The Test of Memory Malingering (TOMM) Overseas: Validation in a Singaporean Sample

Ivana Chan, National University of Singapore
Isabelle Shu Min Ong, National University of Singapore
Kenji Gwee Dpsych., National University of Singapore; Institute of Mental Health, Singapore

The present research aims to establish the predictive validity of the Test of Memory Malingering (TOMM) in a Singaporean sample with mood and psychotic disorders. Its clinical utility was compared against other performance validity tests: Dot Counting Test (DCT), Rey 15-Item Memory Test II (Rey II), Medical Symptom Validity Test (MSVT), Nonverbal Medical Symptom Validity Test (NV-MSVT), Word Memory Test (WMT) and Victoria Symptom Validity Test (VSVT). Findings revealed that TOMM, MSVT, NV-MSVT, WMT and VSVT preserved the ability to distinguish between Clinical participants’ simulation groups, and TOMM outperformed all tests except NVMSVT. Overall, findings supported continued use of TOMM.

Feigning, malingering, performance validity test, Singapore, clinical, TOMM

“…The purpose of developing performance validity tests (PVTs) is to detect feigned cognitive impairment within forensic settings, wherein putting forth optimal effort becomes the subject of interest…Scores below a recommended cutoff score on PVTs would indicate a suspect cognitive profile, suggesting the biased lack of effort or the expended effort to perform poorly…the failure of PVTs is not equivalent to identifying malingering…Due to the difficulty of measuring motivation to malinger, the term feigning will be used for the rest of the study…The TOMM [Test of Memory Malingering]…is a non-verbal PVT that is based on recognition memory of 50 pictures of common everyday items. It comprises of two learning trials and an optional retention trial. The TOMM has frequently been lauded as the gold standard in performance validity testing…Norming data established by Tombaugh (1997) resulted in a criterion core of 45 out of 50…validation of several PVTs in the local context is an important cause to pursue. Hence, in addition to the TOMM, the current study also examined the following PVTs: Dot Counting Test (DCT)…Rey 15-Item Memory Test II (Rey II)…Medical Symptom Validity (MSVT)…Nonverbal Medical Symptom Validity Test (NV-MSVT)…Word Memory Test (WMT)…and Victoria Symptom Validity Test (VSVT)…” (p.1-2).

“The present study aimed to establish the predictive validity of the TOMM in a sample of patients with mood and psychotic disorders, while comparing its performance to other established PVTs…we also aimed to add incremental ecological value to the sparse literature of PVTs in the context of Singapore…Lastly, given the inconclusive evidence on the link between IQ and PVT scores, we aimed to examine the relationship between these two factors in our study. It was hypothesized that: (1) PVTs would be sensitive to psychiatric disorders; (2) TOMM would outperform the other tests in terms of predictive validity; (3) there would be a significant relationship between CTONI-2 and PVT performance in all Honest groups, such that lower CTONI-2 scores related to poorer PVT performance” (p.3).

“The study sample consisted of two groups of participants: 65 adult inpatients with mood and psychotic disorders from a psychiatric unit…and 60 healthy adults living in the community…participants were randomly assigned to one of two groups – Honest…or Simulated Feigning group…participants in the Simulated Feigning group were told to feign cognitive impairment evaluation tests in a way that the cognitively impaired would…The present research established that the TOMM had a relatively higher classification accuracy in comparison to the other PVTs, and this was maintained even in the Clinical sample. In addition, the TOMM retention was one of the only tests that was less affected by IQ compared to other PVTs. Hence, use of the TOMM Retention, especially in low IQ populations, would likely be a valuable addition in a forensic evaluation…” (p.3-10).

“Firstly, the present results provided evidence that DCT and Rey-II measures obtained classification accuracies of no better than chance levels in the Clinical group. In addition, both the sensitivity and specificity of all the PVTs employed in the current study were greatly reduced in the Clinical sample. Thus, Hypothesis 1…was supported…Next, results revealed that the TOMM outperformed all other PVT measures, except for the NV-MSVT, in terms of classification accuracy, while also maintaining a high specificity rate in the Clinical sample…neuropsychologists generally agree that the specificity of a PVT should take precedence over sensitivity…the TOMM…displayed outstanding specificity…overall, the present results provided compelling evidence that the TOMM outperformed the WMT, DCT, Rey-II, MSVT, NV-MSVT, and VSVT in both community and clinical groups. Thus, Hypothesis 2 was supported…In the Community Honest group, contrary to predictions, participants with lower FSIQ scores were more likely to result in lower scores (higher errors) only on three of the PVTs. In the Clinical Honest group, however, there were significant correlations between FSIQ scores and all the measures except for TOMM retention and VSVT Easy…” (p.12).

“…use of the TOMM Retention, especially in low IQ populations, would likely be a valuable addition in a forensic evaluation…the implication of the current findings [regarding hypothesis 1] is that caution should be taken when utilizing the PVTs, especially the Rey-II or DCT in a clinical setting…practitioners may be concerned about how to proceed in the event that the TOMM and NV-MSVT produce conflicting results…In this case, it may be more practical to include more than these two tools…it [has been] reported that the use of multiple measures (up to seven) did not seem to cause a significant increase in false-positive errors beyond the per-test false positive rate, implying minimal drawbacks for the increase in classification accuracy…” (p.10-12)

“The susceptibility of PVT results to differences in FSIQ highlighted the extent of impact that IQ has on performances of Honest participants. As such, a measurement of IQ is warranted when conducting PVTs, especially for samples that are known to be skewed toward the lower IQ bands. Forensically, practitioners in the local context are known to use only the TOMM, even with patients with low IQ. Among the battery of tests, the TOMM appeared to be the most robust PVT, presenting the highest levels of specificity in the Clinical group, which was found to have lower intellectual functioning. Despite that, although the TOMM Retention was one of the only two PVTs that had no significant association with IQ in our study, TOMM Trials 1 and 2 were still found to have a significant effect on IQ. This may be a cause for concern because patients with lower intellectual functioning may be at a higher risk of being miscategorized as feigning. Hence, these results indicate the need for caution in cases where responders may have limited intellectual functioning” (p.13).

“Alternative cutoff scores based on the present community and clinical samples were identified. These revised scores were selected to hold specificity at approximately 90% or above, taking into account the high cost of misidentifying an honest person as a feigner…Application of the revised cutoff scores produced moderately higher sensitivity and specificity in the community sample. For instance, the revised cutoff score of 48 for TOMM Trial 2 increased sensitivity by 3%…the revised cutoff score of 37 for TOMM Retention maintained sensitivity at 93% while increasing specificity by 17%…these revised cutoff scores would require further validation in a different prevalence setting where feigning is likely to occur and should be regarded cautiously. Nonetheless, comparing our optimal cutoff scores to that of other studies conducted primarily in the United States, we found that the optimal scores were relatively similar…” (p.13).

As always, please join the discussion below if you have thoughts or comments to add!