Practice 2

What have you learned so far in R?

After completing the first practice, you should have learned these key concepts:

• R is an object-oriented programming language.
• R is free and open source, and you are able to install as many packages as you want for free.
• You learned the concept of “package”.
• You learned about data frames.

In this practice we will work with data frames , we will create new variables, and we will use extensively pipes along with the package tidyverse.

Good News!

In this practice you will be able to run R code straight from your web browser. Isn’t it awesome!

Let’s put a frame on those data….

The data frame is one type of object in R, and it is extremely useful to represent information as a data frame. This R object looks very similar to a spreadsheet in Excel, each column represents a variable, while each row represents an observation.

In this practice session we will wrangle the data set named ruminationToClean.csv. You may download the data set by Clicking Here. This is a dataset that was part of a large project that I conducted in Costa Rica in collaboration with a colleague and friend David Campos. Participants were high school students from the Central Valley in Costa Rica. The aim of this study was to explore the relationship between rumination and depression in teenagers. We added metacognitive beliefs to the study, but we will not talk about metacognition in this practice.

You may follow me and enroll my Cognitive Processes course to know more about metacognition

When I say wrangle I mean to manipulate the data set and clean the data. In real life, data sets are not perfectly clean. Many times we need to delete observations that are extreme (careless answers, mistakes, etc.), or sometimes we need to delete new variables. Additionally, we might need to compute new variables. We will cover all these topics in this practice using R.

The package tidyverse is your friend

If you run the following code, you’ll see the variable names in the data set ruminationToClean.csv:

I’m reading in the data file from my personal repository, fancy right?

As you can see, we have a large list of variables. We are using the function colnames() to get a list of column names, which in this case are research variables. Several of these variables could be dropped from the data set, others will need to be renamed.

Dropping variables

In this example we will use the function select() from the package tidyverse. Remember that you’ll need to call the package by running library(tidyverse).

What is happening here? You may have noticed that inside the function selection() I am adding the name of the variable I want to delete from the data set, and then I added a minus sign. When you add a minus sign before the name of the variable, the package will delete the variable that you list with a minus sign. You can do the same with several variables at the same time:

Now the variables, age, school, and grade are not longer in the dataset, if you don’t believe me click on RUN CODE.

Excercise 1 (30 points)

1. Download the data set ruminationToClean.csv then open the data file using ruminationRaw <- read.csv(file.choose()) or you may copy the code to open the data from the online repository. After opening the data in R, delete from the data set all the variables that start with th stem EATQ_R_.

Change variable names

In some occasions, the research variable was named wrongly or the name does not reflect the content of the variable. In the data set ruminationToClean.csv there are variables with names in Spanish, we can change them:

In this example, you can see that I’m using the function rename() from the package dplyr which is inside the package tidyverse. When you use rename() you’ll need to declare the new name first and then the old variable name that will be replaced.

Excercise 2 (35 points)

2. In psychological measurement it is common to implement questionnaires with Likert-response options. We typically sum the items corresponding to the latent variable or construct that we are measuring. For example in the ruminationToClean.csv the construct of “Rumination” is measured by the Children’s Response Styles Questionnaire (CRSQ), however the variable names are not self-explained. Your task is to change the name of the variables with stem CRSQ, the new variables should have the stem rumination_itemNumber. For example rename(CRSQ1=rumination_1). Remember to save the changes in a new object.

Compute a new variable

As always in programming languages, there are several ways to create a new research variable in a data frame. But, we will use the tidyverse method by using a function that actually works like a verb. The function mutate() will help use to create new variables.

I mentioned in Exercise 2 psychology studies latent variables also known as hypothetical constructs, they have many names, but at the end we study variables that are implicit; we can measure them but we don’t see them. For example, we have studied variables such depression or rumination in this course, these are latent variables because you cannot capture depression or you don’t see it walking, however you are able to measure depression by asking about symptoms and thoughts.

Example of a latent variable

In the figure above you can see a representation of a latent variable. In psychology we often assume that the latent variable is the common factor underlying our questions, -in this example- questions related to depression.

When we analyze data that corresponds with latent variables we have two options:

1. We can create a Structural Equation Modeling (We won’t cover this topic).

2. We sum all the items to compute a total score. This total score will be our approximation to the latent variable. (Yes, this is covered in this practice).

I’ll show an example in the next chunk of R code:

In this example we are using the function mutate() to create a new column named depressionTotal, this new depressionTotal column will contain the depression total score for each participant. In this study we included the “Child’s Depression Inventory (CDI)”, that’s why the columns follow the stem “CDI”. You can see the new column in the next table:

Code

ruminationRaw <- read.csv("ruminationToClean.csv", na.string = "99")
ruminationDepresionScore <- ruminationRaw |>
  mutate(
    depressionTotal = rowSums(across(starts_with("CDI")))
    )
ruminationDepresionScore |> select(starts_with("CDI"),  
                                   depressionTotal ) |>
  gt_preview()

	CDI1	CDI2	CDI3	CDI4	CDI5	CDI6	CDI7	CDI8	CDI9	CDI10	CDI11	CDI12	CDI13	CDI14	CDI15	CDI16	CDI17	CDI18	CDI19	CDI20	CDI21	CDI22	CDI23	CDI24	CDI25	CDI26	depressionTotal
1	0	2	0	0	0	0	0	1	0	0	1	0	1	1	1	1	1	1	1	0	0	1	1	1	1	0	15
2	0	0	1	1	0	0	0	1	0	0	1	0	2	0	1	0	2	1	1	0	0	1	2	1	0	0	15
3	0	1	0	0	0	1	0	0	1	0	0	0	0	0	0	0	0	0	1	0	0	0	0	0	0	0	4
4	0	1	0	0	0	0	0	0	0	0	0	1	1	0	0	0	1	1	0	0	0	0	0	1	0	0	6
5	0	1	0	1	0	1	0	0	1	1	0	2	1	1	0	1	1	2	0	1	0	0	1	1	1	0	17
6..211
212	0	0	0	0	0	0	0	0	0	1	0	0	0	1	0	0	0	1	0	0	0	1	0	1	0	0	5

Excercise 3 (30 points)

3. In the data set ruminationToClean.csv you’ll find seven columns with the stem “DASS”. These columns correspond to the instrument named: Depression, Anxiety, and Stress Scale (DASS). This is an scale that measures three latent variables at the same time, but only the anxiety items were added in this study.

Your task is to change the name of the “DASS” items, you should think a more informative name for these items in the data frame. After that, you will have to compute the total score of anxiety for each participant in the data frame.

Let’s cleanse this data set

It is always satisfactory to clean a data set and then see the final product, but so far we have cleaned these data in small steps. I’ll clean the data set in one step. In this example I will delete unnecessary columns (aka research variables), and I will translate the variables to English when necessary:

Excercise 4 (10 points)

4. Describe the steps taken in the R code above.

Examine your data

In this second part we will create several plots to examine and explore our data. But first we’ll need to compute total scores again, because after all, the main point is to make conclusions about our latent variables (a.k.a latent factors).

We have now three total scores corresponding to three different latent factors: depression, anxiety, and depression. We are in a good place to explore the data. It is healthy to start looking at histograms because they are now continuous variables:

We can see in the histogram that low values are more frequent. This is expected because all the participants are healthy teenagers without any history of psychopathology. Also, pay attention that we are using the function hist() in R.

We can also generate a density plot to have an idea of the observed shape:

Notice the function plot() which is a generic function in R it is used as a method in the environment. This functions comes with R as a base function. If you need to know how to use it you can run ?plot(), but the help section might not be helpful for new users. Additionally, pay attention on how you can change the labels in your plot.

Excercise 5 (15 points)

5. Create a density plot for the depression total score. Where is most of the density located?

Note

R Studio has an option to save plots. Just click on the tab named “Plots”, then click where it says “Export”. You may click on “Copy to clipboard”, this will allow you to paste the image on any software. You may also select to export the image as PDF file or image.

Option to save generated plots

Box plots are also useful to understand our data better, you may revisit the box plot’s anatomy in this website: link.

We can see the next example (Click on Run Code):

Excercise 6 (15 points)

6. Can you tell which group in the box plot has the highest median in rumination? What else can you see in the box plot?