Psychometric Tools for Diagnosing Adult Temperament

"A plea for the study of individual differences, then, becomes a plea for the recognition of parameters."

Jan Strelau (1991) in exploring the renaissance in research on temperament identified 25 paper-and-pencil techniques for diagnosing temperament traits in adults. The traits measured include:

Inventories which are labeled as temperament diagnostic tools...even if some of the traits...can be considered as belonging to the domain of temperament.

Inventories which are aimed at measuring a single trait/domain...believed to belong to the structure of temperament or to so-called primary (biologically determined) personality characteristics considered synonymous for temperament.

Personality inventories based on the assumption that temperament and personality are interchangeable concepts.

Huelsman (1983) identified three models of cognitive, learning and teaching styles: the Multiple Bi-Modal Dimensions Model, as exemplified by Jungian theory; the Bi-Polar Dimensions Model of which the work of Herman Witkin is an example, and; Multiple Dimensions Model as exemplified by the work of Joseph Hill. These same schemes have been proposed as complementary to temperament models. Similarly, the instrments designed to measure these proposed temperament models have also been proposed for cognitive and learning preference.

Instruments designed to measure the Jungian/Multiple Bi-Polar model include Isabel Myers and Katherine Briggs' Myers-Briggs Type Indicator (MBTI), David Keirsey's Temperament Sorter (KTS) David Kolb's Learning Style Inventory (LSI), and Anthony Gregorc's Gregoric Style Delineator (GSD). Exemplary of the Bi-Polar Model is Herman Witkin's research on Field Dependent-Independent reflected in the Embeded Figures Test (EFT) and group version (GEFT), and Jerom Kagan's Matching Familiar Figures Test (MFFT) and Conceptual Style Test (CST). The Multiple-Dimensions Model includes: the works of Rita and Kenneth Dunn and Gary Price's Learning Style Instrument (LSI) for children 3-12 and the PEPS instrument for adults; Albert Canfield's Learning Style Inventory (CLSI); Anthony Grasha and Sheryl Reichmann's Student Learning Style Scales (GRSLSS); Joseph Hill's Cognitive Style Inventory (CSI), and Marie Elkin's PACE Temerament Sorter Preference Indicator.

The psychometric instruments usd as indicators of temperament preference in children are typically obser-rated and exhibit low to moderate validity and reliability. One interpretation of the literature might be that these instruments are meant more for predicting unacceptable social behavior, predicting coping abilities and responses to stressors, or as an indicator of delinquent behavior, and less for matching learner preference with teaching style.

Similarly, adult temperament measures can arguably be categorized in this same stmulus-response manner. From the earliest categorization of the four temperaments by Hippocrates and galen to the works of Kretchmer, Jung and others, temperament is viewed as choices, tendencies, or preferences between behavior polarities. Teh research in adult temperament assessment tends to associate temperament preference to some measureable preference to some measureable form of success. Several instruments identified as adult temperament inventories are deisgned for children orare modification of research originally conducted with children. The majority of temperament instruments for adults utilize a self-rated scoring of scenarios using some form of Likert scale. With the exception of the MBTI, the majority of adult inventories lack statistical validity and reliability.

The term temperament in the literature for both children and adults is frequently used interchangeably with the encompassing term psychological or personality type. Temperament instruments overwhelmingly categorize children and adults into one of four, or multiples of two, modes. These categories represent behavior polarities-mirror images. The multiplicity of type suggests that the major dimensions of temperament cannot be regarded as so many factors, each of which being capable of independent appraisal. Their combinations do not give rise to even sets of two or four but to dozens of distinctive patterns as suggested by Diamond (1957) who concludes that it is unsatisfactory to classify individuals according to his one strongest temperamental disposition for the following reasons:

1. the various temperaments are in a state of constant fluctuation, and
2. environmental stimuli may influence individual responses.

A more promising attempt to classify individuals, notes Diamond (1957), would be to select a minimum number of dimensions as a basis for description and allow these to vary independently in the individual case. Diamond adds that the most useful way to characterize the temperamental pattern of an individual is to discover this dyanmic pattern which tends to repeat itself in his experience under many diverse circumstances. This sentiment is parroted by Sheldon (1942), Thomas and Chess (1977), and other researchers who concude that an effective measure of individual temperament preference would require longitudinal studies under a variety of conditions. The literature suggests that existing temperament inventories, at best, represent a snapshot assessment of an individual for a particular time and environment.

Matching teaching style and learning styles may, as noted by Zarghani (1988), be erroneous since the methods for determining learning and teaching styles may not have sufficient research base to support their application in a learning environment. Zarghani also notes that neither learning styles nor teaching styles cluster neatly into packages and that matching may, in fact, be a disservice in not exposing students to a variety of teaching styles.

As many studies can be cited to support the advantages of matching teaching strategies as studies showing no statistical significance for creating a "goodness of fit." McCall (1986) in investigating the stability and continuity in temperament research poses the question,

"Are we really wiser after several decades of studying mental behavior? In some repects, we are; in a shocking number of other respects, we are not. And will a vagrant from another field, one unfamiliar with most of the written literature and all of the unwritten lore about temperament, not look like a kibitzing duck out of water? Undoubtedly" (p.13).

The usefulness of temperament concepts might best be summarized by Bates (1994):

"For me, temperament concepts are useful along with other kinds of concepts, inproviding a structure for description, a way of simplifying how I describe individual differences in development. If I keep this in mind, I am better able to tolerate ambiguity in the concepts and emasures of temperament...TO the extent that our measures of temperament describe meaningful phenomena they will be useful" (p.9).

The merit behind the measurement of temperament is noted by Sheldon and Stevens (1942):

"It does not take a science to tell that no two human beings are identiccally alike, but it does require the discipline of systematic inquiry to give, in terms of scales and categories, a useful description of individual differences" (p.2).>

"A plea for the study of individual differences, then, becomes a plea for the recognition of parameters..." (p.3)

"We may ask, 'What are the effects of culture (or environment or learning) upon temperament (or personality or attitudes)?' Or we may ask, 'What are the effects of different types of nurture upon various aspects of behavior for persons of differing constitutional endowment?' The importance of putting the question in this ,atter form is expressed by the old saying, 'What is one man's meat is another man's poison'" (p.3)

Finally, as Thomas and Chess (1977) note,

"Man's increased understanding of himself and his world has been a major ingredient in the history of human society. The intellectual characteristics of this history has been man's ability to increasingly remove himself from the concrete experience of the phenomenological her and now and place himself in an abstracted world of concepts and logic" (p.vii).

From the literature of published temperament, humoral, character, and psychological type theory and their related inventories, it can be seen that no one single temperament-based indicator for adults currently exists. The literature on temperament preference-as a component of the individual's cumulative tendency of perception, retention, and organization-has been associated with learning style preference. It is my theory that temperament preference is a characteristic indication of the learner's natural tendency of responding to and using stimuli in the context of learning. If temperament preference can be identified and measured by means of a paper-and-pencil questionnaire, then a matching of learner needs to the learning environment may avert a mismatching of interactions.