The Fourth Step Addendum: The George Kelly Personality Inventory

 A previous blog discussed whether the Fourth Step essay is an "essay" or a table.  (You hear both).  

When I got to AA in 1990 and saw the Fourth Step Table Template (Big Book Ch 5), it immediately reminded me of the George Kelly Personality Inventory.   (I knew of that from a 1979 personality psychology class taught by Prof. Walter Mischel, a student of Prof. George Kelly.)

Kelly has you list about 20 persons and then evaluate how they are similar or different.  The point is not the persons lists (e.g. favorite high school teacher) but how you frame and structure what axes in thinking about people.

I would say:  Why not the same for the fourth step  The point of the fourth step is not the particular facts in each of 30 rows, but the fact that 15 of your examples are about sex and 10 about money, so your two main issues are sex and money.   (Use those in steps 6-7).

Here is some Chat GPT facts about the Kelly Inventory.

###

Yes. You are remembering George Kelly’s Role Construct Repertory Test pretty well. It begins with a Role Title List: the person fills in the name of someone in each slot, such as mother, father, a liked teacher, a disliked teacher, someone hard to work with, a boy or girl from high school, and so on. Then those named people are compared in triads to elicit personal constructs.

The original “Minimum Context Card Form” role-title list was:

1. A teacher you like
2. A teacher you disliked
3. Your wife/present girlfriend or husband/present boyfriend
4. An employer/supervisor/officer you found hard to get along with
5. An employer/supervisor/officer you liked
6. Your mother
7. Your father
8. Your brother nearest your age
9. Your sister nearest your age
10. A person you worked with who was easy to get along with
11. A person you worked with who was hard to understand
12. A neighbor you get along well with
13. A neighbor you find hard to understand
14. A boy you got along well with in high school
15. A girl you got along well with in high school
16. A boy you did not like in high school
17. A girl you did not like in high school
18. A person of your own sex you’d enjoy having as a travel companion
19. A person of your own sex you’d dislike having as a travel companion
20. A person recently close to you who appears to dislike you
21. The person you’d most like to help
22. The most intelligent person you know personally
23. The most successful person you know personally
24. The most interesting person you know personally

So your memory of a structured list like “favorite grandparent / first boss / friend in high school” is quite close, although the classic Kelly list uses teacher, parent, sibling, boss/supervisor, neighbor, high-school peer, trip companion, and notable person categories rather than exactly “favorite grandparent.”

There was also a shorter Personal Role Form with 15 roles, including mother, father, brother, sister, liked/disliked teacher, spouse or current closest romantic partner, stressful supervisor, someone who seems to dislike you, someone you’d like to know better, someone you’d like to help, and the most intelligent/successful/interesting person you know/

##

And the point is...

###

In Kelly’s Role Construct Repertory Test, the named people are the “elements.” Then the examiner presents them three at a time and asks a question like: “In what important way are two of these alike but different from the third?” The answer is then put into a bipolar construct, such as warm vs. cold, reliable vs. unreliable, strong vs. weak, accepting vs. critical.

So the test is not mainly about the 20 or 24 people themselves. The people are really a scaffold or sampling device. The real target is the person’s own habitual dimensions of judgment—the internal lenses through which they interpret other people. Kelly called these personal constructs, and the test is designed to reveal the person’s implicit personal theory of social reality.

In other words, if someone repeatedly says things like “safe vs. threatening,” “competent vs. helpless,” “genuine vs. fake,” “controlling vs. relaxed,” that begins to map the person’s private coordinate system for understanding human beings. That is exactly why the method was so distinctive: it tried to uncover the person’s own categories rather than forcing them into a prefabricated questionnaire.

A slightly finer point: Kelly’s method usually asks for two alike / one different, so it elicits a contrast dimension, not just a free description. That contrast is crucial, because it turns a vague impression into a structured axis of meaning.

So your recollection is right: the whole point is to reveal the subject’s internal frame of reference for construing persons.

A modern paraphrase would be: “Show me the categories through which you sort people.”