MMPI Design, History, etc Flashcards
Beginnings
Objective tests relied on individuals’ willingness and capacity to accurately report feelings and experiences and also depended on reading and intellectual ability.
Response to the need for improving validity of objective personality tests.
MMPI standardization began in the 1930s
1943: First version in 1943
1949: most common version
1970s began work on a revision, MMPI-2 :1989
Stark Hathaway & Jay McKinley
• goal was to establish instrument to perform diagnosis in order to save clinicians time
• approach – empirical method of test construction
o face valid questions
o novel approach versus the projective/rational tests of the time
Patients at Minnesota hospitals were separated to create scales
• SI scale added in the 1960s (Drake) – first scale that was not specifically a pathology
• Control group was anyone visiting the medical areas of hospital
o Random selection, but more limited to Scandinavian population (Minnesota)
There is no meaning to the arrangement of the scales, any associations are coincidental
Originally it was two standard deviations 30-70, anything outside was pathological
Gliberstadt & Duker – evaluate MMPI: “pretty good”
MMPI was meant to diagnose pathology
• it never really worked like this
• correlations with diagnostic labels were okay, but it gives you lots of other useful information
o Milan prob better at dx info
MMPI-2, range included 40-65
Test items administered to 2 or more groups of participants-a criterion group selected for homogeneity w/ respect to a certain diagnosis, cluster of features, traits, characteristics, and a normal comparison group that does not share the same characteristics or only in base-rate amounts.
For example: People diagnosed with depression versus normals
• which questions the depressed answer differently than normals resulted in the depressive scale
• some questions are face valid, some are not – it simply a function of the depressives answered a certain way versus controls. Content is secondary
• Depression scale face-valid questions are okay – people don’t lie very much
o other diagnostic categories lie, but not depression as much, face valid questions are relevant for scale 2
The Normals
Normal reference group was 724 friends and relatives of patients at U. Minnesota
All white, “underprivileged,” and from Minnesota (overrepresented lower educational and occupational groups. Norm groups since have typically scored about 0.5 SDs higher above mean)
Additional normal and patient groups consisted of: high school graduates attending pre-college conferences, med patients, skilled WPA personnel, inpatients in psychiatric unit.
Original normals reflected a fixed reference group rather than a true normative sample and practical implications are important
By not using a new sample of normals to establish the normative standard, Hathaway and McKinley introduced serious constructional deficiency in design of test.
MMPI Restandardization: The MMPI-2
Published in 1989
MMPI-2 not intended for use in adolescents.
[MMPI-A released in 1992 (14-18 yo)]
MMPI-2 minimized changes in composition of original validity and clinical scales.
K factor left intact
Changes made:
○ Contemporary sample of normals→ new clinical data and thus modified scales
○ Items replaced or modified to modernize language.
○ New scales to measure contemporary clinical problems
○ Improved metric replaced linear T scores so scaled scores can be better compared.
New Normative Sample
2600 adults between 18-84 yo
Ethnic diversity greater in this sample, but Asian Americans, Native Americans, and Hispanics underrepresented.
Men and women 18-19 underrepresented and older adults between 70-84.
Higher SES in this sample. General skew in new sample toward high SES. Accurately reflects acceptable representation of US pop, despite accusations that higher education overrepresented.
MMPI General Points
Character scales: Mania, Psychopathic Deviate, Hysteria, Paranoia, K scale
After psychotherapy, expect a decrease on all standard clinical scales on MMPI except K, 3, and 9.
Harris and Lingoes subscales not developed empirically, developed rationally
T scores: everything between 40-65 normal;
Scales 3 & 6 are the best: hardest to fake and rarely give you false positive.
2-7-8 is one of most common profiles.
2-4-6-8: worst male profile
3-6-9: worst female profile
Short forms of the MMPI:
○ Mini Mult: 100 or 160 questions; some validity
○ Micro Mini Mult: 69 questions; validity highly suspect
Behavior is situation specific
Looking at traits is looking at a generalization→ not always valid
o Trait models may not be very good or useful
o Make predictions based on traits, but if traits don’t hold up, then what’s the point?
Hartshorne & May (1928)
o Actual tests were given in different settings: home, church, school, clubhouse
o Correlations between paper and pencil tests and behavior were very low
o Progressive change of situation = progressive decrease in correlation between test and behavior
Walter Mischel – do we have consistency of behavior?
Bemin & Campbell: USAF attitude toward authority
o Highest correlation was between subjects on the same test
o No correlation between attitudes toward different authority figures
• No large correlations between sexual identity and masculinity/femininity
*Data tends to correlates more strongly over time than between situations
Gender representation
Gender is a significant variable
1970s – were women being diagnosed incorrectly? Was there a bias?
More women in state hospitals – but this was actually accounted for by men being in VA hospitals
Women with psychopathy diagnosis were more frequent in state hospitals than men with psychopathy diagnosis…
Male psychopaths are more often put in prison
• Limited sentence in prison, whereas hospitals were indefinite
2 worst MMPI patterns
Males: Sawtooth: elevated 2,4,6,8
• acting out, nasty
• psychopath is complicated dx
o primary psychopath – do not experience negative emotions other than anger
o secondary psychopath – experience anxiety and depression too late (Catholic schoolgirl conscience)
Females: 3-6-9
• borderline personality
• manicky, acting out paranoid hysteric
• classic hysteric
Differentiating Schizophrenia from Mania
Hallucinations/delusions are definitions of psychosis
**you don’t have to be psychotic to be schizophrenic or vice versa
Past designations of schizophrenia:
1970s/80s – more flat/inappropriate affect definition
*split between reality and emotions
Today– “thought disorder” is defining characteristic of SZ
Mania: delusional and hallucinatory – psychotic, not schizophrenic
*schizophrenia implies emotionality
Old Model of progressive deterioration, “continuity model”
Normal→ OC → Paranoid
Normal →Hysterical → Manic
**This model does not hold up—though this absolutely occurs, it’s not always perfectly linear
