodle for grading explanation).

5. Install R Packages

This section will have you pre-install many packages we will use during the honours unit.

Note, during installation, you may get asked if you want to install more recent packages from source or something like: “there is a binary version available but the source version is later”. In R, binary packages are pre-built on servers and you download a ready-to-use version. Like buy a complete bookshelf. Conversely, source packages are the raw code and have to be built after you download them. Your computer does this, like buying a bookshelf from Ikea where you get the parts but spend some time putting it together. In this unit, we do not need the very latest versions so getting the binary version is fine and generally easier/faster. So for now, if given the option to install from source, you can type “no”. That said, the reason you have to install so many other tools (like clang) is so that your computer has all the tools need to build packages from source, just in case. The goal is to get your machine setup as much as possible so that during the semester you run into fewer issues installing packages. Here is an example of what this might look like while installing packages:

install from source screenshot

in that screenshot example, you would type at the console “no” (without the quotes) and press enter/return at which point R should install the binary version.


## some general language features, used in other packages
install.packages("rlang", dependencies = TRUE)

## for working with R and markdown files
install.packages("rmarkdown", dependencies = TRUE)

## general data management
install.packages("data.table", dependencies = TRUE) 

## data operations
install.packages("extraoperators", dependencies = TRUE) 

## read text files into R
install.packages("readr", dependencies = TRUE) 

## read Excel files into R
install.packages("readxl", dependencies = TRUE)

## read SPSS, STATA, and SAS files into R
install.packages("haven", dependencies = TRUE) 

## reshape data (for longitudinal datasets)
install.packages("reshape2", dependencies = TRUE) 

## beautiful graphs in R
install.packages("ggplot2", dependencies = TRUE) 

## create panels of plots in R
install.packages("ggpubr", dependencies = TRUE) 

## visreg app helps visualize regression models easily
install.packages("visreg", dependencies = TRUE)

## diagnostics and effect sizes for GLMs
install.packages("JWileymisc", dependencies = TRUE) 

## mixed effects models
install.packages("lme4", dependencies = TRUE) 

## mixed effects models diagnostics and effect sizes
install.packages("multilevelTools", dependencies = TRUE)

## multiple imputation
install.packages("mice", dependencies = TRUE) 

## load the packages (basically open/run the apps)
library(data.table)
library(extraoperators)
library(readr)
library(readxl)
library(haven)
library(reshape2)
library(ggplot2)
library(ggpubr)
library(visreg)
library(JWileymisc)
library(lme4)
library(multilevelTools)
library(mice)

The package installation may generate a lot of output. Finally loading the packages (the library() code) will not do anything, although you may get some notes or messages. If you do not receive any errors or messages like “Error in library(readr) : there is no package called ‘readr’” That is considered a successful outcome of this step.

Note that generally any text entered at the console is assumed to be a command to R. The exception is that text following a hashtag, #, is treated as a comment, not a command. This is a helpful way to document code, so you know what the purpose of a particular piece of code is.

6. Summary and Checklist

This module covers installing and setting up R and RStudio. You have a functioning version of R on your own computer. Before the first lecture, make sure you’ve done everything by going through this checklist.