Chapter Three

CHAPTER 3
Assessing the Validity of MMPI-2 Profiles

The MMPI-2 and the MMPI-A can both be faked
However, in most cases, when we know a person is trying to distort his/her responses, we’re able to appraise the person’s message and the extent of their response distortion

The determination of scale invalidity from a test score is a somewhat arbitrary process. Scores on a particular scale are continuous distributions; the determination of “valid” vs. “invalid” is a dichotomous process
In clinical interpretation, it is customary to use “cutoff scores” to determine the validity of performance on a validity scale.
Such cutoff scores represent a “best guess” estimate

Item Omissions, Inconsistent Responding, and Fixed Responding
Cannot Say Score (?)
The total number of items the test-taker didn’t answer
Provides insight into the subject’s cooperativeness
If the person omits more than 30 items within the first 370 questions, the protocol is considered invalid
Possible reasons for omissions include: test defensiveness, indecisiveness, fatigue, low mood, carelessness, poor reading skill, or perception that the item is irrelevant
Variable Response Inconsistency (VRIN)
VRIN is made up of 67 pairs of items, used to detect inconsistent responding; the scale is scored by summing the number of inconsistent responses
VRIN >= 80 indicates inconsistent random responding; MMPI protocol is invalid
VRIN 70-79 are potentially invalid

True Response Inconsistency (TRIN)
Consists of 23 pairs of items to which the same response is inconsistent
TRIN is scored as follows: inconsistencies in true pairs - inconsistencies in false pairs + 9
TRIN >= 80 indicates inconsistent responding due to “yea”- or “nay”-saying
“yea-saying” = TRIN>=80T
“nea-saying” = TRIN>=80F
VRIN 70-79 suggests possible inconsistent responding

All-True or All-False Pattern
An extremely low true or false percentage, less than 20%, represents a highly distorted response pattern (due to conscious manipulation or careless responding)
All true:
Elevates Pa, Pt, Sc, and Ma scales
F scale elevation is very high
L and K scores are very low
All scales (except SOD) are significantly elevated
All false:
L and K extremely elevated
F scale elevated
Produces a “neurotic” appearing profile
Most scales are attenuated (HEA is elevated)

Measures of Random and Exaggerated Responding
Infrequency (F) Scale
Theory is that people attempting to claim psychological problems that they don’t have will go to extremes and endorse symptoms from broad and inconsistent problem areas—in excess of what most patients would endorse (called “plus-getting”)
60 items representing a wide-range of symptoms and aberrant attitudes
“normal” persons usually endorse <5

note for our child people: the F scale for adults does not work for adolescents
F scale is also a good indicator of random responding
Endorsing approx. 30 items means the person may be engaging in random item endorsement, or committing an error in response recording

In sum, high-ranging F scores may reflect:
1. Possible recording error
2. Random responding
3. Possible disorientation
4. Severe psychopathology
5. Possible malingering
6. Different cultural background (see Table 3-4, on pg. 47 of text)

Infrenquency Back (Fb) Scale
Used to detect deviant/random responding after item 370
Suggested interpretations include:
1. If both Fb and F are elevated over T = 110, no interpretation of Fb is necessary as the clinical and content scales may be invalid by F scale criteria
2. If the F scale T score is <89, and Fb is <89, a generally valid approach is indicated
3. If the F scale T score is <89,>90, an interpretation of Fb is needed (see Table 3-5, pg. 48 of text, for guidelines)

Psychopathology Infrequency Fp Scale
Measure of infrequency that has been added to the basic profile; provides perspective on the veracity of the client’s symptom claiming
Fp items are rare or extreme within a sample of persons with severe psychological disability
Fp thus assesses the extent to which the subject is claiming more problems than people in an inpatient psychiatric facility

F-K (Dissimulation) Index
Another way of evaluating symptom exaggeration is to contrast performance on the F scale with performance on the K scale
High symptom checking (high F) plus low defensiveness (low K) suggests an invalid performance
Determined by subtracting the raw score of K from the raw score of F
Gough (the creator of the F-K scale) said that F-K score of 9 or greater suggests an invalid profile due to symptom exaggeration
Others have since suggested a score of 12 or higher to be invalid
Using the F-K index to discover “fake good” profiles has not worked well in practice, and is not recommended for clinical use

Measures of Defensiveness and Claims of Extreme Virtue
Lie (L) Scale
Measures the tendency to distort responses by claiming excessive virtue
15 items center around the assertion of great virtue (“I do not always tell the truth” [False])
T > 65 suggests the individual is claiming a degree of virtue not commonly found
L scale is also associated with personality characteristics that suggest naivete, lack of psychological mindedness, rigid thinking, unrealistic self-image, and neurotic defensiveness.
See Table 3-7, pg. 51 of text, for guidelines on interpreting L

K Scale
Measures test defensiveness (modifies 5 scales: Hs, Pd, Pt, Sc, and Ma)
Items are much less “obvious” than the L scale
Assesses willingness to disclose personal information and discuss problems
High scores (>65) reflect uncooperative attitude and reluctance to disclose personal information
Low scores (<45) suggest openness and frankness K s positively correlated with intelligence and educational level; this should be taken into consideration when interpreting scores Adjustment is probably necessary for persons with less than a high school education Some debate has emerged re: whether K actually improves discrimination However, non K corrected scores are not recommended for clinical interpretation, due to a lack of research on them See Table 3-8, pg. 52 of text for K interpretive guidelines Superlative Self-Presentation (S) Scale Measures test defensiveness Five subscales offer a breakdown of item content: 1. S1 – Beliefs in human goodness – items concerned with basic human goodness 2. S2 – Serenity – e.g., “I have never felt better in my life than I do now [True]” 3. S3 – Contentment with life – morale-related items 4. S4 – Patience/ denial of irritability – endorsing content on this scale suggests the person is calm, cool, and patient 5. S5 – Denial of moral flaws – items deal with denial of flaws Elevations on these subscales suggest particular content areas wherein the client is denying difficulty. Persons taking the MMPI-2 in personnel screening situations are typically defensive on all 5 subscales; persons evaluated in family custody cases tend to have high elevations on S4 and S5 Patterns of Response Invalidity Fake-good profile Presenting an overly-favorable self-view L is prominent, suggesting extremely high virtue This profile is associated with child-custody evaluations, personnel screening, and injury litigation Defensive Profile Found in persons unwilling to disclose personal information (e.g., reluctant therapy cases; persons being assessed against their will) High K and S elevations This profile doesn’t clearly explain the person’s adjustment problems; a cautious statement re: the subject’s reluctance to report problems should be made When basic or content scale elevations occur in the context of a defensive record, the results should not be considered invalid, but rather should be interpreted with the understanding that test profles are likely to underrepresent problems Exaggerated Symptom Pattern Shown below is a marginally valid profile; it should be interpreted with great caution, as the individual clearly intended to present more problems than he or she actually is The highly exaggerated profile shown below is a clear case of a mixed and confused clinical picture, in which a number of extreme and possibly unrelated symptoms are endorsed The below profile was obtained in a worker’s comp. case wherein a person was claiming physical injury, when no work-related injury incident had been reported Invalid Exaggerated Pattern The extremely exaggerated profile shown below should not be interpreted, save to note that it is an exaggerated, and likely faked, record. The persons has endorsed a broad range of unrelated symptoms, and the high VRIN score further supports the notion that the respondent has not answered items in a consistent, selective, task-oriented manner.