Read an Excerpt

Comprehensive Handbook of Psychological Assessment, Volume 2, Personality Assessment

John Wiley & Sons

Chapter One

DANIEL L. SEGAL AND FREDERICK L. COOLIDGE

HISTORICAL OVERVIEW 3
RANGE OF TESTS 4
THEORETICAL CONSIDERATIONS 5
Theoretical Bases 5
Empirical Bases 5
Diagnostic Bases 5
PSYCHOMETRIC CONCERNS 6
Standardization 6
Reliability 7
Validity 7
RANGE OF POPULATIONS 9
CROSS-CULTURAL FACTORS 10
ACCOMMODATION FOR POPULATIONS WITH
DISABILITIES 10
LEGAL AND ETHICAL ISSUES 11
COMPUTERIZATION 11
FUTURE PERSPECTIVES 12
REFERENCES 12

Psychological assessment pervades nearly every aspect of
clinical and research work in the broad area of mental
health. In general, psychological assessment techniques are
designed to evaluate a person's cognitive, emotional, behavioral,
and social functioning. One specific group of tests,
called personality tests, strives to uncover the structure and
features of one's personality, or one's characteristic way of
thinking, feeling, and behaving. Another group of tests is
designed to measure signs and symptoms of psychopathology
or psychiatric disorders. Tests of personality andpsychopathology
can be further subdivided into two specific
types: objective and projective. Objective tests include standardized,
clear, specific items and questions that are presented
to the respondent, as well as a limited choice of
responses (e.g., choosing "yes" or "no" to a test item). In
contrast, projective tests present novel or ambiguous stimuli
and include an open-ended response format, such as a story
from the respondent (an overview of projective tests is presented
in Chapter 23). In this chapter, we discuss the major
issues concerning the objective assessment of personality and
psychopathology, including analysis of theoretical issues in
test development, range of tests, cross-cultural factors, ethical
and legal concerns, and the status of computerization of objective
tests. We begin with a synopsis of the history of objective
tests.

HISTORICAL OVERVIEW

Objective tests of personality and psychopathology received
their first important recognition during World War I. With the
immediate and sudden burden of large numbers of recruits,
the U.S. armed services were in dire need of a means of
assessing the capabilities of recruits quickly and efficiently
and required a classification system for making determinations
of who was mentally fit for service and who was not.
The American Psychological Association volunteered its services
and developed (with psychologist Lewis Terman, who
developed the Stanford-Binet Intelligence test) the Army Alpha
(verbal) and Army Beta (nonverbal) tests of intelligence
for literate and nonliterate recruits, respectively. At the same
time, American psychologist Robert S. Woodworth at Columbia
University was developing a paper-and-pencil test of
psychiatric fitness for the armed services, called the Personal
Data Sheet (Woodworth, 1920).

The Personal Data Sheet became one of the first personality
tests ever to be administered on a large basis. Woodworth, an
experimental psychologist by training, had designed his test
to detect Army recruits who might be vulnerable to emotional
breakdowns during combat. He first created 200 questions
based on neurotic symptoms described in the literature and
on symptoms common to soldiers who had emotional and
behavioral problems in the service. The questions covered
abnormal fears, excessive anxiety, depression, impulse problems,
sleepwalking, nightmares, memory problems, hypochondriasis,
compulsions, shyness, and depersonalization. In
the final version of the Personal Data Sheet, items were included
only if the symptoms occurred twice or more frequently
in a group of neurotics (according to prior psychiatric
diagnosis) compared to purportedly normal people. The original
200 test items were subsequently reduced to 116 "yes-no"
items. The test yielded a single score, which Woodworth
considered a measure of psychoneurosis. One innovative feature
of the test was that it was based on norms, including
education, ethnicity, and clinical versus normal samples. The
average new recruit scored about 10 (10 positive psychoneurotic
symptoms out of 116). Those who were deemed unfit
for service generally had scores of about 30 or 40. By the
time Woodworth had finished the final improvements on the
Personal Data Sheet, it was too late in the war to use the test
to screen recruits. The test later became known as the
Woodworth Psychoneurotic Questionnaire, and it became the
forerunner for later personality inventories.

One other interesting and innovative personality test
proposed during this same period was the Cross-Out Test
(Pressey & Pressey, 1919). Also known as the X-O Test, respondents
were given lists of pleasant and unpleasant words.
They were instructed to cross out or make an X over words
they considered wrong, unpleasant, inappropriate, or worrisome.
The Presseys believed that the resulting pattern could
help categorize emotional states. They also emphasized the
test could be administered in a group format.

The first commercially sold test that yielded more than one
score was the Bernreuter Personality Inventory (Bernreuter,
1933), which consisted of 125 items answered in a "yes,"
"no," or "?" format and was also based on normative samples.
The Bernreuter yielded six subscales: Neurotic Tendency,
Self-Sufficiency, Introversion-Extraversion, Dominance-Submission,
Sociability, and Confidence. The test became and
remained popular in the first half of the twentieth century and
was actually still commercially available (with 1938 norms!)
into the 1990s (see Aiken, 1989). Interestingly, the Personal
Data Sheet and the Cross-Out Test have some features that
heralded some of the current objective tests of psychopathology
such as the Minnesota Multiphasic Personality Inventory
(MMPI) and the Symptom Checklist-90. As noted
earlier, the scoring of these early tests was based on published
literature, patient interviews, and intuition. As such, the items
could be said to be logically keyed (i.e., the test makers used
their subjective judgment based on the origin of the items
and an item's face validity to decide what answers were
pathological or not). However, none of these early tests were
applied widely in the clinical setting.

Notably, the numerous challenges associated with defining
and measuring personality came into much sharper focus
in the 1930s with the publication of two famous books on
the subject. Gordon Allport's Personality (1937) and Henry
Murray's Explorations in Personality (1938) analyzed the
topic from different perspectives, but collectively, they focused
the field on the measurement of individuality and
personality and paved the way for more sophisticated measurement
of the constructs. Around that same time, two clinicians
associated with the University of Minnesota Hospital
began work on the most widely employed test in the history
of objective testing, the MMPI. Starke R. Hathaway, a psychologist,
and J. Charnley McKinley, a psychiatrist, wanted
to provide a more efficient way, other than a one-on-one clinical
interview, of obtaining a psychological diagnosis. Like
Woodworth and the Presseys, they wanted to create a pencil-and-paper
objective test of psychopathology that could be
group administered. However, one unique feature of the
MMPI was that it was not to be logically keyed but empirically
keyed. The problem with logical keying for Hathaway
and McKinley was that the items could be too easily faked
or manipulated by the test takers. Instead, Hathaway and
McKinley chose to use empirical keying where items were
grouped on the empirical basis of their ability to differentiate
between known psychiatric and normal groups. The full history
and nature of the MMPI will be dealt with in Chapter 3,
but the creation of the MMPI set the standard for innovative
and empirical objective test development that has persisted
to the present day.

RANGE OF TESTS

The type and nature of objective tests is astoundingly diverse.
It is safe to say that an objective test has been developed to
evaluate all of the major psychiatric disorders, most of the
relatively uncommon disorders, and almost all of the major
constructs that are relevant in clinical psychology. Major distinctions
among tests are whether the test is designed for
children or adults as the respondent group and whether the
test evaluates mental illness (psychopathology) or normal-range
personality traits. Yet another distinction is whether the
test focuses on a single construct or disorder of interest (e.g.,
potential for child abuse, depression, or anxiety) or on multiple
constructs or disorders (e.g., 10 clinical scales are included
in the MMPI-2). The final types of objective tests
included in this volume are structured and semistructured interviews.
Although they are not classically defined as objective
tests, they are objective tests from the standpoint that the
questions are clear, standardized, and presented in a specified
order, and responses are coded in a specified way. Notably,
tests in each of the categories described here are well represented
in this volume.

THEORETICAL CONSIDERATIONS

Although the referral or research question is perhaps the most
important reason for the selection of an objective psychological
test, it is also important to note that objective tests vary
considerably as to their theoretical bases for construction, and
this basis may also aid in the selection process. There are
three broad methods by which tests are constructed: theoretical,
empirical, and diagnostic. It should be noted at the
outset that these methods overlap, and it could be argued that
no objective test completely lacks a theoretical basis and no
objective test can be judged sufficiently reliable and valid
without strong empirical methods. Yet, as will be shown by
the following examples, objective tests may be driven by one
method more than another.

Theoretical Bases

All objective tests are constructed on some theoretical basis.
A test maker must have some prior conception of what a test
is designed to measure, and test items are initially picked
based on some theoretical relationship the test item has to the
construct being measured. However, some tests are more
tightly linked to a particular theory or theoretician, and other
tests have been created with a more general purpose in mind.
An exceptional example of a theoretically driven test is the
Millon Clinical Multiaxial Inventory-III (MCMI-III; see
Chapter 9), which is based on Theodore Millon's innovative
and comprehensive theory of personality disorders. Another
good example is the Child Abuse Potential Inventory (see
Chapter 19), which is derived from psychological factors
noted in the literature that are theoretically related to child
physical abuse. A final example is that of a recent operationalization
of Karen Horney's tridimensional interpersonal theory
of personality that postulates three basic personality
styles: Moving Towards People, Moving Against People, and
Moving Away From People. Coolidge, Moor, Yamazaki,
Stewart, and Segal (2001) recently created a new test, called
the Horney-Coolidge Type Indicator, that is based on Horney's
theory and has demonstrated the usefulness of her three dimensions
in the prediction and understanding of modern personality
disorder features.

Empirical Bases

Empirical models, although having some theoretical basis,
are usually driven by their statistical methods or procedures
and they frequently use factor analyses. Factor analyses involve
the testing of large groups of participants. In the initial
stage of a factor analysis, a correlation matrix is examined
between every item on a test with every other item on the
test. The second stage is the identification of clusters of related
items. The goal of a factor analysis is usually to reduce
the number of items on a test to only its nonredundant items
or to identify the underlying factor structure of a test. Empirical
models also frequently employ discriminant studies
where particular traits are demonstrated statistically to pertain
more to one identified group than another (e.g., 8-year-olds
as opposed to 12-year-olds, or males as opposed to females).

For example, in the 1940s, psychologist Raymond B.
Cattell sought to understand the basic building blocks of personality
by studying and cataloging all of the words in language
that describe personality features. Based on several
decades of research and factor-analytic techniques, the Sixteen
Personality Factor (16PF) Questionnaire (see Chapter 4)
was created. In a similar vein, Tupes and Christal (1961), in
a review of thousands of English words describing personality
traits, theorized through factor analyses that personality
traits could be summarized by as few as five factors. Later,
Costa and McCrae (1985) created a famous test of the five-factor
model and claimed that it could be extended to abnormal
personality traits as well. Thus, 5-factor models and
16-factor models are initially driven by only a general theoretical
framework (i.e., do 5 factors underlie personality trait
descriptions?). The subsequent creation of an objective test
of 5 factors or 16 factors is largely empirically and lexically
driven; that is, the authors were concerned only with what
the factor analyses (empirically driven) revealed regarding the
relationships among the words (lexically driven).

Diagnostic Bases

The main purpose of a diagnostically based objective test is
to produce a psychiatric diagnosis. The Beck Depression Inventory
(see Chapter 5) was created to measure the severity
of depressive symptoms that the test taker is experiencing
(e.g., mild, moderate, or severe levels of symptoms). It was
not created to be an "official" diagnostic measure of depression,
although it may be useful in that endeavor. In order to
become an "official" diagnostic objective test, a measure
should be aligned with a current diagnostic system like the
Diagnostic and Statistical Manual of Mental Disorders (4th
ed., text revision; DSM-IV-TR) published by the American
Psychiatric Association (2000) or the International Statistical
Classification of Diseases and Related Health Problems
(10th ed.; ICD-10) published by the World Health Organization
(1992). The MMPI could certainly be considered a
diagnostic test, although it is also famous for its innovative
empirical underpinnings. It has been the most widely used
objective diagnostic test for the past 55 years; however, it is
not diagnostically aligned with either the DSM or ICD. The
original Millon Clinical Multiaxial Inventory was created to
diagnose personality disorders, and its most recent version is
aligned closely with the DSM-IV. Certainly, all of the structured
and semistructured diagnostic interviews are designed
specifically to aid in psychiatric diagnosis.

Continues...

---

Excerpted from Comprehensive Handbook of Psychological Assessment, Volume 2, Personality Assessment

Copyright © 2003 by Mark J. Hilsenroth, Daniel L. Segal, Michel Hersen.
Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---