FuzzyDelphiR

An R package for conducting Fuzzy Delphi Method analysis. This package provides comprehensive tools for expert consensus analysis using fuzzy triangular numbers.

Features

• Convert Likert-5 and Likert-7 scale responses to fuzzy triangular numbers
• Calculate fuzzy distance (d-values) for expert agreement analysis
• Determine consensus percentage among experts
• Perform defuzzification to obtain crisp values
• Rank items based on defuzzified values
• Export results to CSV files
• Generate comprehensive analysis reports

Installation

From Source

1. Clone or download this repository
2. In R, install the package:

# Install devtools if not already installed
install.packages("devtools")

# Install FuzzyDelphiR from local source
devtools::install("path/to/FuzzyDelphiR")

Build from Package Directory

# Navigate to the package directory and build
setwd("path/to/FuzzyDelphiR")

# Install required dependencies
install.packages(c("roxygen2", "knitr", "rmarkdown"))

# Generate documentation
roxygen2::roxygenise()

# Install the package
devtools::install()

Quick Start

Basic Usage

library(FuzzyDelphiR)

# Load sample data (27 experts, 10 items with Likert-5 scale)
data(expert_data)
head(expert_data)

# Run complete Fuzzy Delphi analysis
results <- fuzzy_delphi(expert_data, likert_scale = 5)

# View results
print(results)
summary(results)

Step-by-Step Analysis

# Your data: rows = experts, columns = items
my_data <- data.frame(
  Question1 = c(5, 4, 5, 5, 4, 5),
  Question2 = c(4, 5, 4, 5, 5, 4),
  Question3 = c(3, 4, 3, 4, 4, 3)
)

# Step 1: Convert to fuzzy numbers
fuzzy_data <- likert5_to_fuzzy(my_data)

# Step 2: Calculate fuzzy scale (d-values)
fuzzy_scale <- calculate_fuzzy_scale(fuzzy_data)

# Step 3: Calculate consensus
consensus <- calculate_consensus(fuzzy_scale, threshold = 0.2)

# Step 4: Defuzzification
defuzz_values <- defuzzify(fuzzy_data)

# Step 5: Rank items
rankings <- rank_items(defuzz_values)

# View results
print(fuzzy_scale)
print(consensus)
print(rankings)

Export Results

# Run analysis
results <- fuzzy_delphi(expert_data, likert_scale = 5)

# Export all results to CSV files
export_results(results, output_dir = "output", prefix = "my_analysis")

# Generate text report
generate_report(results, output_file = "fuzzy_delphi_report.txt")

Main Functions

Data Conversion

• likert5_to_fuzzy() - Convert Likert-5 scale to fuzzy triangular numbers
• likert7_to_fuzzy() - Convert Likert-7 scale to fuzzy triangular numbers

Analysis Functions

• calculate_fuzzy_scale() - Calculate fuzzy distance (d-values)
• calculate_consensus() - Calculate expert consensus percentage
• defuzzify() - Convert fuzzy numbers to crisp values
• rank_items() - Rank items based on defuzzified values

Main Wrapper

• fuzzy_delphi() - Perform complete Fuzzy Delphi analysis

Utilities

• export_results() - Export results to CSV files
• generate_report() - Generate text summary report
• print.fuzzy_delphi() - Print method for results
• summary.fuzzy_delphi() - Summary method for results

Understanding Results

d-value (Fuzzy Distance)

• d ≤ 0.2: Excellent agreement (consensus achieved)
• 0.2 < d ≤ 0.5: Moderate agreement
• d > 0.5: Low agreement (further rounds needed)

Consensus Percentage

• ≥ 75%: Strong consensus
• 50-75%: Moderate consensus
• < 50%: Weak consensus

Example Output

results <- fuzzy_delphi(expert_data)
print(results)

# Fuzzy Delphi Method Results
# ============================
#
# Overall Results:
#   - Overall d-construct value: 0.15
#   - Overall consensus percentage: 82.5%
#
# Item-level Results:
#   - Average d-value per item:
#     Item1  Item2  Item3  Item4  Item5
#      0.12   0.14   0.13   0.18   0.16
#
#   - Item rankings:
#     Item1  Item2  Item3  Item4  Item5
#        1      3      2      5      4

Data Requirements

Your input data should be structured as:

• Rows: Individual experts/respondents
• Columns: Items/questions being evaluated
• Values: Likert scale responses (1-5 or 1-7)
• Column names: Can be any descriptive names (e.g., "Question1", "Factor_A", etc.)

Example:

   Item1 Item2 Item3
1      5     4     5
2      4     5     4
3      5     5     5

Package Dependencies

This package uses only base R functions and has no external dependencies for core functionality. Optional dependencies:

• knitr and rmarkdown for building vignettes

License

GPL-3

Citation

If you use this package in your research, please cite:

Chan, LK (2025) FuzzyDelphiR  [R library]. Retrieved from https://lerlerchan.github.io/FuzzyDelphiR/

The jamovi project (2025). jamovi. (Version 2.6) [Computer Software]. Retrieved from https://www.jamovi.org.

References

For more information on the Fuzzy Delphi Method: Md Jani, Noraniza & Zakaria, Mohd & Maksom, Zulisman & Shariff, Md & Mustapha, Ramlan. (2018). Consequences of Customer Engagement in Social Networking Sites : Employing Fuzzy Delphi Technique for Validation. International Journal of Advanced Computer Science and Applications. 9. 10.14569/IJACSA.2018.090938

N. A. M. Saffie, N. M. Shukor and K. A. Rasmani, "Fuzzy delphi method: Issues and challenges," 2016 International Conference on Logistics, Informatics and Service Sciences (LISS), Sydney, NSW, Australia, 2016, pp. 1-7, doi: 10.1109/LISS.2016.7854490

Author

Chan Ler-Kuan (lkchan@sc.edu.my) ClaudeCode

Contributing

Contributions are welcome! Please feel free to submit issues or pull requests.