Interactivity

Exercise 11 — PMAP 8551, Summer 2025

---

Task 1: Session check-in

Three interesting or exciting things:

1. It’s interesting how small changes to how we present data can have a big impact. Swapping static tables for visual elements—trendlines, target markers, and bullet graphs—resulted in significantly improved clarity and actionable insights, despite only being three tweaks.

2. Dashboards can be unintentionally misleading. By oversimplifying, single-page dashboards often strip away necessary context, such as missing data or changes in measurement, which can lead individuals to make decisions without fully understanding the context or caveats.

3. Integrating narrative elements into the dashboard is a great approach to provide helpful context, which is critical for users to guide their decision-making process based on the data presented. A good rule of thumb is to pair metrics with a brief explanatory text box to improve clarity and avoid any misleading interpretation of data, without overwhelming readers.

Three muddy or unclear things:

1. I still don’t fully understand the procedure in Quarto to create the dashboard. Everything seemed straightforward, but my plots are still appearing as usual and not in cards. I re-watched the videos, and it’s still unclear what I’m doing wrong.

2. It’s not clear whether the “dashboard report” is a static multipage PDF or a dynamic interactive dashboard. I’m unclear about whether narrative supports still apply in the interactive contexts and how much of it is necessary, especially for large datasets, which can be overwhelming for users.

3. I believe there are different design needs for a real-time dashboard compared to a static one. I’m unclear about those specific requirements, especially regarding the inclusion of narratives alongside the graphics.

Task 2: Interactive plots

library(tidyverse)
library(plotly)

interactivity <- read_csv ("data/unemployment.csv") 
interactivity_clean <- interactivity %>% 
  filter(year== 2010) %>% 
  group_by(region, state, na.rm = TRUE) %>% 
  summarize(total_unemployment = sum(unemployment)*100,1) %>% 
  ungroup() %>%
  mutate(
    percent_unemployment = round((total_unemployment / sum(total_unemployment)) * 100, 1)
  )

Do the following:

1. Make a plot. Any kind of plot will do (though it might be easiest to work with geom_point()).

static_plot <- ggplot(interactivity_clean, aes(x= percent_unemployment, y=region, color=state)) +
  geom_point()+
 theme_minimal()+
theme(legend.position = "NONE")+
labs(title = "Regional unemployment in 2010", x= "Unemploymet (%)", y= "Region")

static_plot

2. Make the plot interactive with ggplotly().

ggplotly(static_plot)

3. Make sure the hovering tooltip is more informative than the default.

Good luck and have fun!

Task 3:

Create a Quarto dashboard in a new .qmd file that shows a plot (static or interactive) in at least three chart areas. See the example for this session for resources on creating dashboards in Quarto.