Analog to Multiple Broadband Inventories (AMBI)

This is an interactive personality test that combines the scales from 8 different well regarded tests into one.


Academic personality psychology settled on the five factor model of personality in the 1990s, but before that there were a large number of competing personality inventories. These inventories have a lot of overlap in the concepts they measure and the questions they use, but each organizes their results in different ways. Given this overlap it is possible to score different scales from the same set of items. The AMBI was created in 2010 by Tal Yarkoni as a demonstration of how this could be done. The original 8 inventories contain a total of 2,019 questions. The AMBI produces scores that are substantially equivalent to all the original scales using only 181. The AMBI is available from the International Personality Item Pool, and is in the public domain.

The 8 inventories the AMBI mimics are these:

  • Revised NEO Personality Inventory
  • HEXACO Personality Inventory
  • Temperament and Character Inventory
  • Hogan Personality Inventory
  • Six Factor Personality Questionnaire
  • California Psychological Inventory
  • Revised Jackson Personality Inventory
  • Multidimensional Personality Questionnaire
  • The average scale in these inventories are reproduces by the AMBI with 72% similarity [1].

    Test Instructions

    The test has 181 statements of opinion that you must rate on a five point scale of how much you agree with each. It takes most people between 10 and 20 minutes to complete.


    This test is provided for educational and entertainment use only. It should not be used as psychological advice of any kind and comes without any guarantee of accuracy or fitness for any particular purpose. Also, your responses may be recorded and anonymously used for research or otherwise distributed.

    • [1] The correlation between each of the scales in the original inventories and their analogs in the AMBI are reported by Yarkoni (2010). This is an underestimate of the true overlap though, so the convergent correlations are divided by the test-retest reliabilities reported in the test manuals. This calculation is ad-hoc, so sould not be taken as an extremely good estimate.
    Page information

    Updated: 17 September 2019


    Privacy policy

    Results page and public comments