In your final paper, you will conduct and interpret analyses with R on a dataset of your choice.
The goal of this assignment is to reinforce the main topics in the course, and to give you one document that you can look back on to help you in your future R adventures. This is not a test – it is an opportunity for you to continue learning R.
You need to analyze one new dataset that we have not already used in class. Ideally, your dataset should have at least 100 rows and 5 columns, where at least two columns are numeric and continuous (like height and age), and two columns are categorical (like sex and favorite movie). If you have a dataset you really want to analyze that does not satisfy these criteria, come talk to me and we’ll find a solution.
You can obtain a dataset from one of the following three methods:
Method 1: Use your own dataset
If you have data for a Bachelor’s or Master’s thesis, you are very welcome to use that. I don’t care if you have analyzed it in the past.
Method 2: Create a randomly generated dataset
If you want to analyse a randomly generated dataset from a fictional psychology study looking at the effects of priming on decision making, look at the following tutorial for instructions: Tutorial for creating your own dataset
Method 3: Find a real dataset online
If you would like to analyze a real dataset, here are a few places to get one. Note: I recommend you look for data in .csv or .txt format as these will be much easier to read than .xls or .sav files.
Journal of Judgment and Decision Making: http://journal.sjdm.org/. The Journal of Judgment and Decision Making requires all recent authors to publish their data. If you look at issues after 2014, you should find raw data for most if not all published articles.
Psychological Science: http://www.psychologicalscience.org/publications/psychological_science. Many articles published in Psychological Science in the past 2 years have open data.
Harvard dataverse: https://dataverse.harvard.edu/dataverse/harvard. Harvard University publishes thousands of open datasets. You can search for one related to psychology.
ClinicalTrials.gov: https://clinicaltrials.gov/ct2/home. ClinicalTrials.gov has hundreds of clinical datasets.
Financial datasets: http://www.r-bloggers.com/financial-data-accessible-from-r-part-iii/. Several financial datasets are available here.
Do not spend too long trying to find a dataset! Many students in past years have spent most of their final class periods just finding data, rather than using that time to write their papers and get help. This is not the best use of your time. Instead, spend an hour or two outside of class looking for a dataset.
You need to submit three separate documents in your final project
| Documents | Description |
|---|---|
Final_LastFirst.pdf (or .doc) |
A text document without code (e.g.; a Word document, or PDF output from RMarkdown or papaja) containing the three sections Introduction, Analyses and Conclusion. Your text document does not need to be formatted in any particular style. However, you must include your name, date, and a paper title at the top of your document! I recommend trying to replicate the overall look of APA style (e.g.; double spaced, with a title page), but you don’t need to. You don’t need to cite any papers or include a bibliography (but you can if you’d like to). |
Final_LastFirst.R (or .Rmd) |
A code document containing all the R code relevant to the tasks. Your code document should contain all of the code that produced the output in your text document, from the statistical analyses to the plots. Ideally, your code should be entirely self-contained and replicable. What does this mean? This means that when I receive your code, I should be able to run all of your code without any errors. |
*.txt |
Any data files (e.g. mydata.txt, survey.csv) referred to in your Final_LastFirst.R script. |
Final_LastFirst.pdf document?You can create your final text document in one of three ways: (standard) R Markdown, APA format with papaja, or writing it directly in Word (or your other favorite word processor).
There should be three sections in your Final_LastFirst.pdf (or .doc) text document: Introduction, Analyses, and Conclusion. Please write the document in paragraph style (e.g.; not just a list of bullet points).
Title
Include a title, your name, and the date at the top of the document.
Introduction (1-2 paragraphs)
Your paper should start with an introduction where you briefly describe your dataset. Where did you get it? For what purpose was the data collected? How was it collected? Be as descriptive as you can, but if you don’t know exactly how the data were collected (perhaps because you got it online) that’s ok.
In addition to describing the dataset, briefly describe the 1-5 main research questions you want to answer with these data. You can frame them as explicit hypotheses, or just exploratory questions.
Analyses (however long it takes)
This is the main section of your document. In this section, you should report the results of all the programming tasks (see below) that are relevant to your research questions.
Label your tasks with [TX]
Any time that you are reporting results from a specific task in your paper, include the [TX] tag, where X is the task number, so I (and you) can quickly find them in your paper. For example, here’s how a paragraph in your analysis section could look when you complete tasks 16 (a t-test) and 14 (a plot).
In order to test if there was a significant difference in the weights of chicks given diets 1 and 2, I conducted a two-sample t-test. The test showed that, indeed, chicks on diet 1 had significantly higher weights than those on diet 2, t(201.38) = -2.64, p < .01 [T16]. A barplot showing the distribution of weights for different diets is shown in Figure 1 [T14]. Next, I conducted an ANOVA to test if there were differences in weights in the different time periods. An ANOVA was significant F(1, 576) = 1349, p < 0.01 [T21]. To see which groups differed, I conducted post-hoc tests…
You only need to report tasks in your paper that provide real interpretable results. Some tasks with the Outside and R Only labels ask you to do some programming or organisation that you don’t need to include in your final text document. For example, task 1 asks you to open a new R project, obviously you don’t need to report this in your paper.
Conclusion (1-2 paragraphs)
Write a brief summary of your main conclusions in 1-3 paragraphs. You don’t need to go nuts here. Just write a few paragraphs that summarize the main conclusions from your results.
Final_LastFirst.R document?Your code document Final_LastFirst.R should be well formatted, contain your name and the date at the top, and include all of the R code necessary to produce your results. When you complete a task, you should include the [TX] tag so you and I can quickly see when you complete each task. Use comments to explain your code!
Important! Your code document should be complete, and without errors or bugs. You should be able to run all of your code without errors in a clean working directory. To test this yourself, I recommend clearing your workspace by running rm(list = ls()), and then running all of your code. It should run without errors!
Here is an example of how your Final_LastFirst.R document could look:
# Final_Phillips.Nathaniel.R
# Nathaniel Phillips
# 15 November, 2017
# Introduction to Statistics with R, Final Paper
# Set up libraries
library(tidyverse)
library(yarrr)
# [T1]
# Set up project
# [T2]
# Created Final_PhillipsNathaniel.R file
# [T3]
# Load mydata.txt as a new dataframe data
data <- read.table(___)
# ----------------
# More tasks....
# ----------------
# [T15]
# Calculate mean X for each level of Y
ChickWeight %>%
group_by(__)
%>% summarise(
__
__
)
# [T16]
# T-test comparing the weights of chickens on diet 1 versus 2
t.test(weight ~ Diet,
data = subset(ChickWeight, Diet %in% c(1, 2)))
# Result: t(201.38) = -2.64, p < 0.01You need to complete each of the following tasks. Tasks with the label Outside don’t require any code, you just need to make sure to do them. Tasks with the label R Only are programming tasks that you need to include in your .R script, but that you do not need to report in your text document. All other tasks should appear both in your .R script and in your final document.
Outside: Open a new project called R_FinalPaper.Rproj. In the main directory of the project, include two folders: R and data.
Outside: Create a new R script called Final_LastFirst.R, where Last and First are your last and first names. Save the file in your R folder. Add your name and the date as comments at the top of the script.
Outside: Put the raw data for your project (probably a .txt or .csv file) in the data folder of your project.
R Only: Load all of the packages you need for your analyses.
R Only: Read data into R from your text file using read.table() (or a similar function).
Using summary(), str(), nrow() and ncol(), Report summary statistics from your data generated by applying the summary() or str() function to your data. How many rows and columns are in your data? What are the names of the columns?
R Only: Change the name of at least one column in your data.
Calculate at least one standard deviation with sd(), one mean with mean(), and one median with median().
Print a table of the frequencies of outcomes of a categorical variable with table().
R Only: Recode the values of at least one column using indexing and assignment (or the case_when() function in dplyr). For example, in a column of sex data, you could change “female” to 1, and “male” to 0.
Count the number of outliers in one numerical vector. Define an outlier as any datapoint that is more than 3 standard deviations away from the mean.
Note: In all plots, make sure to include appropriate axis labels and plot titles
Create at least one plot showing the relationship between a numeric independent variable and a numeric dependent variable (maybe a scatterplot?). Include a regression line. If you’d like, show different plots for different groups of data (e.g.; different colored points for different groups).
Create at least plot showing the distribution of a numeric variable. Maybe a histogram?
Create at least one plot showing the relationship between one or more categorical independent variables (e.g.; sex, eye color) and one numeric dependent variable (e.g.; height). Consider including error bars representing confidence intervals.
Use aggregate() or dplyr to calculate descriptive statistics across groups of data. For example, you could calculate the mean of a numeric variable for each level of a categorical variable.
Calculate at least one t-test with t.test(). Write your results in APA format.
Calculate at least one t-test test on a subset of data by combining subset() with t.test(). For example, you could conduct a t-test comparing results from a performance test between two experimental groups, but only for females. Write your results in APA format.
Calculate at least one correlation test with cor.test(). Write your results in APA format.
Calculate at least one correlation test on a subset of data by combining subset() with cor.test(). For example, you could conduct a correlation between height and weight but only for males. Write your results in APA format.
Calculate at least one chi-square test with chisq.test(). Write your results in APA format.
Calculate at least one one-way ANOVA analysis with aov() (or another ANOVA function such as Anova()) with 1 independent variable. Calculate post-hoc tests for any significant effects. Write your results in APA format. If there are more than 4 paired comparisons in your analysis, you don’t need to describe every paired difference, just explain the most important ones.
Calculate at least one multiple-variable ANOVA analysis with 2 or more independent variables. Calculate post-hoc tests for any significant effects. Write your results in APA format. If there are more than than 4 paired comparisons in your analysis, you don’t need to describe every paired difference, just explain the most important ones.
Repeat your ANOVA from the previous task (22), but now include interaction terms. Write your results in APA format. Do you find an interaction?
Calculate at least one regression analysis with lm() or glm(). Write your results in APA format.
Report summary statistics of the residuals from your regression analysis in task 22. That is, on average, how far away are the model fits from the true values? What is the standard deviation of these residuals?
Here are my answers to frequently asked questions
| Question | Answer |
|---|---|
| Can I work with someone else on the same dataset? | Yes! You are welcome to use the same dataset as another student in the class, and you are welcome to work together on your analyses. However, you must each write your own code and text and turn in your own work! If I suspect that you did not write your paper / code, I reserve the right to ask you to explain it to me personally. If you cannot convince me that your code is your own, you will not pass the class. |
| Do I have to refer to all tasks in the final text document? | No. Some tasks (like task 1) are for setting up your analyses, or manipulating your data in R. You don’t need to mention these in your text document. However, any R code should be in your R document. |
| Do I have to complete the tasks in order? | No. If it makes more sense for you to do the tasks in a different order, that’s fine. But do make sure to label them with the [TX] tag. |
| How do I submit my final assignemnt? | Submit your paper like a regular WPA by completing the WPA submission page, and emailing me all your documents. |
| Does my document need to be in APA style? | No. You can use any style you want. However, I would prefer that you write your document in paragraph style, as it would generally look in a real manuscript, rather than simply a list of all the tasks. |
| Can I turn in the assignment before the deadline? | Yes, of course. In the past, many students have even finished their final assignment in the final class periods. |
| What do I do if I can’t turn in the paper before the deadline? | You have lots of time to finish your assignment, so this should not be an issue. However if you tell me well before the deadline that you won’t be able to turn it in on time, let me know your reason and we may be able to arrange something. If you tell me shortly before the deadline, it is unlikely that I will give you more time (unless of course there was a serious emergency). |
| What if some of these tasks don’t make sense for my data? Do I still have to do them or can I skip them? | If some of these tasks don’t make sense for your data, explain why, then replace the task with another one at the appropriate level of difficulty that displays your R knowledge. |
| What if I don’t know how to complete a task, or I’m not sure if I did it correctly? | Just ask for help! The final project is not a test of your knowlege, it’s an opportunity for you to continue learning. Ask your colleagues, or come to my office and ask for help directly. |
| The paper is due tomorrow and I have an error! Can you help me? | No. I won’t be able to help you the day before the assignment is due. |
| Do I have to complete all tasks perfectly to get a good grade? | No. If you can convince me that you worked hard on your project, and asked for help when you needed it, you’ll be fine, even if you did not successfully complete all tasks. |
| I want to include some additional analyses in my document that aren’t listed as one of the tasks. Can I include them? | Of course, yes! |
I will grade your paper based on how well you followed the instructions above. Here is a checklist I will use when grading your paper.
If the answer to all these questions is “Yes” (or “Mostly Yes”), you will be just fine.