Communicate

LAW Module 4: A Code Along

Welcome to the LAW Code Along for Module 4

• The Communicate phase of the LAW is about selecting, polishing, and presenting findings to an audience.

• In the Communicate phase of the Learning Analytics workflow, the focus is on effectively presenting and sharing the results of your analysis with stakeholders. This involves creating clear and compelling visualizations, writing comprehensive reports, and delivering presentations that highlight key insights and actionable recommendations.
• The goal is to ensure that the findings are easily understood and can inform decision-making processes, ultimately leading to improved educational outcomes. Utilizing tools like ggplot2 for visualizations and packages like RMarkdown for report generation can enhance the clarity and impact of your communication.

Module Objectives

By the end of this module, learners will:

• Develop an intuition for the Communication phase of the learning analytics workflow.

• Understand the utility and power of dashboards for data communication.

• Build a dashboard of their own, experimenting with its modular components (e.g., tabs, plots, sidebars, etc.).

Preparing Findings for Communication

Earlier in the LAW, such as in the Explore phase, you might have wondered how you could ever find a compelling or concrete finding from all of the data. In the Communicate phase, you might find yourself with the opposite problem: Which findings should I pick and share?

Krumm et al. (2018) have outlined the following 3-step process for communicating with education stakeholders what you have learned through analysis:

• Select: Communicating what one has learned involves selecting among those analyses that are most important and most useful to an intended audience, as well as selecting a form for displaying that information, such as a graph or table in static or interactive form, i.e.a “data product.”

• Polish: After creating initial versions of data products, research teams often spend time refining or polishing them, by adding or editing titles, labels, and notations and by working with colors and shapes to highlight key points.

• Communicate: Writing a narrative to accompany the data products involves, at a minimum, pairing a data product with its related research question, describing how best to interpret the data product, and explaining the ways in which the data product helps answer the research question.

Dashboards for Communication

• Overview
• Components
• YAML

• Dashboards are one of the most flexible ways you can communicate findings with R!

• Like other data products, they can be coded using R and R Markdown, and knitted using Quarto.

• Navigation bar: Icon, title, and author along with links to sub-pages if needed

• Pages, rows, & columns: Defined using markdown headings (with optional attributes to control height, width, etc.)

• Tabsets: Used to further divide content within a row or column

• Cards: Containers for plots, data display, and free form content, typically mapped to cells in your notebook or source document

• Sidebars and toolbars: Present inputs within interactive dashboards

To make your markdown file a dashboard, use the YAML header to specify “flexdashboard”.

---
title: "ADD A TITLE"
output:
  flexdashboard::flex_dashboard:
    theme: 
      version: 4
      bootswatch: pulse
    source_code: embed
  html_document:
    df_print: paged
editor_options: 
  chunk_output_type: console
---

• Dashboards combine the strengths of the other data products we have discussed.

• They are essentially little websites that give viewers the ability to explore your findings, view and download data, and even adjust and interact with visualizations.

• Since they are built with HTML, dashboards:

  • Are essentially universal and readable by any device or browser.

• Keep in mind that they have no password protection. This makes distribution easy but requires extra care when sharing your data (particularly “raw” data).

The Basic Dashboard

The boilerplate dashboard format uses basic web design principles, with titles and overview subjects at the top and your findings below. Basic dashboards are not unlike poster presentations, but they can be built upon to be much more than that.

Navigation Bar and Pages

--- 
title: "Palmer Penguins"
author: "Cobblepot Analytics"
format: 
  dashboard:
    logo: images/penguins.png
    nav-buttons: [linkedin, twitter, github]
---

# Bills

# Flippers

# Data

Using the markdown language for headers, dashboards give users the ability to navigate through different pages both vertically and horizontally.

Sidebars: Global

---
title: "Global Sidebar"
format: dashboard
---
    
# {.sidebar}

Sidebar content (e.g. inputs)

# Plot

```{r}
```

# Data

```{r}
```

Global sidebars are shown at the top of the dashboard by default, and give users the ability to switch between pages. If you had three major findings, for example, this is where users would go between them.

Sidebars: Page Level

---
title: "Sidebar"
format: dashboard
---
    
# Page 1

## {.sidebar}

```{r}
```

## Column 

```{r}
```

```{r}
```

You can nest sidebars within individual pages. For example, you may have two distinct takeaways, visualizations, etc. to better convey a single finding, so you would make one page for your finding, then use the sidebar to show different aspects of that finding.

Layout: Rows

---
title: "Focal (Top)"
format: dashboard
---
    
## Row {height=70%}

```{r}
```

## Row {height=30%}

```{r}
```

```{r}
```

You can arrange dashboard components in rows…

Layout: Columns

---
title: "Focal (Top)"
format: 
  dashboard:
    orientation: columns
---
    
## Column {width=60%}

```{r}
```

## Column {width=40%}

```{r}
```

```{python}
```

…or columns!

Tabsets

---
title: "Palmer Penguins"
format: dashboard
---
    
## Row

```{r}
```

## Row {.tabset}

```{r}
#| title: Chart 2
```

```{r}
#| title: Chart 3
```

Tabsets are tab structures nested within a single component on the page. You can see how you can end up giving viewers a lot of places to click, so use all of these methods judiciously! Less is more. And remember: You can always add more sidebars and tabsets later.

Plots

```{r}
#| title: GDP and Life Expectancy
library(gapminder)
library(tidyverse)
ggplot(gapminder, aes(x = gdpPercap, y = lifeExp, color = continent)) +
  geom_point()
```

You can imbed visualizations via ggplot directly into your dashboards. Each code chunk makes a card, and can take a title.

Tables

```{r}
library(gapminder)
library(tidyverse)
library(gt)
gapminder |>
  group_by(continent, year) |>
  summarize(mean_lifeExp = round(mean(lifeExp), 1)) |>
  pivot_wider(names_from = continent, values_from = mean_lifeExp) |>
  gt()
```

The same goes for tables. Each code chunk makes a card, and doesn’t have to have a title.

Other Features

• Text content

• Value boxes

• Expanding cards

• Interactive elements

  • For interactive exploration, some dashboards can benefit from a live R backend.

  • To do this with Quarto, add interactive Shiny components.

  • Deploy with or without a server!

👉 Your Turn ⤵

Create a data story with our current data set.

• Develop a research question.
• Add ggplot visualizations.
• Model your visualizations.
• Add a short write up for the intended stakeholders.

Resources

Quarto Dashboards in R

Quarto Dashboard gallery

Interactive dashboards for R

What’s Next?

• Complete the Communicate part of the case study
• Complete the Data Products badge
• Optional: Complete the LAW Microcredential