Two Validation Studies of a Performance Validity Test for Autism Spectrum Disorder

Thursday, May 12, 2016: 5:30 PM-7:00 PM
Hall A (Baltimore Convention Center)
L. M. Olde Dubbelink1,2, A. M. Scheeren1, A. G. Lever1 and H. M. Geurts3, (1)Dutch Autism & ADHD research center (d'Arc) Dept of Psychology, Brain and Cognition, University of Amsterdam, Amsterdam, Netherlands, (2)Dept. of Research, Development & Innovation, Dr. Leo Kannerhuis, Doorwerth, Netherlands, (3)Dutch Autism & ADHD research center (d'Arc) Dept of Psychology, Brain and Cognition, University of Amsterdam; Dr. Leo Kannerhuis, Amsterdam, Netherlands
Background: People seeking treatment for psychological symptoms can feign or exaggerate their functioning in order to receive external incentives such as financial aids or reduced responsibility for criminal actions. To detect malingering and increase diagnostic accuracy, performance validity tests (PVT’s) are used in (neuro)psychological assessments. Currently, no PVT exists for autism spectrum disorder (ASD). Therefore, in two studies, we examined whether the Morel Emotional Numbing Test (MENT), a simple emotion recognition test, has potential as a PVT for ASD in adulthood. As intelligent adults with ASD usually do not have difficulties in recognizing simple emotions, we hypothesized that they would perform near to perfect on the MENT. Furthermore, we hypothesized that instructed malingerers would perform poorly on this test. 

Objectives: To examine the validity of a simple emotion recognition task, the MENT, as a PVT for adults with ASD. 

Methods: In Study 1, 19 adults with ASD and 19 typically developing (TD) controls(age: 27-54; IQ>80) completed the MENT. In Study 2, 26 TD students instructed to simulate ASD and 26 TD controls (age: 18-46) completed the MENT. 

Results: In Study 1, Bayesian hypotheses testing showed that adults with and without ASD performed equally well on the MENT. In Study 2, group comparisons demonstrated that instructed malingerers made more errors on the MENT than those instructed to do their best. Moreover, instructed malingerers performed worse on the MENT than adults with ASD. ROC curve analyses demonstrated that a cut-off point of 7 errors reached a specificity of 100% and a sensitivity of 96%. 

Conclusions: These study findings suggest that the MENT may be considered as a PVT for ASD, because it appears to be face-valid measure of ASD, is insensitive to ASD (i.e. has high specificity), and  detects malingerers by luring them to perform beneath their true level of ability (i.e. has high sensitivity). However, one should note that these findings are preliminary as the performances of intelligent adults with ASD (Study 1) may not generalize to individuals with ASD with lower intellectual abilities or a broader age range. Moreover, performances of healthy students instructed to malinger as reported in Study 2 may not be comparable to the performances of those suspected of malingering in a real diagnostic setting. Therefore, we encourage future research on the validity of the MENT as a PVT for ASD in a clinical or a forensic context.