The Success and Happiness Attributes Questionnaire (SHAQ) And Cognitive Systems Projects
Tom G. Stevens PhD
SHAQ is currently a LispWorks delivered app that can be downloaded (free) from my website at web.csulb.edu/~tstevens. SHAQ has 81 scales and subscales and is one of the most comprehensive personality questionnaires ever written. The results of SHAQ are automatically sent back to my website [users must consent] and are now completely anonymous. Users can either complete all scales, their choice of scales, or allow SHAQ to select the scales for them depending upon how they answer a few questions [an expert system]. SHAQ was based upon my book, You Can Choose To Be Happy: Rise Above Anxiety, Anger, and Depression (revised 2010).
My original study using 3446 SHAQ Internet users (2009, see website) has a huge amount of data and found very strong support for both the validity of SHAQ and its association with important life outcomes. The SHAQ scales had moderate to high positive correlations with almost all outcome measures. SHAQ's subscales had surprisingly high multiple correlations with the emotional outcomes; with Overall Happiness, R = .865; with Low Depression, R = .730; with Low Anxiety R = .675; Low Anger-Aggression, R = .701 (N = 1123 for users completing all scales). I devised the Happiness Quotient (HQ) to get an overall predictor of happiness. Overall SHAQ could account for over 75% of user happiness variance. SHAQ correlated with the Relationship Outcomes scale, R = .693; the Health Outcomes scale, R = .816; Highest Income, R = .486; and Educational Attainment, R = .458.
Development of software
I originally wrote SHAQ in 2000. I wanted to write it in Common Lisp then. I had studied AI during post-graduate work at University of California, Irvine in their CS & AI Department. I (along with most of their faculty) concluded that Lisp was far superior to any other computer language for (at least) AI applications. However, in 2000 I did not know how to use Lisp on my website for users to complete SHAQ. Therefore, I learned Java and wrote a very large Java Applet to deliver SHAQ to my website users. It was torture using Java after using Lisp. To name just a few problems for me: rigid classes where functions apply only to a class without having to add long dotted names to them, no macros, NO LISTS for dumping all sorts of objects into, limited use of functions as data, etc, etc. To summarize: Lisp is more like natural English writing. You can just create functions that follow the algorithm that is in your mind. You don't have to add a lot of data declarations/typing, or create special names that don't relate to your meanings. And certainly—no attention to memory per se. To a new user, using all those parentheses may seem awkward, but with a good editor (e.g. the LispWorks editor) there is no problem, and it makes nesting functions, data, etc into huge hierarchies/trees very easy. Lisp allows you to store any kind of data any place in them.
The human brain operates in a giant hierarchical structure with extremely flexible and complex connections to model and control our worlds. Lisp allows us to model the world or write any kind of control mechanism in a very similar way. Java and other languages do not [at all or not so easily].
When Applets went kaput, I decided to revise SHAQ and make it a Lisp application. After trying several other Lisp versions, I chose LispWorks for a variety of reasons. ANSI Common Lisp was a must. I liked everything about LispWorks — including its editor and IDE. For me it was essential to have no license fee for deliverables. After all, I'm giving SHAQ away for free. I wrote Lisp code to convert as much Java to Lisp as possible to minimize manual new code and data writing. That was not much of a problem, because Lisp makes it so easy to manipulate strings in sophisticated ways. [I treated Java code mostly as strings.] Soon I had SHAQ in Lisp on my website sending back user results from their computers. SHAQ has been functioning well in the years since. I have also lately been extending it to do much more for my upcoming project.
Now in my "retirement" I have been pursuing a long-term interest to develop a cognitive systems computer model of human personality—my Cognitive Systems (CS) Model. I have used LispWorks to:
- Write a questionnaire that explores users' value (and other) categories [based on Personal Construct Theory]. Each subject is tested to develop a unique CS model network that reflects his/her own unique values/category network. There is no limit to how extensive that network can be.
- I have combined that flexible questionnaire with a CS model framework based upon SHAQ scales and other theoretical ideas I have used from the AI and Cognitive Science literature (e.g. Stephan Grossberg and John Anderson).
The result is a network system of nodes and paths. Each node/cell [csym] has meaning, connections, and functional relations with other csyms though paths. One difference with other neural network models is that this one attempts to model human thinking at the "personality level" of functioning.
Summary
In my experience and opinion no other language can compare to Common Lisp, and LispWorks has the best overall package/pricing of any on the market. I think that if the world of computing turned to Lisp—especially in this day of AI—software engineers and software would be far more efficient. In the 1990s I wrote a Lisp program that could analyze CSULB college student progress toward all degrees by myself in a few hours/week. CSULB gave me some released time to develop this program since the $1 million plus they were paying to IBM to develop a similar program in COBOL was delayed by several years. In my consulting with their software people, I concluded that a huge advantage of Lisp was being able to so easily modify code (in only one place) when adjustments had to be made for changes in degree requirements, etc. The IBM folks couldn't keep up with the changes for the entire university. I had two student assistants, who knew nothing about programming, typing in changes in Lisp lists that looked like English, and all they had to do was change these requirements and my core program didn't need any changes.
Tom G. Stevens PhD
Counseling Psychology Professor Emeritus
California State University, Long Beach
web.csulb.edu/~tstevens