In respect to the MeSA-AE malingering is defined as deliberately making test responses that feign impairments of attention or response control for personal gain. Published research has found that individuals who malinger on this test produce extreme quotient scale scores. Nevertheless, the determination of malingering requires that a clinical decision be made by the examiner. In almost all cases, additional tests of malingering will need to be administered in order to accurately identify its occurrence.
In respect to the MeSA-AE Test, the possibility of malingering or an attempt to exaggerate symptoms is evident in unusually long Test A and Test B completion times which result in extremely low test scale scores. Malingering is defined as attempting to feign impairments of cognitive deficits for personal gain. Research completed by Egeland & Langfjaeran (2007), Ruffolo et al. (2000), and O’Bryant et al.(2003) found that individuals in experiments who are instructed to “fake bad” on this test or litigants who are suspected as malingering based on recognized tests of malingering clearly have extremely low test quotient scale scores. Those classified as malingering demonstrated quotient scale scores that were significantly less than individuals with recognized neurological impairments.
None of the current malingering research for the Trail Making Test reported making adjustments that took into account age and education differences in evaluating test completion time scores. However, as noted in the MeSA-AE normative discussion above, age and education factors greatly influence the interpretation of test completion time scores. In light of this fact the standard quotient scale scores for Test A and Test B based on the research studies by Egeland & Langfjaeran (2007), Ruffolo et al. (2000), and O’Bryant et al. (2003) were calculated based on the MeSA-AE normative database. These calculations included age and education correction factors and were based on the mean test completion time scores, age group means and education level means reported in these three studies. The mean Test A quotient scale score (i.e., ACQ) for the malingerers in these three studies, was 41. Neurologically impaired individuals were found to have a mean quotient scale score for ACQ of 71. This thirty point difference revealed a significant difference between the two groups of two standard deviations (SD). In respect to Test B age and education corrected quotient scale score (i.e., CFQ) the three malingering groups had a mean quotient score of 54 and the impaired groups CFQ score was 78. These two groups differ by 24 points (1.6 SD) which was considered significant. Based on this research it was determined that standard quotient scale scores for both Test A and B when they fell in the extremely impaired range (i.e., less than or equal to 60, 2.65 SD) would be the best cut-off measure to use for differentiating individuals who were suspected of malingering from those who were truly impaired. Empirically, this cut off level means that less than one percent of a non-impaired normative sample would be incorrectly labelled as possibly malingering.
O’Bryant’s et al. (2003) research did support the validity of using test completion times to differentiate suspected litigant malingerers and classified 63% of them correctly while also accurately identifying 85% of non-malingerers. Unfortunately, detailed test data available in the three studies discussed above was not sufficient for completing a similar discriminate function analysis in determining the specificity and sensitivity of age and education corrected quotient scale scores in the differentiation of suspected malingerers from neurologically impaired individuals. Until further malingering research is completed using the criteria discussed, the MeSA-AE test by itself cannot be used solely by itself in identifying suspected malingering. However, it can be used in conjunction with other tests of malingering and in this way will provide examiners with useful information to help them in making a determination of malingering in conjunction with their comprehensive evaluation and clinical observations.
As discussed above in the test overview section, only one type of error can be made during Test A which is referred to as a Sequential Error. For Test A the total number of Sequential Errors made was used in the assessment of malingering. This assessment for Test B includes both Sequential and Perseverative Errors. In the existing research exploring the use of test errors in determining malingering, no mention has been found that specifies the different types of possible errors or whether one type of error was included or excluded. Consequently, it can be assumed that the total number of errors for either test, regardless of the type of error made is the variable used in assessing whether malingering may possibly have occurred.
The research findings reported by Ruffolo et al. (2000) found that individuals who were instructed to simulate malingering made significantly more errors on Test A and Test B. Research by O’Bryant et al. (2003) only revealed a trend when comparing the number of errors made of litigants suspected of malingering and individuals with recognized neurological impairments. However, this study differed in that the subjects used were not simulators instructed in malingering. The problem of using the total number of errors in identifying suspected malingerers is further complicated by the fact that individuals without any impairment have been found to make a number of errors depending on their age and education level on both Test A and Test B (Ashendorf et al., 2008). Consequently, the research supports only using the number of total errors for either Test A or Test B as a secondary indicator of possible malingering. In addition, the cutoff criterion for the number of errors made also needs to be in the extreme range (i.e., greater than 2.65 SD) in order to avoid misclassifying normal or impaired individuals as possible malingerers. An analysis was made of the four studies discussed above in order to determine error cutoff scores for Test A and Test B that met this criterion. It was determined that for individuals less than 55 years old or any individual who was a college graduate (regardless of their age) that made either two or more errors on Test A or three or more errors on Test B was indicative of possible malingering. For individuals who were 55 years or older or who had not completed college a slightly higher cut off level was determined. For these individuals, suspected malingering was indicated by three or more errors on Test A and four or more errors on Test B. Additional support for using these cut off scores is based on the research by Irverson et al. (2002) who found that it is highly unusual for individuals with moderate to severe traumatic brain injury (TBI) to have two or more errors on Test A and four or more errors on Test B.
In summary, the research supports that extreme ACQ and CFQ quotient scale scores can provide the examiner with useful information regarding possible malingering. Secondarily, an examination of the number of the Sequence and Perseverative Errors, when extreme, can provide further indication of possible malingering for individuals who meet this primary cutoff quotient score criterion. Individuals who are motivated for one reason or another to intentionally perform poorly on these tests can “take their time” in completing them without overtly arousing suspicion. In addition, some but not all of them may also decide to intentionally make an unusual number of mistakes during either test. The research clearly supports that only when either an individual’s ACQ or CFQ test performance falls in the extremely impaired range are examiners justified in considering the test results as indicating possible malingering. However, by itself the MeSA-AE test cannot be the sole means of detecting malingering and in all cases other clinical data and observations must be considered in making this determination.