Fred Borgen whose major professor was David Campbell. Fred's doctoral dissertation was the factor analysis that resulted in the basic interest scales in what was called at that time the Strong-Campbell Interest Inventory. My relationship with Fred Borgen grew, and I began to be involved with his research projects. The following spring I took my first course in psychometrics (it used Frederick Brown's text and was taught by Leroy Wolins). Dr. Wolins was involved with work on the Kuder Career Assessment and was a leader in the field of measurement analysis and psychometrics. He affectionately called me "boy" in class, even though I already had one grandchild, a delight to me that my younger classmates could not comprehend. Dr. Wolins taught me to question everything, to make no assumptions about measurement and psychometrics. These were welcomed words for a critical rebellious thinker like myself. He told me to continuously and critically evaluate a test item.
I began to design and run research projects, hoping to collect data that I might one day use for a Master's Thesis. I voraciously researched questionnaires and wrote my own test items for everything that I could not find to complete my research. I was asking nearly 800 questions to my research participants (victims) and Fred Borgen accused me of attempting to develop the theory of everything. He didn't know—I was. It had become obvious to me that questions could each have their own particular meaning for each and every individual, and I would have to be careful about all my assumptions of measurement, psychometrics and the intersection with people.
I was offered a position in the PhD. Program at Iowa State. At that time I was running three research projects and was the paid undergraduate research assistant/department computer consultant. I entered the Counseling Ph.D. program in the fall of 1991, continuing to collect research data and trying to find theoretical orientation for behavior, personality and psychotherapy that suited me. I was exposed to most of the common measures of personality, including the MBTI® Form G.
Iowa State did not have a clinical program, and I desperately wanted to know how to help individuals resolve the serious problems that can occur in life. I left Iowa State after the spring semester of 1992 and enrolled as a graduate student-at-large, taking clinical courses at the University of Nebraska-Lincoln. While my statistics training at Iowa State was excellent, the psychometric courses at Nebraska seemed more in-depth and more to my liking.
In the spring of 1993 I began working on a Strong Interest Inventory revision team with Fred Borgen, traveling back and forth to Iowa State. During the same time I finished my Master's thesis, and transferred graduate credits back to ISU to earn my MS in Counseling Psychology. Because of my computer and psychometric ability, CPP Inc. (formerly Consulting Psychologists Press, Inc.) hired me as a contract consultant to do the statistical analysis for the 1994 revision of the Strong. The work lasted a year and finished with me writing the 13,000 line scoring routine that is used in the 1994 Strong Interest Inventory®. During that time I read virtually all of the SAS manuals, the software program for statistical analysis and deepened my understanding of the statistical analysis and the software, but began to question many of the assumptions about the work that I was being asked to do on the Strong.
I applied and was admitted to the Ph.D. program in Counseling Psychology in the fall of 1994. I began taking all the psychometric courses that they would let me. Once again I found myself constantly running two or three research projects at a time, doing analysis on many instruments and many projects, but having a consistent nagging feeling that all was not well in the Camelot of measurement.
In 1996 I was asked by Allen Hammer to do the data analysis for the MBTI® (Myers-Briggs Type Indicator®) Form M revision. It was during that time that I recognized the need to attend to the fact that individual differences will violate the assumptions of classical measurement theory. When the Form M project was completed, I began to work on my dissertation. I knew that I had to design a project that would prove the assumptions that were currently in vogue in classical measurement were false. Specifically, it seems that a belief existed that if you simply collect a large enough sample, run a factor analysis to form scales and get a good alpha for the items on your scale, you had a good measure. This is not true. I discovered personality, interests, gender and other important factors may have their impact overwhelmed when you have a large sample, but they not go away. I wanted to know what is going on for small homogeneous (similar in important characteristics) groups of people for which measures don't work. Could new steps in measurement compensate for this and include what was sometimes called "outliers"? There is an entire collection of descriptive words that are used in psychometrics for individuals and information that does not fit the classical mode. I knew it was the classical mold that needed to be amended and that calling names at the things that did not fit your design would get you nowhere.
I finished my work at Nebraska and defended my dissertation in the spring of 1998 before going to my internship at the University of Maryland where I did research and published with William Sedlacek. It was there where I used some of my new thinking in trying to look at structure within entering freshmen classes.
Halfway through my internship I was offered a part-time position as a research scientist at CPP. I was assigned the Myers-Briggs product line. One of the first things that I did was to perform a confirmatory factor analysis on the 1997 Form M revision data. This led me to a relationship with some of the statisticians at Scientific Software International (SSI) who schooled and guided me into some of the nuances and fallacies of structured equation modeling. Before I finished my internship I also spent a week in Montreal, Canada learning advanced IRT (item response theory) and how to use some of the new software that was available. This would later be beneficial in doing the development work on the MBTI® Form Q.
I completed my Ph.D. in May of 1999 and in June I moved to Minnesota to be a full time research scientist on the CPP research team. I was assigned the Form Q development project. Even though IRT was an improvement over classical measurement, it still was prone to the same psychometric fallacies, namely the assumption that if an instrument looks good psychometrically on paper, it works well for everybody.
In the spring of 2000 I accepted a position as Director of Research at CAPT, which allowed me to move back to my childhood home of Florida, and gave me an opportunity to be more involved with psychological type. I finished the Form Q project as a contract consultant for CPP while entering my new role at CAPT (Center for the Application of Personality Type). In January of 2001 I was asked to head up the MBTI® Step III research project for both CPP and CAPT. I had a great sense of frustration with Step II and Step III because no one was listening to me about the psychometric issues.
In the summer of 2001 Step II was rolled out at the Minnesota APT (Association for Psychological Type) Conference and shortly thereafter we had the first large Step III research meeting. I presented what I had found and discovered. Then I realized something very important. It was my thinking and psychometrics that was driving the project. I made the decision at that time to create my own instruments and try to overcome the measurement errors and psychometric fallacies that I had realized over the last decade.
As I wrote the items and collected the data for the MajorsPTI™ (Majors Personality Type Indicator™), I recognized three things: forced choice was not effective; a graduated response would allow the person to directly weight their choice and the person was the only one who could determine if an item was good. This connecting the measurement item with the reality of the person I termed "Reality Based Assessment Accuracy." The use of a neutral and weighted response I called "Differential Intensity Weighting." When the electronic version of the MajorsPTI™ was released in 2003 I was finally able to incorporate the information that I knew about measurement fallacy and introduce an adaptive testing module referred to as a Type Precision Module that further improves the PTI's ability to accurately assess type (92% accurate). This module introduces items that do not fit the big population norm. They only work for the small group who fall near the middle, those outliers—the difficult ones.
In 2005 I began collecting data on an expanded type assessment that later became the Majors PT-Elements. Having spent many years looking at the original 1923 text "Psychological Types" I knew that only half of the Jungian story was being told. There was no assessment of the Jungian eight mental functions and no assessment was providing information on the blockages to the natural expression of a person's psychological type. In 2008 the MPT-Elements was released and contained the Majors/Jungian 8-Process Scores. These scores represent the level of access and utility that the individual holds on each of the eight mental functions. This dynamic information expresses what experience has given to the innate personality type. The Personality Formation Information (Scores) was the second innovation presented in the MPT-Elements. These scores indicate blockages to the natural expression of type that are both internal and external. The assessment is very complex and took many hours of advanced statistical analysis on large sets of data to complete. The full use of the Majors PT-Elements requires advanced training and understanding of Jung's original text. It seems that challenging psychometrics can present a clearer and more complete picture of the individual.
Since beginning my own work I have continuously been able to innovate by reducing error and avoiding the pitfalls of classical measurement. I break classical rules in order to do a better job; providing accurate information to individuals who want to learn more about themselves and grow. Nowadays I find myself rebelling against my last measurement thought, never satisfied, knowing it can always be better. I have embraced the direction and enjoy the process. Are there new things on the horizon? Yes, and I hope to finish them in the next few years. New thoughts and new innovations. |