Lab 2: Data Wrangling Beginnings

Instructions

We have made it to Lab #2! We are going to keep practicing the skills we started using in the last week, except with using some new data. Before, we were using pre-installed data, but that won’t be the case IRL. When working with your own data, you will want to create a workflow of cleaning and wrangling your data in a reproducible way. These steps will likely occur for every new dataset that you work with.

Please complete the exercises below. Create a new .Rmd file and include the following at the top:

---
title: "Lab 2: Data Wrangling Beginnings"
author: "Your Name Here"
date: "`r Sys.Date()`"
output: html_document
editor_options: 
  chunk_output_type: console
  markdown: 
    wrap: 72
---

You should then be able to copy/paste everything below into your document.

Here is a .Rmd file that you should be able to download and use as well.

Download Lab 2 (.Rmd)

When you are finished, click the Knit button to turn your work into an HTML document. You will submit both this .Rmd file and the 🧶knitted .html file.

Scenario and Goal

Congratulations, you’ve just collected data for a study on “Personality”! You administered a 10-item personality questionnaire, where participants responded on a 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree). This measure is called the “Ten Item Personality Inventory” (TIPI) and more information about the measure can be found here: TIPI Scale Info. Please refer to this page and documents to help with scoring the data and getting familiar with the measure.

However, the raw data from the survey software is messy. Your goal in this lab is to import, clean, and score the data to prepare it for analysis. This process of turning raw data into usable data is called data wrangling, and it’s what researchers spend most of their time doing.

You will learn to: * Import a CSV file. * Rename variables for clarity. * Filter out participants based on data quality checks. * Reverse-score negatively worded items. * Compute a composite score for a psychological scale.

Getting Data

Be aware of your file structure and how things are organized. For some refreshers, take a look at the Resources and rstats.wtf

Download your data

Download Data from Drive (CSV)

Download Questionnaire From Drive (DOC)

Exercise 1: Importing and Inspecting the Data

First, you need to load the appropriate packages and import your data. Use the tidyverse, rio, and here libraries.

# Load the appropriate libraries.   

# Write your import code here:   
  ## This will look like:     
  # your_data_name <- import(here("path", "to", "file", "filename.csv))    


# Use glimpse() to get a first look at the raw_data. 
# Write your code here:

Question 1: How many participants (rows) are in the raw, imported dataset?

❓Your Answer: [Type your answer here]

Exercise 2: Renaming Variables for Clarity

The column names are a mixture of naming conventions. Let’s rename them to be consistent and convey the appropriate information.

First, let’s take a look at what the names are. You can do this by using View(), but let’s use the names() function to list out all the column names.

names(Whatever_you_Named_your_data)

This will give you your list of names. You can see which ones we may want to rename. What does Q85 even mean? Thanks Qualtrics. Review the documentation for the survey to get a sense of what the questions are asking to properly rename the variables.

For now, let’s update the names as follows. It is helpful to keep everything lowercase to make it easier to type (but this is a personal preference), and make sure there aren’t any spaces in your variable names:

ID -> id
Progress -> progress
Duration (in seconds) -> duration
Consent -> consent
Q85 -> sex_orient
Q85_6_TEXT -> sex_orient_txt
Sleep Quality -> sleep_qual
Hours of Sleep -> sleep_hours

# Task: Use the rename() function to change the variable names as listed above.
# Create a new object called `renamed_data`.
# Hint: The syntax is: new_object <- old_object %>% rename(new_name_1 = old_name_1, new_name_2 = old_name_2)

# Write your code here:
# This part is a placeholder as the actual column names from the URL will differ.
# Be sure to update the following template with the info from your own data
renamed_data <- raw_data %>%
  rename(participant_id = country, 
         grit_1 = year,
         grit_2 = population)
         # ... and so on for the other variables

# Print the first few rows of your new `renamed_data` object to check your work.
head(renamed_data)

Exercise 3: Filtering for Data Quality

Our survey includes variables that allow us to see how long they took and what percentage they completed. We should remove participants who did not finish the survey, as well as those who finished it too quickly.

This will be done using the filter() function (more about filter). Use filter to keep participants where their progress is equal to 100. You will also want to remove participants who completed the survey in less than 7 minutes (note: the Duration variable is in seconds).

# Create a new object called `filtered_data`.

# Example code:
# filtered_data <- renamed_data %>%
#   filter(Progress == 100) %>% 
#   filter(Duration ==, <, > 1000)

Question 2: How many participants remain from your original dataset? How many participants did you remove with your filters?

❓Your Answer: [Type your answer here]

Exercise 4: Reverse-Scoring Items

In a lot of psychological research, we need to reverse score variables. They are often worded in a negative/positive way compared to the rest of the items. Review the TIPI documentation to see what items need to be reverse scored.

For example, Item 2 is a item that should reflect “Agreeableness”, but the rated words are “Critical, quarrelsome”. Therefore low score on this item would reflect high amounts of Agreeableness.

The formula for reverse scoring an item is: (Maximum Possible Value + 1) - Original Score. So, for the TIPI, it’s 8 - TIPI_2.

To create/compute a new variable we use the mutate() function (more about mutate). I like to remember this function name because we are “mutating” the data and introducing another “growth” or something extra that wasn’t there before.

# Task: Use the mutate() function to create new, reverse-scored variables.
# Use a new name to indicate which items are reverse-scored.
# Create a new object called `scored_data`.
# Hint: The syntax is: 
  # new_object <- old_object %>% 
  #   mutate(new_variable = computation, 
  #          new_variable2 = computation)


# Exmple code:
# scored_data <- filtered_data %>%
#   mutate(TIPI_2r = 8 - TIPI_2)



# Print the first few rows, showing only the original and new reverse-scored item to check your work.

# scored_data %>% select(TIPI_2, TIPI_2r) %>% head()

Question 3: If a participant’s original score on TIPI_4 was a 6, what would their score be on the new TIPI_4r variable?

❓Your Answer: [Type your answer here]

Exercise 5: Computing and Finalizing the Scored File

Now we are ready to compute the final score! There are individual subscales for each of the 5 factors of the Big 5 Personality Inex. Compute the 5 scales. Remember to use the reverse-scored items (TIPI_2r), not the original one.s

# Task 1: Use mutate() to calculate the total grit score.
# Sum the items corresponding to each scale.
# Call the new variables 'extra', 'agree', 'consc', 'emo', 'open.
# Overwrite your `scored_data` object with this new version.

# Task 2: Create a final, clean dataset.
# Use select() to keep only the 'id' and Big 5 scale columns.
# Call this object `final_data`.
# Conceptual code:
# final_data <- scored_data %>%
#   select(id, agree, ...)



# Task 3: Calculate the mean and standard deviation of each of the subscale scores.

Question 4: What would be the highest possible Emotional Stability score a participant could get? What would be the lowest?

❓Your Answer: [Type your answer here]

Question 5: Report the mean’s and standard deviations of each of the subscale scores

❓Your Answer:

Extraversion:
Agreeableness:
Conscientiousness:
Emotional Stability:
Openness to Experience:

Exercise 6: Visualize Relationships

We often want to see the relationship between 2 variables. This is often done using a scatterplot. Select 2 scales that you would like to see the relationship between. Use the code from the previous lab/class to create a scatterplot of the relationship. Take a look at this cheat sheet to help with ggplot2!

# Using ggplot and the scored data you have, generate a scatterplot. It can be as simple or as fancy as you would like (try putting a title and changing the axis names)



# Use the cheat sheet and other materials to put a straight line to the data (Hint: add another layer with a smooth geom)

Question 6.1: Visually inspect the chart you have and describe the relationship below. Be sure to include the two variables and an estimate of your correlation in your answer.

❓Your Answer:

# Compute a correltion for your variables
# Hint: use the `cor.test()` function. You will need to specify the variable names as `name_of_data$name_of_variable`
  
  # Example
    # cor.test(starwars$height, starwars$mass)

Question 6.2: Calculate the correlation coefficient for your two variables of interest (refer here for more info). Report the correlation, and reflect on how close or how far away your initial estimate was. Do your best here. I recognize that we haven’t really gone over this just yet.

❓Your Answer:

End of Lab. Don’t forget to Knit! 🧶