Mark Majors, Ph.D. — Author of the Majors Personality Type Inventory™
Dr Mark S. Majors is a counseling psychologist with extensive psychometric credentials. He is the author and developer of the Majors PTI™, Majors OEM™ and Majors PT-Elements™) and principal developer of the Interstrength® X-Styles Assessment. He provided the data analysis on the 1994 Strong Interest Inventory and the MBTI® Form M and Form Q, developed the IRT scoring for the MBTI® Form Q, and co-authored the MBTI® Form Q Manual. In addition to the test development, Mark trains pastoral counsellors with an emphasis on the use of personality assessment for conflict resolution through the acceptance of differences and personal growth. He has developed and presents leadership training seminars that train leaders to serve others by using personality and individual differences to facilitate optimum performance. He has provided 19 years of successful individual and couples/marital counselling and coaching using personality differences and has authored numerous books, manuals and articles on personality differences and biblical psychology. He received his Bachelor of Science and Master of Science degrees in psychology from Iowa State University and his PhD in counselling psychology from the University of Nebraska–Lincoln.
Establishing the reliability of an instrument is accomplished by showing that the scales themselves are reliable. The two forms of reliability assessment most commonly used in instrument development are internal consistency and test re-test reliability. The internal consistency of an instrument is a statement of how well the items on a scale seem to be consistently measuring the same thing. It does not tell you what is being measured, just that all of the scale’s items are measuring it. Test re-test reliability tells us how well the instrument measures consistently the same thing over time. The internal consistency alphas of the 52-item MajorsPTI are presented in Table 3.
|MajorsPTI Scale||Best- Fit||Balanced By Type||Total Sample||F||M||Post Dev. Sample||2009 International Sample||# items/ scale|
The measure of internal consistency was Cronbach’s coefficient Alpha. All of the samples that were collected were included in the total sample (this does not include the post development or 2009 publishers sample). The best-fit sample is comprised of those individuals in the total sample that have reported, and are known to understand, their psychological type. The balanced by type sample was randomly drawn from the total sample and was balanced by not only type, but also sex (N = 260). This sample provides a good indication of how well the internal consistency of the MajorsPTI hold up across all types. The post-development sample was collected over a 3-year period and represents data from across gender, ethnic and socioeconomic strata.
All of the alphas indicate that the MajorsPTI has very good internal consistency. Given that the number of items on each scale range from 10 to 15 these alphas are remarkable. The range of alphas was .92 to .95 for the best-fit sample, .90 to .91 for the balanced by type sample, .89 to .91 for the total development sample, .90 to .93 for the postdevelopment sample and .92 to .93 for the 2009 sample. In table 4 below the alphas for the four scales are given for the development sample divided by the range of level of clarity found in the reported type.
Table 4 shows that the scale reliabilities for all scales improve as the range of clarity moves to the Very Clear end of respondents. The improvements in reliability for the Mod + range (all individuals above Slightly Clear) are .02 for all scales. For the Clear + (all individuals above Moderately Clear) it is an improvement of .03 for the SN scale and .02 for the other three. In the Very Clear category (all above Clear) the increase is again .02 for all scales. These results demonstrate the MajorsPTI’s scale score relationship with scale reliability. Even at the Slight (which includes those with ties) the internal consistency is excellent, and increases across groups are even, indicating similarity of functioning on the four scales. These results point to an extremely important issue regarding clarity of results:
It is the measure’s clarity in evaluating the type of certain individuals; not the clarity of individuals in reporting their type.
Establishing the consistency or stability of measurement over time is known as test-retest reliability. Test-retest analysis is accomplished by correlating (Pearson product moment correlations) the raw scores on the MajorsPTI that were collected using the same sample of people over a period of time (Table 5).
A test-retest (30-day interval) sample was collected from a Protestant church in a moderate size Florida town. The sample was made up of 35% males and 65% females and had a mean age of 35 years with a range from 19 to 76. The first MajorsPTI administration collected from this sample was included in the total sample use in the internal consistency analysis above. The second column of Table 5 presents the percentage of agreement of the dichotomous results between administrations. Results from the test-retest sample indicate that the MajorsPTI is very reliable over time. The range of Pearson correlations was from .88 to .92.
The E/I scale demonstrates the greatest stability over time and the S/N and T/F scales have the lowest. Yet, even the .88 correlation for the two lower scales indicate very stable measurement over time. The percentage of agreement of dichotomous results show that the MajorsPTI is providing the same four letter-code consistently.
Conclusion of Reliability Analysis
The reliability data that has been presented demonstrates that the MajorsPTI has a high degree of internal consistency and strong test-retest stability over a 30-day interval. Compared to other personality measures of similar length the MajorsPTI is very reliable. It is important to keep in mind that the number of items on a scale greatly influences the statistical calculation of reliability (the more items the higher the alpha reliability). The internal consistency that is reported here is similar to other measures of nearly twice the length as the 52-item MajorsPTI.