Methods for Deliberation Analysis.

deliberr: Deliberation in R

Overview

deliberr is an R package designed for analyzing deliberation data in political science and communication research. The package provides tools for measuring and evaluating deliberative quality, including the Deliberative Reasoning Index (DRI) and other metrics commonly used in deliberation studies.

Whether you're analyzing focus groups, town halls, deliberative polls, or online discussions, deliberr helps you assess the quality of deliberative discourse with both quantitative metrics and visual outputs.

Key Features

• Deliberative Reasoning Index (DRI) calculation
• Statistical analysis tools for deliberation data
• Data visualization capabilities for deliberative metrics
• User-friendly graphical interface (no coding required!)
• Support for multiple data formats (CSV, Excel, etc.)

Installation

Requirements

• R version 3.5 or higher
• Internet connection for package installation

Install from GitHub

You can install the development version of deliberr directly from GitHub using the devtools package:

# Install devtools if you haven't already
install.packages("devtools")

# Install deliberr
devtools::install_github("gumbelino/deliberr")

Load the Package

After installation, load the package:

library(deliberr)

Usage

For Non-Programmers: Graphical Interface

You don't need to know how to code to use deliberr! The package includes a user-friendly graphical interface that you can access in two ways:

Option 1: From R

library(deliberr)
open_gui()

Option 2: Web Browser (No R Installation Required)

Visit: https://gumbelin.shinyapps.io/deliberr/

The graphical interface allows you to:

• Upload your data files
• Select variables for analysis
• Calculate deliberation metrics
• Generate visualizations
• Export results

For R Users: Code-Based Analysis

If you're comfortable with R, you can use the package functions directly in your scripts. Documentation for all functions is available within R:

# View package documentation
?deliberr

# Get help on specific functions
?dri  # Example: Help for DRI calculation

Getting Started Guide for Researchers

Step 1: Prepare Your Data

• Organize your deliberation data in a spreadsheet (Excel or CSV)
• Ensure each row represents an observation (e.g., a speaker turn, a participant)
• Include columns for relevant variables (e.g., speaker ID, statement content, coded categories)

Step 2: Open the Interface

• If using R: Install the package and run open_gui()
• If using web browser: Go to https://gumbelin.shinyapps.io/deliberr/

Step 3: Upload and Analyze

• Upload your data file through the interface
• Follow the on-screen instructions to select variables
• Run your analysis and view results
• Download outputs for your research

Support and Documentation

Getting Help

• Package documentation: Type ?deliberr in R console
• Function help: Type ?function_name (e.g., ?open_gui)
• Issues and bugs: See "Reporting Bugs" section below

Reporting Bugs

If you encounter any issues or bugs while using deliberr, please report them on GitHub:

1. Go to: https://github.com/gumbelino/deliberr/issues
2. Click "New Issue"
3. Provide a clear description of:
   • What you were trying to do
   • What happened (error messages, unexpected behavior)
   • Your R version (type R.version.string in R console)
   • Your package version (type packageVersion("deliberr") in R console)

For questions about usage or interpretation of results, please open a discussion on the GitHub repository.

License

This package is licensed under the MIT License. See the LICENSE file for details.

Author

Gustavo Kreia Umbelino
Email: [email protected]

Dependencies

The package automatically installs all required dependencies, including:

• dplyr, ggplot2, tidyr (data manipulation and visualization)
• shiny, DT (graphical interface)
• stats, rstatix, psych (statistical functions)
• And other supporting packages

You don't need to install these separately.

Acknowledgments

Thank you to all researchers and practitioners who have contributed feedback and suggestions to improve this package.

Questions? Open an issue on GitHub or contact the package maintainer at [email protected].