Skip to content

09. Visualization

Module items

R Script file code

  • [[Copy the code]] below ➜ Paste into [[RStudio console]] ➜ Hit Enter 

    • source(url("https://raw.githubusercontent.com/ttezcann/ssric-reg/refs/heads/main/docs/assets/attachments/data/0-packages-data.R")); 
(function(f="09-visualization.R"){if(!file.exists(f)){download.file("https://raw.githubusercontent.com/ttezcann/ssric-reg/refs/heads/main/docs/assets/attachments/modules/09.-visualization/09-visualization.R",f,mode="wb");file.edit(f)}else{download.file("https://raw.githubusercontent.com/ttezcann/ssric-reg/refs/heads/main/docs/assets/attachments/modules/09.-visualization/09-visualization.R",gsub(".R","-original.R",f),mode="wb");file.edit(gsub(".R","-original.R",f))}})()
      • When this R script file opens in a new tab, [[Save R script file|save your previous R script file(s)]], and
        • Close the previous tabs (R Script files), which you can find later in the [[Files tab]].

Lab assignment

Visualization

Sample lab assignment

Sample: Visualization

Suggested reading


  • Healy, Kieran Joseph. 2019. “Look at the Data.” Pp. 1–23 in Data visualization: A practical introduction. Princeton: Princeton University Press.

Learning outcomes

  1. Learn how to generate and interpret a stacked bar graph for multiple categorical variables
  2. Learn how to generate and interpret a grouped bar graph
  3. Learn how to customize graph appearance by modifying color themes, titles, and font sizes

[[Stacked bar graph for multiple variables]]

  • A stacked bar graph for multiple variables displays:
    • Multiple [[categorical]] variables with the exact same response categories at the same time.
      • Each row represents one variable and shows the percentage breakdown across response categories.
      • It is useful when you want to compare distributions across several related variables.
      • When interpreting stacked bar graphs, we generally interpret one response category.
  • We will create a stacked bar graph for confidence in major US institutions variables, then interpret it.

Find the variables in Variables in GSS page

  1. We want to make sure all selected variables are categorical.

  2. [[Search]] the variable names, conbus, coneduc, confed, conmedic, conarmy, and conjudge in Variables in GSS page.

    • Variable name Variable label Variable type Question wording and response categories
      conbus Confidence level in major companies Ordinal Would you say you have confidence in major companies?
      (1: A great deal; 2: Only some; 3: Hardly any)
      coneduc Confidence level in education Ordinal Would you say you have confidence in education?
      (1: A great deal; 2: Only some; 3: Hardly any)
      confed Confidence level in executive branch of fed. govt. Ordinal Would you say you have confidence in executive branch of the federal government?
      (1: A great deal; 2: Only some; 3: Hardly any)
      conmedic Confidence level in medicine Ordinal Would you say you have confidence in medicine?
      (1: A great deal; 2: Only some; 3: Hardly any)
      conarmy Confidence level in military Ordinal Would you say you have confidence in military?
      (1: A great deal; 2: Only some; 3: Hardly any)
      conjudge Confidence level in United States Supreme Court Ordinal Would you say you have confidence in Supreme Court?
      (1: A great deal; 2: Only some; 3: Hardly any)

[[Stacked bar graph for multiple variables]] #code

  • [[Model code]]

    • 1
2
3
4
5
6
7
8
9
      graph <- gss |> 
select(variable1_here, variable2_here, variable3_here) |>
plot_stackfrq(sort.frq = "first.asc",  geom.colors = "Blues",
show.total = FALSE, title = "title_here")
graph + theme(
axis.text.x = element_text(size=11),
axis.text.y = element_text(size=11),
plot.title = element_text(size=12),
legend.text = element_text(size=11))

  • [[Working code]]

    • 1
2
3
4
5
6
7
8
9
      graph <- gss |> 
select(conbus, coneduc, confed, conmedic, conarmy, conjudge) |> 
plot_stackfrq(sort.frq = "first.asc",  geom.colors = "Blues",
show.total = FALSE, title = "Confidence in major US institutions")
graph + theme(
axis.text.x = element_text(size=11),
axis.text.y = element_text(size=11),
plot.title = element_text(size=12),
legend.text = element_text(size=11))
      • Line 2: We put the variable names here ➜ variable1_here, variable2_here, variable3_here, separated by a comma.
        • conbus, coneduc, confed, conmedic, conarmy, conjudge.
      • Line 4: We put the graph title here ➜ title_here ➜ Confidence in major US institutions.
      • Line 6: Change font size of x-axis labels ➜ size=11
      • Line 7: Change font size of y-axis labels ➜ size=11
      • Line 8: Change font size of graph title ➜ size=12
      • Line 9: Change font size of legend ➜ size=11
        • [[Find this working code in the R script file]].
          • [[Highlighting and running]] this code will generate the output below (which will appear in the [[plots tab]] of RStudio).

[[Stacked bar graph for multiple variables]] #output

  • A stacked bar graph showing confidence in major U.S. institutions, including the military, medicine, education, the Supreme Court, major companies, and the executive branch of government. Each bar is divided into response categories representing different levels of confidence.

[[Stacked bar graph for multiple variables]] #interpretation

  • Stacked bar graph for multiple variables interpretation sample

    Of the GSS respondents, 40.8% have a great deal of confidence in the military; 27.2% have a great deal of confidence in medicine; 19.6% in education; 16.4% in the supreme court; 14.4% in major companies; and 12.5% in the executive branch of government.

  • Stacked bar graph for multiple variables interpretation template

    Of the GSS respondents, xx.xx% have / report / say [[value label]] in [[variable label]] 1; xx.xx% have / report / say value label in variable label 2; xx.xx% have / report / say value label in variable label 3...

  • Interpretation explanation

    • When interpreting stacked bar graphs, we pick one response category, use the percentage shown on the graph, and compare it across all variables.
      • In this example, we interpret the "a great deal" category (light blue).
      • Then, we report the percentage for the same response category in each variable by tweaking wording to make it understandable. That's why we always read the [[question wording]] first.
        • For example: "40.8% have a great deal of confidence in the military..."

[[Grouped bar graph]]

  • We can create a grouped bar graph to:
    • Compare a [[categorical]] variable's distribution across groups.
  • We will create a stacked bar graph for confidence in medicine by perceived health groups.

    • Since we use two variables, we want to see the connection between the two variables. Therefore, we need to decide which one is the [[factor variable]], and which one is the [[outcome variable]].

      • flowchart LR
    subgraph F["Factor variable (Categorical)"]
        A[Perceived personal health quality<br/>1: Excellent;<br/> 2: Very good;<br/> 3: Good; <br/> 4: Fair; <br/> 5: Poor]
    end

    subgraph O["Outcome variable (Categorical)"]
        B[Confidence level in medicine<br/>1: A great deal;<br/> 2: Only some;<br/> 3: Hardly any]
    end

    A ==>|May have an effect on| B

Find the variables in Variables in GSS page

  1. We want to make sure all selected variables are categorical.

  2. We check this information in the Variables in GSS page.

    • Variable name Variable label Variable type Question wording and response categories
      health Perceived personal health quality Ordinal Would you say that in general your health is Excellent, Very good, Good, Fair, or Poor?
      (1: Excellent; 2: Very Good; 3: Good; 4: Fair; 5: Poor)
      conmedic Confidence level in medicine Ordinal Would you say you have confidence in medicine?
      (1: A great deal; 2: Only some; 3: Hardly any)

[[Grouped bar graph]] #code

  • [[Model code]]

    • 1
2
3
4
5
6
7
8
9
      graph <- plot_xtab(
gss$factor_here, gss$outcome_here, show.n = FALSE, margin = "row",
geom.colors = "Dark2",
show.total = FALSE, title = "title_here")
graph + theme(
axis.text.x = element_text(size = 11),
axis.text.y = element_text(size = 11),
plot.title = element_text(size = 12),
legend.text = element_text(size = 11))

  • [[Working code]]

    • 1
2
3
4
5
6
7
8
9
      graph <- plot_xtab(
gss$health, gss$conmedic, show.n = FALSE, margin = "row",
geom.colors = "Dark2",
show.total = FALSE, title = "Confidence level in medicine by perceived personal health quality")
graph + theme(
axis.text.x = element_text(size = 11),
axis.text.y = element_text(size = 11),
plot.title = element_text(size = 12),
legend.text = element_text(size = 11))
      • Line 2: We put health here ➜ factor_here, and conmedic here ➜ outcome_here.
      • Line 4: We put the graph title here ➜ title_here ➜ Confidence level in medicine by perceived personal health quality.
      • Line 6: Change font size of x-axis labels ➜ size=11
      • Line 7: Change font size of y-axis labels ➜ size=11
      • Line 8: Change font size of graph title ➜ size=12
      • Line 9: Change font size of legend ➜ size=11
        • Now, [[find this working code in the R script file]], and
          • [[Highlighting and running|Highlight and run]] the working code in this module's R script file to generate the output shown below, which will appear in the [[viewer tab]] of RStudio.

[[Grouped bar graph]] #output

  • Bar graph of confidence in medicine by perceived health. In every health group, “Only some” is the most common response (44.0–56.4%); “A great deal” falls from 32.1% (excellent) to 22.4% (fair), and “Hardly any” is highest in fair health (33.6%).

[[Grouped bar graph]] #interpretation

  • Grouped bar graph interpretation sample

    Of the GSS respondents who perceive their health as excellent, 20.3% have hardly any confidence in medicine, 47.6% have only some confidence, and 32.1% have a great deal of confidence.

  • Grouped bar graph interpretation template

    Of the GSS respondents who are / have / feel / perceive / think / say / report selected [[value label]] of the [[factor variable]], xx.x% are / have / feel / perceive / think / say / report value label 1 of the [[outcome variable]], xx.x% are / have / feel / perceive / think / say / report value label 2 of the outcome variable, and xx.x% are / have / feel / perceive / think / say / report value label 3 of the outcome variable.

  • Interpretation explanation

    • Each value label of the factor variable is a group, and the bars within each group add up to 100%.
    • When interpreting stacked bar graphs by groups, we pick one value label of the factor variable (one group of bars), use the percentages shown on the graph for that group, and report them across all outcome variable value labels.
      • In this example, we interpret the "excellent" group.
      • First, we name the selected factor variable group. Then, we report the percentage of that group in each value label of the outcome variable by tweaking wording to make it understandable. That's why we always read the [[question wording]] first.
        • For example: "Of the GSS respondents who perceive their health as excellent, 20.3% have hardly any confidence in medicine, 47.6% have only some confidence..."