Structured Personality Tests

Some Definitions
Personality
the relatively stable and distinctive patterns of behavior that characterize an individual and his or her reactions to the environment.
Personality tests attempt to measure personality traits, states, types, and other aspects of personality (such as self concept). Some Definitions
Personality Traits
relatively enduring dispositions
tendencies to act, think, or feel in a certain manner under any given circumstance
distinguish one person from another
Personality States
predominantly emotional reactions that vary from one situation to another. 1 Some Definitions
Personality Types
refer to general descriptions of people
e.g., avoidant, depressive
Self-Concept
a person’s self-definition; an organized set of assumptions one has about him or herself.
Most structured personality tests attempt to assign a personality type based upon measurements of someone’s personality traits. Strategies to Structured Tests
In general, subject is asked to respond to an objective, written statement that is designed to minimize ambiguity.
Different from projective tests, where subjects respond to purposely ambiguous stimuli.
Two approaches to structured tests
Deductive
Empirical Strategies to Structured Tests Deductive Logical Content Content Approach Intuitive Approach Rational Approach Theoretical Empirical Criterion Group Contrasted-Group External strategy Empirical strategy Criterion-keying Factor Analytic 2 Deductive Strategies
Two types
Logical-content

Uses reason and deductive logic in the development of personality measures. Test designer attempts to logically deduce the type of question that should be asked to measure the hypothetical concept.

e.g.,
I frequently worry about my weight. T/F
I feel bad after I’ve eaten a good meal. T/F Relies heavily on face validity. Deductive Strategies
Two types
Theoretical


Start with a theory Ask questions that are consistent with the theory. Assume that every item in a scale is related to a characteristic that you are measuring.

Attempt to create homogeneous scales Frequently use item analyses to confirm Empirical Strategies
Criterion-group strategy
start with a group of people who share a common characteristic (e.g., aggressiveness, depression)
select and administer a group of items to everybody in the criterion-group and a control group
choose those items that distinguish between the criterion and control groups; which items best contrast the groups 3 Empirical Strategies
Factor Analytic Strategy
uses factor analysis to derive empirically the basic dimensions of personality
asks a large number of questions
looks for correlations among questions
if groups of questions correlate with each other, this is evidence of an underlying latent factor Logical Content Tests
Woodworth Personal Data Sheet
The first personality inventory
Based on faulty assumption that responses can be taken on face value
Produced a single score Logical Content Tests
First multidimensional scales
Bell Adjustment Inventory
Assessed adjustment in different areas of life (e.g. home life, social life, emotional functioning)
Bernreuter Personality Inventory
Items pertaining to six personality traits (e.g. introversion, confidence, sociability)
Mooney Problem Checklist (1950)
One of few still in use
Much like the Woodworth 4 Criterion-Group Tests
Minnesota Multiphasic Personality Inventory (MMPI; originally developed in 1943)
More detail to come
California Psychological Inventory (CPI; originally developed in the late ‘50s; circa ‘56/ ‘57)
Originally developed to identify personality traits of normally adjusted individuals
Uses criterion groups for some of the subscales
Compared (men and women, homosexual men and heterosexual men)
Produces personality continuums (e.g. intro-extroverted, conventional vs. unconventional, etc.) The MMPI
MMPI: Minnesota Multiphasic Personality Inventory (MMPI, MMPI-2, MMPI-A)
MMPI: original MMPI 1943
MMPI - 2: first revision in 1989
MMPI - A: adolescent form
Purpose - to distinguish “normals” from “abnormals” The MMPI
MMPI - 2 requires an 8th grade reading level
original MMPI required a 6th grade reading level (!)

Reading skills since 1943 have dropped. 8th grade 1989 reading level = 6th grade 1943 level
average reading difficulty for the MMPI-2 items is approximately 5th grade although
90% of the items require less than a 9th grade education.
Auditory or interview forms are available. 5 The MMPI
Original MMPI began with a pool of 1000 questions drawn from case histories, psychological reports, textbooks, and existing tests of
Starke Hathaway - medical psychologist
Jovian McKinley - neuropsychiatrist
both of the University of Minnesota Hospitals
Narrowed pool down to 504 items thought to be relatively independent. The MMPI
These 504 questions were administered to 800 patients representing the following psychopathologies:
Hypochondriacs: patients who are overly oncerned with bodily symptoms and express conflicts through bodily (somatic) symptoms.
Depressives: patients with depressed mood, loss of appetite, anhedonia, suicidality
Psychopathic Deviates: patients who are antisocial and rebellious & exploit others without remorse or anxiety. The MMPI
These 504 questions were administered to 800 patients representing the following psychopathologies:
Paranoids: patients who show extreme suspicions and delusions
Psychasthenics: patients plagued by excessive self- doubts, obsessive thoughts, anxiety, and low energy
Schizophrenics: patients who are disorganized, highly disturbed, out of contact with reality, hallucinating, and have poor relatedness skills 6 The MMPI
These 504 questions were administered to 800 patients representing the following psychopathologies:
Hypomanics: patients who are in a highenergy, agitated state with poor impulse control, inability to sleep, and poor judgment.
In addition, 700 controls - visitors and relatives of patients at the University of Minnesota Hospital - were also administered the same 504 questions. The MMPI
After an item analysis, items that separated the patients from the non-patients were included on one or more of the clinical scales.
The items were then cross-validated on independent samples of the criterion and control groups (administered again).
Those items that discriminated between the two groups significantly were retained. The MMPI
In addition to the eight clinical scales, two additional scales were later added:
Masculinity-Femininity (MF): containing items differentially endorsed by men and women.
Social Introversion (Si): measures introversion and extraversion. 7 The MMPI
Finally, three validity scales were also included, to address concerns regarding the logical-content approach.
Lie (L)
Infrequency (F)
Defensiveness (K) The MMPI
Lie (L)
designed to detect individuals who attempt to present themselves in a favorable light (fake good).
Not empirically constructed but logically constructed.
Infrequency (F)
designed to detect individuals who attempt to present themselves in an unfavorable light (fake bad) The MMPI
Defensiveness (K)
Measured test-taking attitude
Designed to detect those who were trying to “fake normal”
To create the K scale, Hathaway & McKinley examined protocols of disturbed individuals who produced normal MMPI patterns.
Those items “left over” that differentiated between the two groups were included in the K scale. 8 MMPI Scales 1. 2. 3. 4. 5. 6. 7. 8. 9. 0. A B C Hypochondriasis Depression Hysteria Psychopathic Deviant Masculinity-Femininity Paranoia Psychasthenia Schizophrenia Hypomania Social Introversion Lie Infrequency Defensiveness (Hs) (D) (Hv) (Pd) (Mf) (Pa) (Pt) (Sz) (Ma) (Si) (L) (F) (K) Scoring the MMPI
Questions that contribute to each scale are added up to obtain raw scores.
Raw scores are then converted to McCall’s T scores (mean 50, SD 10) based upon scores from the control group.
Scores above T=65 are considered clinically significant. Scoring the MMPI
Original goal - a single “spike” in a patients scores will lead to a clinical diagnosis.
E.g Schizophrenics would show a spike on the schizophenia scale
E.g. histerics would show an elevation on the hysteria scale
This assumption turned out to be false in that in reality most profiles produced multiple “spikes”
Sometimes a person would show elevated levels on all of the scales (????) 9 Interpreting the MMPI
Single scale interpretation is therefore generally not possible.
Configural Interpretation (pattern analysis)
Meehl (1951): Two-Point Code
Started research looking for common characteristics of individual profiles with common two highest T-score scales.
General strategy: if there is a defined “spike”, interpret it first. Then look for two-point codes. MMPI-2
The MMPI was re-standardized in 1982 and what resulted was the 2nd version
Purpose
Revise the norms MMPI Psychometrics
Median split-half reliability coefficients run in the .70s, some as high as .96, most lower.
Median test-retest reliability coefficients range between .50 and .90 (median .80s).
Factor analytic coefficients running in the high .90s.
Reliability is generally considered adequate for a psychological measure. 10 MMPI Psychometrics
Tens of thousands of validity studies point to diagnostic specificity for a range of problems including
substance use, alcoholism, post-traumatic stress, delinquent behaviors.
Probably the most widely studied personality test world wide. Factor Analytic Strategy
Cattell’s 16PF
Began with all adjectives applicable to humans
4504 “real” traits (Allport and Odbert, 1936)
Cattell reduced to 171 items he believed accounted for all the other items
The 171 items were administered and came back with 36 surface traits
Subsequent factor analysis produced 16 distinct factors that accounted for all the variables Factor Analytic Strategy
Problems with the factor analytic strategy
The subjective nature of naming factors
Since the main goal factor analysis is to identify common variance, what is identified as common as opposed to unique may be a product of which items are being utilized, the extraction, the rotation, etc. 11