Skip to content

BellaIngenue/Divorce_Rates_And_Big_5

Repository files navigation

Divorce Rates and Big 5/OCEAN Models

Using MATLAB, let's look into Divorce Rates and the connection with Big 5 Personality Assessments

MIT License Open in MATLAB Online

Documentation

Overview of the Research Project

Research Question

When it comes to Typology and the limited research associated, what is the possible connection between Divorce rates and the different levels of Big 5 Typology types? Focusing more specifically on Neuroticism, Extraversion/Introversion, and Openness in a person in a couple/relationship. The other two: Agreeableness and Conscientiousness can go either way in regards to marital success.

Hypothesis

Our hypothesis is that Relationship Divorce/Marital Rates and Typology have some slight connection or correlation with one another. Specifically, Big 5's OCEAN Model measured by the IPIP300. We hypothesize that there will be the highest correlation between Divorce rates and higher Neuroticism counts, Extraversion counts and lower Openness counts.

Ethics and Privacy

Do you have permisssion to use this data? As all of the data we have collected is open to the public for fair use, we have permission to use this data and for this purpose. All of our datasets for our variables are from Kaggle, Figshare, and Google Data.

Are there privacy concerns regarding your datasets that you need to deal with, and/or terms of use that you need to comply with? All of the data compiled and used for our project was collected from different resources, studies, and websites. We are looking for this data through websites such as Kaggle or Figshare and through government websites that provide open-source, free use datasets, so we expect that we shouldn't run into too many complications with privacy, as much of the data has already been released to adhere to data privacy laws.

Are there potential biases in your dataset(s), in terms of who it composes, and how it was collected, that may be problematic in terms of it allowing for equitable analysis? There is a high probability that there would be some significant biases. Due to the idea and fact that Typology in general is known as a Pseduoscience, makes it apparent that it can be hard to correlate and causate between a hard science (such as Psychology) and something like Typology. Having said that, Big 5 is well known as the most empirically-supported Typological Model.

What will you be doing with this data? We do not intend to ever use this project for monetary use, this is strictly for educational and fun purposes only.

Feedback and Contact Section

If you have any feedback, questions, comments, or concerns, please reach out to me at sienasrivera@gmail.com

Screenshots of Data Visualizations

App Screenshot