Summary: The traditional DISC assessment—a common tool in recruitment and team development—has been updated using machine learning. New research from the University of East London shows that AI models can reproduce DISC results with high accuracy while reducing the number of questions and the time needed to complete the test.
The study demonstrates that a standard 40-question DISC questionnaire can be reduced to a compact 10-question form while retaining most of its predictive power. In addition to greater efficiency, the machine learning approach identifies blended personality profiles rather than forcing respondents into a single category, giving organizations a more nuanced view of behaviour.
Key Facts
- 10-question short form: Researchers selected the most informative items and produced a short-form DISC test that preserves more than 91% accuracy compared with the full instrument.
- Blended profiles: Machine learning exposes hybrid patterns—for example, people who score strongly on both Dominance and Conscientiousness—rather than assigning everyone to a single DISC type.
- Data-driven clusters: Analysis of responses from over 1,000 participants found four natural personality clusters that align closely with the classic DISC categories while highlighting subtle overlaps.
- Practical application: Shorter and smarter assessments are better suited to fast-paced contexts like high-volume hiring, leadership workshops, and rapid team diagnostics.
Source: University of East London
Overview
DISC assessment groups behaviour into four styles—Dominance, Influence, Steadiness and Conscientiousness—and remains popular because it offers quick, practical insights into how people communicate, lead and work together. However, the traditional method relies on rule-based scoring that typically places individuals into a single dominant category. That simplicity is useful but can overlook the complexity of people whose behaviour spans multiple styles.
Researchers at the University of East London explored whether modern data science could preserve the practical benefits of DISC while improving accuracy and flexibility. Using responses from more than 1,000 participants, the team trained several supervised machine learning models to predict DISC classifications from answers to the standard 40-question instrument. The best-performing models reproduced DISC labels with accuracy above 93%, indicating that AI can reliably emulate conventional scoring.
Beyond classification accuracy, the study focused on reducing the length of the questionnaire. Using recursive feature selection, the researchers identified ten high-information items that preserve most of the original test’s discriminative power. A 10-item version maintained more than 91% accuracy, offering substantial time savings while still delivering meaningful personality profiles suitable for workplace use.
Clustering methods were also applied to discover how responses group naturally. The unsupervised analysis revealed four clear clusters that correspond to the established DISC types but also revealed overlaps that traditional scoring can miss. These hybrid patterns reflect real-world behaviour where people may combine traits from different DISC categories depending on context.
Dr Mohammad Hossein Amirhosseini, Associate Professor in Computer Science and Digital Technologies at the University of East London, emphasizes that the approach keeps DISC’s practical simplicity while enriching its insights. According to Dr Amirhosseini, a shorter, AI-informed assessment makes personality profiling more practical for recruitment, leadership development and team-building activities where time is limited.
Rather than replacing human judgement, the AI tools are intended to provide better data to inform decisions. By highlighting blended behavioural tendencies and supplying fast, reliable profiles, machine learning can help managers and coaches place people in roles and teams that align with their natural strengths.
Key Questions Answered
Q: Can a 10-question test match a 40-question test?
A: According to the UEL data, the reduced form performs very well. By selecting the most informative questions, the short version achieves roughly 91% accuracy, which the authors argue is acceptable for many workplace settings given the significant time savings.
Q: Does the AI recognise mixed DISC types?
A: Yes. The model identifies blended profiles and does not force respondents into a single category. This better reflects the complexity of human behaviour and helps reveal how individuals may combine traits across DISC dimensions.
Q: Will AI replace human decisions in hiring?
A: No. The research frames AI as a decision-support tool, offering richer, faster information to help people make more informed placement and development choices rather than automating hiring decisions.
Editorial Notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full.
- Additional context was added by staff.
About this AI and psychology research news
Author: Kiera Hay
Source: University of East London
Contact: Kiera Hay – University of East London
Image: The image is credited to Neuroscience News
Original Research: “Reinventing DISC personality assessment: machine learning approaches for deeper insights and greater efficiency” by Fatima Kalabi; Mohammad Hossein Amirhosseini. DOI: 10.52768/3067-7947/1037. The paper is open access.
Abstract
The DISC personality framework is widely used in applied settings but traditionally relies on fixed, rule-based scoring that can oversimplify behavioural profiles. This study evaluates whether machine learning can offer a more flexible, efficient, and accurate approach to DISC classification. Using a dataset of over 1,000 participants, multiple supervised models—including Logistic Regression, XGBoost, SVM, MLP, Random Forest, and K-Nearest Neighbours—were tested alongside unsupervised clustering techniques. Logistic Regression produced the strongest performance with accuracy around 93.5% and robust cross-validation results.
Recursive Feature Elimination revealed a reduced set of ten key items that maintain over 91% accuracy, enabling a concise assessment that is practical for fast-paced organisational contexts like recruitment, leadership coaching and team composition. Clustering analyses confirmed that the same latent trait structures are recoverable from the reduced questionnaire and that hybrid profiles can be identified. The findings show that machine learning can both replicate and extend conventional DISC assessments, preserving conceptual integrity while offering a scalable, empirically grounded tool for modern psychometrics.