AAAI HCOMP 2021 - Conjoint Analysis of Human Preferences

This repository contains code to support the analysis of the paper: "Modeling Simultaneous Preferences for Age, Gender, Race, and Professional Profiles in Government-Expense Spending: a Conjoint Analysis" published in the 2021 Proceedings of AAAI Conference on Human Computation and Crowdsourcing.

This repository and associated work may be cited as follows:

@inproceedings{ibrahim2021modeling,
  title={Modeling Simultaneous Preferences for Age, Gender, Race, and Professional 
    Profiles in Government-Expense Spending: a Conjoint Analysis},
  author={Ibrahim, Lujain and Ghassemi, Mohammad M. and Alhanai, Tuka},
  booktitle={Proceedings of the AAAI Conference on Human Computation and Crowdsourcing},
  year={2021}
}

Directory Structure

• datasets: contains three csv files:
  • suggestedadvisor-expense-pairs.csv: the file that is sampled to show advisor profiles and expenses during the experiment
  • results.csv: the file with all the data collected from the experiments
  • users.csv: the file with information on each of the experiment respondents
• ip-map-datasets: contains files used to create heatmap of geographical locations of respondents
• requirements.txt: required python libraries for analysis
• demographic-analysis.ipynb: jupyter notebook to perform demographic analysis of respondents
• global-conjoint-analysis.ipynb: jupyter notebook to perform conjoint analysis on responses from all respondents
• global-plots.ipynb: jupyter notebook to generate sankey plots and bubble plots on respondents' choices for age, gender, race, and profession
• political-party-conjoint-analysis.ipynb: jupyter notebook to perform conjoint analysis on responses from (a) democrats and (b) republicans
• political-party-plots.ipynb: jupyter notebook to generate sankey plots and bubble plots on (a) democratic respondents' and (b) repiublican respondents' choices for gender and race

Running Analysis

System Requirements

• To install Python 3, follow these instructions.
• To install Pip, follow these instructions.
• To install Jupyter Lab/Notebook, follow these instructions. To run Jupyter Lab/Notebook, follow these instructions.
• To set up a virtual environment and use it in Jupyter Lab/Notebook, follow these instructions.

To install requirements:

1. Clone this github repository

git clone <url-to-this-repo>
cd <cloned-repo>
cd public-repo

2. Get Python requirements needed

pip3 install -r requirements.txt

Understanding Datasets

There are two datasets used in the analysis:

1. datasets/results.csv:
   • session_id: unique id of the each respondent's session
   • response_id: unique id of each task response in a respondent's session
   • expense_category: category of the expense presented in task
   • expense: specific expense presented in task
   • age_suggested: age of suggested advisor
   • age_selected: age of customized advisor
   • age_changed: 0 if age_selected=age_suggested, 1 otherwise
   • gender_suggested: gender of suggested advisor
   • gender_selected: gender of customized advisor
   • gender_changed: 0 if gender_selected=gender_suggested, 1 otherwise
   • race_suggested: race of suggested advisor
   • race_selected: race of customized advisor
   • race_changed: 0 if race_selected=race_suggested, 1 otherwise
   • profession_suggested: profession of suggested advisor
   • profession_selected: profession of customized advisor
   • profession_changed: 0 if profession_selected=profession_suggested, 1 otherwise
   • profession_option1: 1st profession option in the customized advisor profession drop down menu
   • profession_option2: 2nd profession option in the customized advisor profession drop down menu
   • profession_option3: 3rd profession option in the customized advisor profession drop down menu
   • profession_option4: 4th profession option in the customized advisor profession drop down menu
   • task_submit_time: task submission time
2. datasets/users.csv:
   • session_id: unique id of the each respondent's session
   • user_state: state the respondent completed the experiment from
   • user_gender: self-reported gender of respondent (survey response)
   • user_age: self-reported age of respondent (survey response)
   • user_degree: self-reported educational level of respondent (survey response)
   • user_politics: self-reported political party affiliation of respondent (survey response)
   • user_vote: self-reported voting behavior (yes/no/not sure) in 2020 national elections of respondent (survey response)
   • user_religion: self-reported religion of respondent (survey response)
   • user_ethnicity: self-reported ethnicity of respondent (survey response)
   • user_income: self-reported annual income of respondent (survey response)
   • session_submit_time: session submission time (survey response)

Running Jupyter Notebooks

All the analysis notebooks used to generate the figures and results used in the publication can be found in this folder:

• Datasets
• Subjects Demographic Analysis
• Global Conjoint Analysis
• Political Party Conjoint Analysis
• Global Sankey & Bubble Plots
• Political Party Sankey & Bubble Plots