Where’d we leave off …

In the previous lab, we started to get acquainted with the layout of RStudio and some of the commands.
In this lab, we’ll learn about different types of variables.
- Such as those that are measured by numbers and others that have values that are categories.
We’ll also look at ways to visualize these different types of data using plots (a word data scientists use interchangeably with the word graph).
Find the History tab in RStudio and click on it. Figure out how to use the information to reload the cdc data.

Variable Types

Numerical variables have values that are measured in units.
Categorical variables have values that describe or categorize our observations.
View your cdc data and find the columns for height and sex (use the History pane again if you need help to View your data).
- Is height a numerical or categorical variable? Why?
- Is sex a numerical or categorical variable? Why?
- List either the different categories or what you think the measured units are for height and sex.

Run the code you used in the previous lab to display the names of your cdc data’s variables (use the code displayed in the History pane to resubmit previously typed commands). Use the code’s output to help you complete the following:
- Write down 3 variables that you think are categorical variables and why.
- Write down 3 variables that you think are numerical variables and why.

One way to get a good summary of your data is to look at the data’s structure.
- One way to view this info would be to click on the little blue arrow next to cdc in the Environment pane.
- Another way would be to run the following in the console:

str(cdc)

Look at the structure of your cdc data and answer:
What information does the str function output?
Were you able to correctly guess which variables were categorical and numeric? Which ones did you mislabel?

Visualizing data is a really helpful way to learn about our variables.
Choose one numeric and one categorical variable from the data and create both a bargraph and a histogram for each variable.
- Which function, either bargraph or histogram, is better at visualizing categorical variables? Which is better at visualizing numerical variables?

Make a graph that shows the distribution of people’s weight.
- Describe the distribution of weight. Make sure to describe the shape, center and spread of the distribution.
Options can be added to plotting functions to change their appearance. The code below includes the nint option which controls the number of intervals in a numerical plot.
- Options, also known as arguments, are additional pieces of information you provide to a function, and are separated by commas.
- Type the command below on your console and then answer the questions that follow:

histogram(~weight, data = cdc, nint = 3)

- How did including the option nint = 3 change the histogram?
- Does setting nint = 3 impact how you would describe the shape, center and spread?
- Try other values for nint. What value produced the best graph? Why?

Make a graph that shows how often people in our data texted while driving.
- What does the y-axis represent?
- What does the x-axis tell us?
- Would you say that most people never texted while driving? What does the word most mean?
- Approximately what percent of the people texted while driving for 20 or more days? (Hint: There’s 17232 students in our data.)

Fill in the blanks with the correct variables to create a side-by-side bargraph:

bargraph(~ ___, data = ___, groups = ___)

Write a sentence explaining how boys and girls differ when it comes to texting while driving.
Would you say that most girls never text and drive? Would you say that most boys never text and drive?
How did including the groups argument in your code change the graph?

To answer this, what we’d like to do is visualize the distributions of heights, separately, for males and females.
- This way, we can easily compare them.
Use the groups argument to create a histogram for the height of males and females.
- Can you use this graphic to answer the question at the top of the slide? Why or why not?
- Is grouping numeric values, such as heights, as helpful as grouping categorical variables, such as texting & driving?

Why does this work for bargraphs but not for histograms?
- The groups argument uses color to differentiate between groups.
- With bargraphs, each group is split with bars next to each other on the x-axis.
- With histograms, the x-axis is a continuous set of numbers so the bars overlap making it difficult to compare center and spread.
Fill in the blanks with the correct variables to create a split histogram to answer the questions below:

histogram(~ ____ | ____ , data = ____ )

Do you think males and females have similar heights? Use the plot you create to justify your answer.
Just like we did for the histogram, is it possible to create a split bargraph? Try to create a bargraph of drive_text that’s split by sex to find out.

In this lab, we looked at the texting & driving habits of boys and girls.
What other factors do you think might affect how often people text and drive?
- Choose one variable from the cdc data, make a graph, and use the graph to describe how drive_text use differs with this variable.