Self-report inventories and other objective assessments are widely used by psychological professionals. These tests offer clinicians a cost-effective and efficient means for diagnosing individuals beyond more subjective methods such as projective tests or pure clinical judgment. The Minnesota Multiphasic Personality Inventory and its successor, the MMPI-2, remain one of the most popular self-report assessments available. The MMPI-2 gains its objective utility from the use of a technique called empirical criterion keying.
Empirical criterion keying examines the differences between honest responses from groups of normative individuals and a criterion group of those already diagnosed with disorders. Questions that effectively delineate between the groups are keyed and added to a clinical scale associated with the particular criterion group. In addition to the development of ten objective clinical scales, empirical criterion keying allows the MMPI-2 to differentiate between honest responders and individuals who are faking.
Validity Scales
Individuals do not always respond with direct and honest answers on the MMPI-2 and other objective assessments (Cassissi & Workman, 1992). Malingering behavior—either in the form of feigning a disorder (faking bad) or concealing negative qualities (faking good)—arises from a variety of personal motives (Rogers, Bagby & Charkaborty, 1993). Self-report psychological assessments are often criticized for their susceptibility to malingering (Wiggins, 1973 from Cassissi & Workman, 1992). In response to these phenomena, the developers of the MMPI-2 utilized empirical criterion keying to establish seven validity scales (four from the original MMPI, three new) that could detect malingering responses.
The ability to detect malingerers remains an important yet difficult task for subjective assessments in the clinical setting, and even more difficult for self-report tests (Lim & Butcher, 1996). The MMPI-2, through its effective use of validity scales, continues to be one of the best, and therefore most important, tools by which clinicians may identify malingering individuals (Bagby, Rogers & Buis, 1994; Rogers, Bagby, Charkaborty, 1993). Research focusing on the validity scales of the MMPI-2 continues to define the strengths and limitations of their detection abilities.
Research studies that examine the detection abilities of the MMPI-2 validity scales (typically, the L, F, and K scales) strive to achieve a variety of different objectives. Certain studies seek to determine the differential effectiveness of validity scales in the detection of faking specific disorders (Bagby, et al., 1997), whereas others observe changes in validity scale effectiveness when subjects receive coaching or different instructions before the test (Rogers, Bagby & Charkaborty, 1993). Some research attempts to identify optimal cutoff scores for increasing malingering sensitivity and specificity (Bagby, Rogers & Buis, 1994). In common, however, all of these studies employ some of their subjects to intentionally fake an assigned profile during an administration of the MMPI-2. Also commonly, validity scale research studies tend to look at two primary conditions: faking bad versus honest responders, and faking good versus honest responders.
Faking Bad versus Honest
Clinical scale results from the MMPI-2 are not necessarily clear-cut indicators of a psychological illness. Honest, normal individuals sometimes show elevated scores on one or more clinical scales, and honest individuals with disorders often show elevated scores on several clinical scales. This phenomena reduces a psychological professional’s ability to interpret data. Fortunately, differences on clinical scales between malingerers, honest responders, and diagnosed patients can be detected. Bagby et al. (1997) found multivariate effects for all ten of the 10 clinical scales between these groups. Depending on strategies or coaching prior to taking the MMPI-2, however, these differences may vary. This further emphasizes the importance of validity scales as the best detectors of malingering.
When faking bad, several studies found subjects tend to over-report symptoms, resulting in significant elevation of the F validity scale T scores (F>100), and lower K scale T scores (K<50) than diagnosed patients (Bagby, Rogers & Buis, 1994; Cassissi & Workman, 1992; Lim & Butcher, 1996). ANOVA and MANOVA consistently find significant differences between faking bad and honest groups on validity scales (Bagby, Rogers & Buis, 1994; Lim & Butcher, 1996; Rogers, Bagby & Charkaborty, 1993). The F scale index is most commonly used to classify possible malingering. Using a raw cutoff F scale score of 17, Lim & Butcher (1996) correctly classified 100% of their Honest and Fake Bad subjects. Malingerers may increase or decrease detection abilities of the validity scales by feigning specific disorders, such depression versus schizophrenia (Bagby, et al. 1997). In general, however, research indicates that successful malingering on the MMPI-2 rarely occurs. Only one Fake Bad group—in which subjects were coached on the MMPI-2, validity scales, and subsequent strategies—managed to go undetected on the validity scales (Rogers, Bagby & Charkaborty, 1993).
Faking Good versus Honest
Because of its efficiency and objectivity, the MMPI-2 has become a widely used tool outside clinical settings, e.g. employers use the MMPI-2 to evaluate job applicants or employees under review. In these situations, individuals often feel obliged to present themselves in an overly virtuous manner. There are at least two groups that fake good: those who are attempting to claim extreme virtue, and those who are attempting to deny/hide a psychological disorder, but MANOVA fails to differentiate the scores of the two groups on the MMPI-2 (Lim & Butcher, 1996). Diagnosed patients that attempt to deny their disorders still show elevated levels on the clinical scales (Bagby, Rogers & Buis, 1994). Evidence suggests that the MMPI-2 cannot detect faking good as well as faking bad (Bagby, Rogers & Buis, 1994; Lim & Butcher, 1996), but the validity scales remain effective indicators of malingering. Significantly elevated L and K scale scores, typically accompanied by F scales scores lower than the normative range, are often the best indicators of a Fake Good profile.
Short Form of the MMPI-2
Although its effectiveness in clinical diagnosis has been called into question, Cassissi and Workman (1992) proposed the utility of an abbreviated version of the MMPI-2 in certain applications. Specifically, their findings supported the effectiveness of the L, F, and K validity scales on a short form version of the MMPI-2. Subjects, instructed to fake good and bad, respond similarly to L, F, and K scale questions on both the full and short form versions of the MMPI-2 (Cassissi & Workman, 1992). This suggests that the MMPI-2 Short Form can offer clinicians a much faster, viable method for identifying malingerers.
Since our study used the MMPI-2 Short Form—focusing on the L, F, and K validity scales—we expected results similar to those found in previous studies. We predicted Fake Bad responders will score significantly higher on the F scale, and lower on the L and K scales, than Honest responders. We also expected the F scale by itself will successfully identify Fake Bad from Honest responders. As an educational tool, the primary purpose of our study was to help students gain a better understanding of validity scales and their function in self-report assessments such as the MMPI-2. By using students as the subjects, the study offered them the opportunity to gain valuable first-hand experience with the classroom material.
Since some studies propose that malingering strategy may play a role in an individual’s ability to go undetected (Lim & Butcher, 1996; Rogers, Bagby & Charkaborty, 1993), we discussed strategies used by our most successful malingerers during a debriefing session. Research also suggests that modifications to instruction sets may alter the effectiveness of MMPI-2 validity scales (Rogers, Bagby & Charkaborty, 1993). The Fake Bad instruction set for this trial outlined a very detailed and believable scenario, a major improvement from the instruction set used in previous trials. An incentive award was offered to subjects who successfully malingered without detection. As a secondary purpose to our study, we hoped to identify any changes in the MMPI-2’s detection abilities when confronted with different types of strategies and improved instruction sets.
References
Bagby, M.R., Rogers, R., & Buis, T. (1994). Detecting Malingered and Defensive Responding on the MMPI-2 in a Forensic Inpatient Sample. Journal of Personality Assessment, 60, 215-226.
Bagby, M.R., Rogers, R., Buis, T., Nicholson, R.A., Cameron, S.L., Rector, N.A., Schuller, D.R., & Seeman, M.V. (1997). Detecting Feigned Depression and Schizophrenia on the MMPI-2. Journal of Personality Assessment, 68, 650-664.
Cassissi, J.E., & Workman, D.E. (1992). The Detection of Malingering and Deception with a Short Form of the MMPI-2 based on the L, F, and K scales. Journal of Clinical Psychology, 48, 54-58.
Lim, J. & Butcher, J.N. (1996). Detection of Faking on the MMPI-2: Differentiation Among Faking-Bad, Denial, and Claiming Extreme Virtue. Journal of Personality Assessment, 67, 1-25.
Rogers, R., Bagby, R.M., & Charkaborty, D. (1993). Feigning Schizophrenic Disorders on the MMPI-2: Detection of Coached Simulators. Journal of Personality Assessment, 60, 215-226.
