Machine Learning Socionics Test

I am now developing a data-driven and self-reported likert style socionics questionnaires. The test items in the inventory are brought together from various different sources, and I have no idea what variables many of them measure, and moreover I have made no attempt to create any answer key to figure out what they measure. The idea of this test is to examine whether different representatives of different types will give systematically similar answers such that a machine learning classifier could predict the type a human diagnostician would assign. To the best of my knowledge, this has never been done before (if you know of anyone that has done a similar project, please let me know in the comments), in any typology (I am aware that Viktor Talanov collects extensive self-reported data in a vaguely similar vein, but does not employ machine learning to classify types, in fact his questionnaire appears to take a totally theoretical rather than data-driven approach in its design).

In order to train the classifier, a large amount of sample responses of people known to be each type is needed. In order to collect a lot of data in a way where I am confident of the ground truth types used to train the classifier, I am offering free diagnostics interviews to anyone who participates in the project and submits their answers to the lengthy questionnaire (and whose type I am not already confident of from another method). In order to do diagnostic work on a large sample, I need to be able to collaborate and discuss any interviews done with other people in private, and likewise it is important that other interested researchers and skeptics be able to examine and potentially reclassify the ground truth types in order to fully evaluate the method. As such, the interviews will be recorded and shared privately for research purposes with some people I trust, but do not need to be publically available to the world as the other interviews on my blog are. Interviews must be done over voice, although video is preferred. 

This free interview service will not last forever, but probably this project will take a very long time.

The questionnaire is divided into five sections which can be taken separately and at different times (since the whole thing is quite long). All five sections must be completed. If you have already completed Vladimir Vincent's quadra research test, you do not need to complete the first section (labeled Section V).

Section V

Section R

Section L

Section P

Section T

To participate and schedule interviews, please first fill out the questionnaire with a good email address to contact you at, and I will contact you. If I do not get in touch with you after you complete the questionnaire, feel free to email me at aestrivex at gmail dot com.