Nearest Neighbors Personality-Type Matching Survey
This is an interactive personality quiz that will compare your responses to a population sample who took the survey and show you what personality types those users who gave responses close to yours identified as. This test reports results in two different personality frameworks, and serves as a good + critical introduction to the various popular personality typologies.
Background
The most popular way for a consumer oriented personality test is to report its results in terms of a "personality type", a category of people the user belongs to. There are many such type systems, the Myers-Briggs Type Indicator and the Enneagram of Personality being the most common on the internet. There is no scientific evidence for personality types, but it is possible that they could nevertheless be useful as a practical way to describe people.
This quiz takes a machine learning approach that assumes personality types are not real. It is based on a dataset of 10,000 people who self-reported their MBTI and Enneagram personality types. Your answers in this survey will be matched to their answers, allowing them to be ranked in terms of closeness to you. Your results will take the 100 closest and explore what personality types they identify as. This is an application of the k-nearest neighbors algorithm to personality types.
The NNPS was published in 2025 and is the first on-line personality test to use this methodology. For more information, read the documentation.
Test Instructions
This test has 125 items, though you only have to answer as many as you want and can submit it at any time to get your results. It's recommended to answer at least 40 to get reliable results. For each item, drag the slider to where you fall on the scale between the two adjectives at either end.
Participation
Your use of this test should be for personal entertainment or educational use only. At the end of the survey you will be asked if you agree to allow your data to be saved and added to the dataset of the test or used for other research purposes. No identifying information will be collected, so risk to your privacy is assumed to be minimal.