Assessment (Normal personality and Abnormal Personality)

Assessment (Normal personality and Abnormal Personality)

Instruments exist to assess whether persons are clinically problematic in one or another combination of polarity extremes. Best known among these is the MCMI (Millon Clinical Multiaxial Inventory, either versions I, II, or III) or the forthcoming MCMI-III; 2nd edition, 2006 (see Grossman & del Rio, 2005). Recent instruments geared to the theoretical model, but focused essentially on nonclinical or normal populations, have also been published. The Personality Adjective Check List (PACL) (Strack, 1987) was the fi rst and most direct form of these tools. More recent is the Millon Index of Personality Styles (MIPS-R), developed by the authors and their associates (Millon, Davis, & Weiss, 1993; Millon, Millon, & Weiss, 2003).
Abnormally Oriented Clinical Instruments
The initial tools developed on the basis of the evolutionary model were oriented primarily
to assess abnormal personality, termed disorders in the DSM, although other clinical syndromes were appraised by these instruments as well. Three initial diagnostic inventories instruments were constructed and published, the MCMI (Millon Clinical Multiaxial Inventory), the MBHI (Millon Behavioral Health Inventory), updated a short time ago as the MBMD (Millon Behavioral Medicine Diagnositc) and the MAPI (Millon Adolescent Personality Inventory), replaced in recent years with the MACI (Millon Adolescent Clinical Inventory).

MCMI
A 175-item true–false self-report inventory, the MCMI and its subsequent revisions, MCMI-II (Millon, 1987), MCMI-III (Millon, 1994), and MCMI-III, 2nd edition, 2006, (Millon, in press) include 14 abnormal personality scales [all of the personality disorders included in the main texts and appendices of the DSM-III, III-R, and IV; American Psychiatric Association (APA), 1980, 1987, 1994)], nine clinical syndrome scales, as well as three “modifying indices” to appraise problematic response tendencies. Within the restrictions on validity set by the limits of the self-report mode, the narrow frontiers of psychometric technology, as well as the slender range of consensually shared diagnostic knowledge, all steps were taken to maximize the MCMIs concordance with its generative theory and the offi cial classifi cation system. Pragmatic and philosophical compromises were made where valued objectives could not be simultaneously achieved (e.g., instrument brevity versus item independence; representative national patient norms versus local base rate specifi city; theoretical criterion considerations versus empirical data).A major goal in constructing the MCMI was to keep the total number of items comprising the inventory small enough to encourage use in all types of diagnostic and treatment settings, yet large enough to permit the assessment of a wide range of clinically relevant behaviors. At 175 items, the fi nal form is much shorter than comparable instruments. Potentially objectionable statements were screened out, and terminology was geared to an eighth-grade reading level. As a result, the great majority of patients can complete the MCMI in 20–30 minutes. Unfortunately, as many have noted (Butcher, 1972), assessment techniques and personality theorizing have developed almost independently. As a result, few diagnostic measures either have been based on or have evolved from clinical theory. The MCMI is different. Each of its personality disorder and clinical syndrome scales was constructed as an operational measure of a syndrome derived from a theory of personality and psychopathology (Millon, 1969, 1981, 1990).
No less important than its link to theory is an instrument’s coordination with the offi cial diagnostic system and its syndromal categories. With the advent of the various recent DSMs (APA, 1980, 1987, 1994), diagnostic categories and labels have been precisely
specified and defined operationally. Few diagnostic instruments currently available are as consonant with the nosological format and conceptual terminology of this official system as the MCMI. Separate scales of the MCMI have been constructed in line with the DSM to distinguish the more enduring personality characteristics of patients (Axis II) from the acute clinical disorders they display (Axis I), a distinction judged to be of considerable use by both test developers and clinicians (Dahlstrom, 1972). This distinction should enable the clinician to separate those syndrome features of psychopathological functioning that are persistent and pervasive from those that are transient or circumscribed. Moreover, profiles based on all 23 clinical scales illuminate the interplay between long-standing characterological patterns and the distinctive clinical symptomatology a patient manifests under psychic stress.
Similarly, it seemed useful to construct scales that distinguish syndromes in terms of their levels of psychopathological severity. For example, the premorbid personality pattern of a patient is assessed independently of its degree of pathology. To achieve this in the recently published MCMI-III (see Figures 1.1–1.3), separate scales are used to determine the style of traits comprising the basic personality structure (Scales 1-8B) and the greater level of pathology of that structure (Scales S, B M1, M2 and P). In like manner, moderately severe clinical syndromes (Scales A, H, D, B, and T), notably those referred to traditionally as “neurotic,” are separated and independently assessed from those with parallel features but more of a so-called “psychotic” nature (Scales SS, CC, and PP).
Worthy of note in the just published MCMI-III, 2nd edtition, is the inclusion of facet scales designed to identify and measure several of the functional and structural personological domains cross-validation data gathered with nondevelopment samples supported the measure’s generalizability, dependability, and accuracy of diagnostic scale cutting lines and profi le interpretations. Large and diverse samples have been studied with the MCMI, but it is still necessary to achieve full domain coverage in ongoing cross-validation studies. Local base rates and cutting lines may continue to be developed for special settings. Nevertheless, validation data with a variety of populations (e.g., outpatients andinpatients; alcohol and drug centers) suggest that the various forms of the CMI can be used with a reasonable level of confi dence in most clinical settings. The MCMI is not a general personality instrument to be used for normal populations or for purposes other than diagnostic screening or clinical assessment. Hence, it contrasts with other, more broadly applied, inventories whose presumed utility for diverse populations is often highly questionable. Normative data and transformation scores for the MCMI are based on presumed clinical samples and are applicable therefore only to persons who evince psychological symptoms or are engaged in a program of professional psychotherapy or psychodiagnostic evaluation. As should be noted, there are distinct boundaries to the accuracy of the self-report method of clinical data collection; by no means is it a perfect data source. The inherent psychometric limits of the tools, the tendency of similar patients to interpret questions differently, the effect of current affective states on trait measures, the effort of patients to effect certain false appearances and impression, all narrow the upper boundaries of this method’s potential accuracy. However, by constructing a self-report instrument in line with accepted techniques of validation (Loevinger, 1957), an inventory should begin to approach these upper boundaries.

MBMD
In a manner similar to the various MCMI forms, these instruments are best employed with persons also being seen in a clinical or medical setting. In this regard, they differ from the PACL and the MIPS-R, to be discussed shortly. Let us briefl y describe the Millon Behavioral Medicine Diagnostic (MBMD:Millon, Antoni, Millon, Meagher, & Grossman, 2001) fi rst. MBMD. Using psychiatrically oriented psychological tests in settings primarily of a medical nature requires that their concepts and indices be translated to fi t new populations and purposes. Medical populations are not psychiatric populations, and viewing patients within traditional mental health constructs may prove neither valid nor useful. Of course, standard techniques can provide general information, such as levels of emotional health, or the presence of distinctive symptoms, such as depression or anxiety. Problems arise, however, because of the unsuitability of norms, the questionable relevance of clinical signs, and the consequent inapplicability of interpretations. In brief, a “standard” interpretation of results obtained with a medical sample on a diagnostic test developed on and designed to assess a psychiatric population may not characterize sound test use. MBHI. Developed in 1979, the MBHI was replaced by the Millon Behavior Medicine Diagnostic (MBMD) in 2001. Both were developed specifi cally with physically ill patients and medical–behavioral decision-making issues in mind. Brevity, clarity, and ease of administration were added to the goal of elucidating salient and relevant dimensions of functioning. The MBMD includes a series of negative health habit indicators (e.g., smoking) and several psychiatric indicators (e.g., depression). Also, 11 basic coping styles scales are included, each derived from the evolutionary theory of personality (1969, 1974, 1981, 1990, 1996). Other scales were developed additionally to refl ect stress moderators found in the research literature to be signifi cant precipitators or exacerbators of physical illness. A fi nal group of treatment prognostic scales were derived either to appraise the extent towhich emotional factors are likely to complicate therapeutic efforts and to predict psychological complications associated with such efforts. All items were selected with data comparing groups of general medical populations or differentiating among subgroups of patients. This shift to a general medical reference population, rather than the usual psychiatric comparison groups, was expected to optimize the discrimination efficiency of the scales; the assumption that these steps should heighten diagnostic accuracy was supported by cross-validation evidence. MAPI, MACI, and M-PACI. A variety of psychological tests have been developed through the years for use with adolescents and preadolescents. Often constructed in accord with the sophistication then available, their shortcomings are now evident as psychometric advances have taken place. Although these older instruments provide useful information, they tend not to be tuned to current issues and behaviors, and lack a theoretically grounded system of personality traits that can integrate the diverse features salient to the teen years. The Millon Adolescent Personality Inventory (Millon,
Green, & Meagher, 1982), also known as the MAPI, was normed on both abnormal (seen in clinical settings) and normal (seen in school and church settings) teenagers. The Millon Adolescent Clinical Inventory (Millon, Millon, & Davis, 1993) was based on 13–19-year old abnormal adolescents only, and developed to quantify relationships between traits and clinical states. The M-PACI or Millon PreAdolescent Clinical Inventory was designed as a self-report tool for youngsters between 9 and 12 years of age (Millon, Tringone, Millon & Grossman, 2005). The items that comprise these inventories were drafted in a language that teenagers and preteenagers use; they deal with matters teens can understand and fi nd relevant to their concerns and experiences. The most elegantly constructed psychometric tool is not likely to be widely accepted if its content, length, and linguistic style make it unwieldy. A major goal for the MAPI, MACI, and M-PACI was to construct inventories with enough items to assess and illuminate accurately a variety of personality traits, psychological concerns, and clinically relevant behaviors, yet be of suffi cient brevity to encourage its use in a variety of outpatient and residential settings. Both reading level and vocabulary were set to allow for ready comprehension by the vast majority of youngsters. The fi nal 160-item MACI inventory, geared to the sixth-grade reading level and the 97-item M-PACI test, geared to the third-grade level, can be completed by most youngsters in less than 20 minutes. The brevity and clarity of the instrument facilitates rapid administration with a minimum of client resistance.
Counselors, clinical psychologists, and psychiatrists were involved with the M-PACI and MACI throughout all phases of their development. At an early stage, clinicians were interviewed to ascertain issues relevant to both typical and troubled adolescents. The M-PACI has seven “emerging personality” scales and seven “clinical syndrome” scales. The MACI includes 12 personality scales that seek to parallel the DSM-IV disorders. It also possesses eight “expressed concern” scales that address the attitudes teenagers have regarding signifi cant developmental problems. In addition, there are seven clinical syndrome scales that refl ect major diagnostic categories associated with behaviors and thoughts that may pose serious diffi culties for the adolescent (e.g., suicide ideation, substance abuse). The capacity to differentiate each of the various clinical problem areas is a key to the effectiveness of the inventory. Hence, MACI item selections were made by comparing the targeted criterion group with a general, but troubled adolescent population. The use of reference groups such as these should substantially increase diagnostic discrimination efficiency. The M-PACI items were selected following substantial evidence of their correlation with clinician judgments of psychologists well-acquainted with their psychological problems.

Normality-Oriented Personality Instruments
As noted, the fi rst group of instruments generated by the theoretical model focused on abnormality and clinical diagnosis. As the theory was broadened to encompass personality traits and characteristics that fell within the so-called normal range (Millon, 1991), attention was directed to the development of suitability coordinated nonclinical assessment tools. Two such instruments have been constructed in line with this goal— the PACL (Strack, 1987), and the Millon Index of Personality Styles and its revision (Millon, 2004; Millon, Millon, & Davis, 1993). PACL. The fi rst instrument specifi cally designed to assess the theoretically derived personality types in a normal population was developed by a graduate research group led by the senior author in the early 1980s. As the project progressed through its early revisions, the major responsibility for its further development was undertaken by one of its members, Stephen Strack. Using the adjective checklist format, the initial forms (both clinician-rated and self-reported) were composed of 405 items that were intended to refl ect the theory’s original basic personality types and its three more severe personality variants.
Item refi nements and initial validation studies were based on data from over 2000 normal adults from a variety of diverse national settings. A wide range of validity data have been gathered and reported (Strack, 1987, 1991), including correlations with various
other gauges of personality and biographic data on current and past behavior. Each scale of the PACL appears consonant with theoretical expectations and their expected personality characterizations. Thus, the PACL Inhibited scale (which refl ects a milder variant of the theory’s active-detached pattern and the DSM’s avoidant personality disorder) is positively correlated with measures of shyness, submissiveness, and social anxiety, although being negatively correlated with measure of sociability, dominance, and emotional well-being. Similarly, the PACL Forceful scale (a milder version of the theory’s and DSM’s antisocial and sadistic disorders) is positively related to gauges of aggressiveness, arrogance, dominance, and negatively related to gauges of deference, submissiveness, and conscientiousness. Owing to the common linkage in theory there is a possibility that the results of the PACL might be erroneously confused with parallel revisions of the MCMI. High scores on the PACL scales do not signify the presence of personality disorders, as they do on the MCMI; rather, they suggest the presence of distinctive personality traits. As Strack (1991) notes, all persons in the PACL development group were presumed to have normal personalities, hence the special utility of the instrument as a measure of normality rather than abnormality.

MIPS-R.
This is a self-report inventory composed of sentence-length items rather than an adjective checklist, as is the PACL. Both are anchored to the theoretical modelformulated by Millon from the mid-1960s to the early 2000s. However, the PACL, as with the MCMI, focused its scales on personality types as a composite or whole, be they normal (PACL) or abnormal (MCMI). In contrast, the MIPS-R focuses on the constructs that underlie these personality types, the latent components that combine to give rise to them. To illustrate: The histrionic personality disorder (MCMI), termed “sociable” personality style (PACL), are the manifest abnormal and normal personality forms that take shape among those who are latently “active” on the active–passive polarity, and oriented to other on the self–other polarity. Focusing on these latent components, the first set of scales of the MIPS measure the three polarity pairs of the theory directly (e.g. pleasure–pain, active–passive, and self–other), rather than the manifest forms into which various combinations of these are exhibited.
Beyond breaking down the theory’s manifest personality types into their constituent latent constructs, the theory, as described in previous pages, has been expanded substantially. Whereas the three polarities of the theory are still considered crucial elements of the model and serve as a particularly important gauge of personality abnormality, they are now judged to be insuffi cient as a comprehensive scaffold for encompassing the highly diverse styles of normal personality. This is not the chapter to elaborate both the full rationale and specifi cs of the expanded model; recent essays on this theme may be found in the MIPS or MIPS-R manuals (Millon, Millon, Davis, & Weiss, 1993; Millon, Millon, & Weiss, 2003).
Briefl y, we should note that cognitive differences among individuals and the manner in which they are expressed have not been a suffi ciently appreciated domain for generating personality traits. We have added a set of four polarities that refl ect different “thinking styles” to the MIPS and MIPS-R. These follow the initial three polarities (e.g., self-other), which have been termed as “motivating styles.” Similarly, we have added a third domain of polarities to those of “motivation” and “cognition,” that are termed as “behaving styles.” We share the view of many who give the manifest forms of the interpersonal dimension a degree of centrality in their personality gauges; in fact, we do judge them no less signifi cant than either motivation or thinking styles, especially if they
are organized in terms of the latent polarities they express. Thus, a third domain, comprising fi ve interpersonal behavior polarities, concludes the MIPS-R test form, although work has begun on the inclusion of “character styles” to further advance the utility of the instrument in identifying positive and healthy orientations.
The following précis of the tripartite structure of the MIPS scales divides the test in the manner in which organisms function in their environment, one which we believe may be a useful theory-based schema for purposes of normal personological analysis. As noted previously, we have termed the fi rst segment in this tripartite sequence as motivating styles, to signify that the behaviors of organisms are prompted, energized, and directed by particular purposes and goals they wish to achieve. The second component of the sequence is labeled thinking styles to indicate the manner in which human organisms seek out, regulate, internalize, and transform information about their environment and themselves, a step necessary if organisms are to achieve their aims effectively. The third segment in the sequence is referred to as behaving styles to represent the different ways in which human organisms relate to and negotiate with other humans in their social environment in light of the aims that motivate them and the cognitions they have formed. To capture personality more-or-less fully we must fi nd ways to characterize all three components of the sequence: the deeper motives which orientindividuals, the characteristic sources they utilize to construct and to transform their cognitions, and the particular behaviors they have learned to relate to others interpersonally.