Predicting MBTI Personality Type of Twitter Users

Sep 17, 2026

Uploaded by Levi

As internet use expands, social media reshapes self-expression and interaction, making personality traits relevant to online behavior. This thesis studies relationships between self-reported MBTI personalities and social media activity, using 4,000 Twitter users and 425,752 tweets. Twenty-six activity features across social connections, activities, content topic and sentiment, and temporal patterns are analyzed with t-tests, stepwise logistic regression, and machine learning models to predict personality. Results show strong associations and a best-model accuracy of 67.6%.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages36

Reading Time1 h 31 min

AccessFree to download

Introduction

Background and Prior Work

Rationale for Predictive Models

Study Proposal

PREDICTING MBTI PERSONALITY TYPE OF TWITTER USERS By WEILING LI A thesis submitted to the Graduate School-Camden Rutgers, The State University of New Jersey In partial fulfillment of the requirements For the degree of Master of Arts Graduate Program in Psychology Written under the direction of Dr. Daniel Hart And approved by | Dr. Daniel Hart | | --- | | Dr. Sarah R. Allred | Dr. I-Ming Chiu Camden, New Jersey May 2021 THESIS ABSTRACT Predicting MBTI Personality Type of Twitter Users by WEILING LI Thesis Director: Dr. Daniel Hart As the internet becomes more pervasive, people shift their self-expression and social interaction activities towards online platforms. It is therefore relevant to relate personality traits to behavior in social media domains. In this study, I investigated the relationship between one’s social media profiles and their MBTI personalities. I performed this research based on 4000 Twitter users who self-report their personalities and 425,752 tweets these users posted. I extracted 26 features that map into five categories of the user’s social media activities: Social Connections, Social Activities, Content Topic, Content Sentiment, and Temporal Patterns of Activities. I used two-sample t-tests and stepwise logistic regressions to examine the association between people’s social media activities and their personality traits. Then I used machine learning methods to predict one’s personality based on their social media profiles. Findings indicate that: 1) there are strong associations between personality and social media activities; 2) MBTI personality could be predicted by social media data and the best model achieved an average testing accuracy of 67.6% . I discuss the implications for further research in broader domains. Keywords: MBTI, Personality, Social Media, Twitter, Prediction, Machine Learning 1 Introduction Personality, deﬁned as a bundle of traits in people’s behavior, cognition, and emotion, is an important topic in psychological studies (Mischel, Shoda, & Ayduk, 2007) . Personality traits are related to a wide range of topics in psychological studies, including identity, depression, anxiety, abuse, and poor health (Bagby, Quilty, & Ryder, 2008) . For example, studies find strong associations between depression and three personality traits – neuroticism, extraversion, and conscientiousness (Klein, Kotov, & Bufferd, 2011), and matching the patient’s personality with different types of treatments (pharmacotherapy or psychotherapy) will maximize the response (Bagby & Quilty, 2007) . In addition, personality has been used to solve many practical problems in various industry domains such as security, advertising, and human resources (Celli, Massani, & Lepri, 2017; Turban et al., 2017) . For example, substantial evidence reveals that personality constructs are strong predictors for work performance (Hurtz & Donovan, 2000) and workplace antisocial behavior (Lee, Ashton, & Shin, 2005) . Therefore, according to Psychology Today, more than 80 percent of Fortune 500 companies use some forms of personality tests for job screening, leadership development, team building, and sales training (Dattner, 2008) . Traditionally, questionnaires have been used for personality assessment. As the internet becomes more pervasive, people shift their self-expression and social interaction activities towards online platforms. As a result, there is a growing trend of studies that focus on using users’ online proﬁles and behaviors to predict their personalities (Bachrach et al., 2012; Gosling et al., 2011; Qiu et al., 2012; Quercia et al., 2012; Wang, Qu, & Sun, 2013) . For example, Bachrach et al. (2012) found that there are signiﬁcant 2 associations between users’ personalities as measured by questionnaire and characteristics of their Facebook proﬁles such as the size and density of friendship network, the number of uploaded photos, and the number of events attended. Moreover, th

FAQs

What data and participant information were used for predicting MBTI types?

The study used 4,000 Twitter users who self-report their MBTI personalities, along with 425,752 tweets posted by those users.

How were social media behaviors represented in the analysis?

The research extracted 26 features mapped into five categories: Social Connections, Social Activities, Content Topic, Content Sentiment, and Temporal Patterns of Activities.

What methods were used to test associations and build prediction models?

Two-sample t-tests and stepwise logistic regressions were used to examine associations between social media activities and personality traits, and machine learning methods were used to predict personality from profiles.

Sep 17, 2026

Uploaded by Levi

Predicting MBTI Personality Type of Twitter Users_Page 1

Open in WPS to view all 36 pages