Tests, Measurement and the MBTI^®

Background

The decision by Isabel Myers and Katharine Briggs to construct a psychological instrment thta would practically apply their adaptation of Jung's theory of Psychological Types meant that they would be entering the world of psychometrics (literally: "measuring the mind").

Psychometrics essentially follows from the idea of the 19th Century mathematician Francis Galton. Broadly, it is the intellectual presumption that the use of mathematics and statistical methods is a valid means of measurement for attributes like intelligence or, in this case, personality. The application of these techniques became very popular in the United States. In retrospect, this seems to be because it meshed with the philosophical pragmatism of William James and, particularly John Dewey, that was so influential in Amercian thought that it could be considered to be part of the contemporary American mythos.

The notion of psychological testing is a key part of American psychology to an extent not shared by other countries, a point reinforced recently by Eduardo Casas, the translator of the MBTI^® into French, during his keynote speech at the APT X International Conference on Psychological Type in July 1993. In the American cultural context, therefore, it was logical for Katharine Briggs and Isabel Myers to venture into this aspect of the empirical scientific method.

At the time Katharine and Isabel began to work on the MBTI^®, there had already been much discussion on personality, albeit preDominantly from the behaviourist perspective (which by definition rejected the innate aspects of Jung's theory), and also from the viewpoint of the medical model, which put the person in the role of a "patient", who invariably needed to be "fixed".

Jung's concepts of Introversion and Extraversion aroused interest, but it was hard to quantify the successfully, in particular as the methodologies used showed a strong correlation between Introversion and what was defined as Neuroticism. This led to many researchers giving up the field of personality as an areas of worthwhile investigation. Notwithstanding this, the Gray-Wheelwright Test, an instrument designed to apply Jung's Psychological Types, was developed in San Francisco in the 1940's. These activities provided the broad intellectual context for the development of the MBTI^® as a psychological instrument.

The MBTI^® as a Psychological Test

Technically, MBTI^® is a particular form of psychological test. The word "Indicator", rather than "test" is used in describing the MBTI^® in order to avoid connotations of "right" or "wrong", "good" or "bad" in completing the instrument. This fits in with the MBTI^® theory and philosophy that all 16 Types are "good". It also makes the MBTI^® a test of typical performance (test of personality, interests, attitudes, values and opinions), something typical of the person at the time of the test, rather than a test of maximum performance (attainment, ability, aptitude and skills) where a personal best performance is sought (Cronbach, 1984).

Anne Anastasi, the author of a standard text on psychological testing, defines a psychological test as "essentially an objective and standardised measure of a sample of behaviour" (p.23 ; 1990). For the MBTI^® to be seen as a valid psychological test, it needs to meet these requirements.

Anastasi's statement simply means that the content of the test means the same to everyone who completes it and scores/interprets it (objectivity) and that the admimistration, scoring and interpretation of the test has a standard format, no matter who is engaged in the activity. In terms of the latter, the MBTI^® Accreditation Programme is designed to facilitate a standard format of administration, scoring and interpretation by accredited users. Reliability and validity are perhaps the most important components of establishing a test's objectivity and there are statistical ways of measuring these categories. Reliability means consistency in results if the test is taken again by the same person; Validity relates to whether the test measures what it purports to measure, in the MBTI^®'s case a questionnaire about personality, an application of Jung's theory of psychological Types.

Statistically, reliability can be measured by:

• Test/Retest: Where the test is given again to the same people over a particular time frame, and the differences are measured. (will it do it again)
• Parallel Forms: Similar to the above except that diferent forms of the same test are used;
• Internal Consistency: Applied to one administration of a test A Split-half technique is used, where the test is divided into two equal parts.The purpose here is to see whether answers to the test are consistent and not contradictory.
• Inter-Scorer Reliability: This is used for tests that have subjective methods of scoring and where the scoring of the test may vary between individual scorers.

Establishing the Validity of a psychological test is limited in that it relates to the stated purpose of the test. This is essentially the gathering of evidence that the test scores actually support the propositions asserted by the test in a meaningful, useful and appropriate way.

Validity can be measured by:

• Face Validity: Where items in the test are read to give an idea of what it appears to measure.
• Content Validity: Whether the sample items represent what the test is meant to measure.
• Criterion-Related Validity: Usually for test without a theory or other base behind their structure but can also be used with theory-based tests like the MBTI^® as a comparison with similar tests.
• Construct Validity: Evidence of how well a theoretical construct or trait is measured; In MBTI^® terms, whether the information the instrument gives is consistent with the theory behind it.

The MBTI^® Manual explains the relevance and application of these statistical concepts to the MBTI^®.

Photograph courtesy of Jamie Johnston, CAPT Library. PETER GEYER (INTP) is a consultant, researcher and writer in the field of C G Jung's theory of psychological types. He conducts MBTI Accreditation programs and presents internationally on a regular basis. Peter is a life member of AusAPT and a professional affiliate of the Australian Psychological Society.