Chapter Ten

Chapter 10:
Interpreting the MMPI-A Validity Measures

The MMPI-A is a effective tool to identify malingering and deception when a patient is psychologically assessed
An accurate interpretation of assessment results depends on how accurately and honestly an individual self-reports
Individuals who distort their answers:
a) compromise precise MMPI-A interpretation
b) constrain validity of the empirically derived standard scale (those punks!)

· Adolescents may still distort responses despite proper test administration and test instructions (“F*#k the Man!”)
· The first step in interpreting a teen’s MMPI-A is to determine if answers were given openly and honestly; this is the only way to obtain an accurate profile

Threats to MMPI-A Profile Validity
I. Random Responding: indiscriminant and nonsystematic responses
Possible reasons for random responding
a) dissimulation (concealing/disguising one’s thoughts)
b) poor reading skills
c) verbal comprehension difficulties
d) difficulty deciding on a response
e) clinical conditions (e.g., psychosis or extreme mental confusions)
II. Non-Random Responding: two types
a) fixed responding (e.g., “yea-saying” or “nay saying” without considering item content)
b) misleading responding: not giving an accurate self-description with an answer (e.g., “faking good” by being defensive, or “faking bad” by malingering or exaggerating responses)

Six Validity Measures of the MMPI-A
(See table 10-1 (p.228) to see how the MMPI-A indicators differ from the MMPI & MMPI-2. It is actually not that interesting, but I point it out anyway)

These 6 Validity measures are divided into 2 categories:

I. Measures of Defensiveness and Fixed Responding
a) Cannon Say (? Or Cs) Score
Not specifically a scale
a count of items either left unanswered, or marked both true AND false
Cs³30 lead to questionable profile validity
Encouraging an individual to answer previously omitted answers may decrease Cs
If omissions occurred in the ‘back half’ of the MMPI-A, original validity (L, F, & K) and standard scores can still be interpreted (these items appear in the ‘front half’’, or the first 350 items)
Computerized reports (e.g, The Minnesota Report ™) can list omitted items and help the clinician determine if there was a pattern to omitted items (e.g, items relating to family problems)
If items are omitted, the clinician should try to figure out why – interviewing often works!
Possible reasons for omitting items:
Not understanding a question’s meaning
Questions considered irrelevant to life experience
Careless
Poor reading skills
Indecision
Mood (e.g., very depressed people often lack energy to complete test)
Overly defensive test-taking attitude

b) Lie (L) Scale
“Any naïve attempt by an adolescent to put him- or herself in a favorable light, particularly with reference to personal ethics and social behavior” (Hathaway and Monachesi, 1953)
Alerts the examiner to various possibilities:
naïve and virtuous presentation of self
unwilling to relatively minor flaws
claiming high moral and ethical code
Somewhat unlikely to act out
“Nay-sayers” more likely to have elevated L scale (use TRIN to rule it out)
T scores on the L scale that are ³65 are potentially problematic and defensive
T-Scores ranging from 60-64 moderately elevated, suggestive of a possibly defensive response set
Absence of psychological problems should not be assumed with either an elevated L score or normal-limits scale score.

c) Defensiveness (K) Scale
Scale developed to identify attempts to deny psychopathology and present an overly favorable picture by individuals with non-test evidence of significant psychopathology
Similar to the L scale in that it attempts to identify individuals who may be:
defensive
not candid
feigning excellent psychological health
T Scores ³65 signify high elevation on K, may indicate presence of a potentially problematic defensive response set
T scores btw 60-64 are moderately elevated, suggesting a possible defense response set
Nay-sayers may have elevated K, use TRIN scale to help determine if this accounts for elevated K
An elevated K does not mean a MMPI-A profile can’t be interpreted
Absence of psychological problems should not be assumed with either an elevated K score or normal-limits scale score

d) True Response Inconsistency (TRIN) Scale
Measure of inconsistent responses
MMPI-A has 24 item-response pairs (e.g, “I have no close friends” and “I have a close friend whom I can share secrets with”)
TRIN scores ³ 100 are significant and indicative of an invalid profile due of indiscriminate yea-saying or nay-saying (see manual for exact guidelines based on gender, etc)
TRIN scores of 75-99 are suspect, indicating possible profile invalidity
Elevations on L and K (³65 T score) combined with elevated TRIN (³75) indicate an invalid and inconsistent “nay-saying” response set, (not defensiveness) and other MMPI-A scores should not be interpreted
Low scores on L and K (40³) combined with elevation on TRIN (³100) indicated an invalid and inconsistent “yea-saying” response set, (not frankness) and other MMPI-A scores should not be interpreted

There are case examples highlighting defensiveness and fixed responding (see pg 234-36)

II. Measures of Random and Exaggerated Responding

a) Infrequency (F) Scale
High scores indicate individuals who present themselves in a negative light or “faking bad”
Opposite of the L scale
Reasons for elevated F scales in adolescents:
Possible overly candid responses
Possible failing to answer carefully or consistently
Possible reading comprehension problems
Possible faking bad or malingering
Possible exaggeration of symptoms
Possible serious psychopathology
F³110 indicate an invalid profile, don’t interpret it
An F-K index ³19 (raw score on K – raw score on F) suggests possibility of malingering

F Scores & VRIN Scores
T scores ³ 90 on F, F1 or F2 (described below) signify very high elevations indicative of a potentially problematic response set.
T scores ranging from 66-89 on F, F1 or F2 (described below) signify high elevations indicative of a potentially problematic response set
T-Scores of 65-79 on F, F1 or F2 (described below) signify moderate elevations reflecting endorsement of a variety of symptoms
Use VRIN T scores to determine possible reasons
If VRIN T-score is <70, rule out inconsistent responding
If VRIN T-score is ³75, inconsistent responding is likely

Infrequency 1 & 2 (F1 & F2) Subscales
Used to determine if a problem response is in only one half
F1: front half (items 1-236)
F2: back half (item 242-end)
Elevation on F2 (but not F1) indicates that a potentially problematic response set occurred in the 2nd half of the test
a possible cause is boredom – implementing two shorter sessions or giving frequent breaks may help alleviate this

b) Variable Response Inconsistency (VRIN) Scale
· Created to address inconsistent responding, (e.g. careless or random responses)
· If F scale is not elevated, VRIN scores ³80 are significant and indicative of an invalid profile that was marked inconsistently
· If this is found, no other MMPI-A scales should be interpreted in the response set
· VRIN scores ranging from 70-79 are suspect, possibly indicating an invalid profile
· Interpret the MMPI-A with caution

See the other case profiles, blah blah blah (pg. 242-247). Don’t worry, there are lots of graphs to fill up the space.

A final note about Cutoff Scores…
Cutoff scores for all validity measures are given, however take them with an empirical grain of salt. Why?
Style of response varies by setting (e.g, impatient treatment, in schools, or juvenile detention centers)
Reasons for testing (e.g., anonymous and voluntary research or for a child-custody eval)
How the teen is treated during the assessment (e.g, informed about why the evaluation is occurring? Frequent encouragement or bathroom/smoke breaks given?)