Correction to: Journal of Neural Transmission (2021) 128:1045–1063 https://doi.org/10.1007/s00702-021-02318-y

Due to an administrative error, 73 individuals were wrongfully assigned to the Simulation Group. Correction of this error consequently reduced its sample size from 242 to 169 individuals.

This did not affect the development of the ADHD Credibility Index (ACI). However, it resulted in an underestimation of the index’ classification accuracy in its initial validation.

The corrected demographic data of the Simulation Group (see amended Table 1) differed from those of the other experimental groups as described in the original publication. Participants of this group were still significantly younger than participants in the ADHD Groups (credible: z = 7.357, adjusted p < 0.01; non-credible: z = 3.819, adjusted p < 0.01) and Control Groups (credible: z = 20.681, adjusted p < 0.01; overreporting: z = 3.557, adjusted p < 0.01). The gender distribution in this group also differed from the Credible (χ2 (1) = 42.518, p < 0.01) and Overreporting Control Groups (χ2 (1) = 20.289, p < 0.01) as well as the ADHD Groups (credible: χ2 (1) = 16.176, p < 0.01; non-credible: χ2 (1) = 13.327, p = 0.01). In terms of education, instructed simulators differed from credible participants in the Control Group (z = − 7.611, adjusted p < 0.01) and the ADHD Group (z =  − 3.660, adjusted p < 0.01), but not from overreporting controls (z = 1.864, adjusted p = 0.623) or non-credible patients with ADHD (z = 0.014, adjusted p = 1.00).

Table 1. Descriptive data by group
Full size table

Correcting the Simulation Group further required the revision of Tables 2, 5, 6, 8, 9, 10, 11. With the exception of the following findings, the pattern of results remained unchanged. The validity indicators under study showed overall higher sensitivity rates and larger effect sizes than previously reported. Rather than the small effect described in the original publication, the ACI yielded a large effect for the comparison of instructed simulators and credible adults with ADHD (d = 1.29, 95% CI [0.49, 2.09]). As was previously the case, the largest effect could be observed on the Supposed Symptoms subscale, followed by Exaggerated Symptoms, Selectivity, and lastly Symptom Combinations (see amended Appendix 3). Additionally, changes in classification accuracy were noted for two DSM scales (see amended Table 6). While previously significant, ROC analysis showed a statistically non-significant result for the DSM Inattention (E) scale. In contrast, the DSM Total (G), yielded a statistically significant result upon correction. These changes did not affect the conclusions drawn from the results.

Table 2. Summary statistics for ADHD Credibility Index (ACI) scores by group
Full size table
Table 5. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the ADHD Credibility Index (ACI) and CAARS Infrequency Index (CII) in the detection of simulated ADHD, non-credible adults with ADHD, and overreport on CAARS DSM Scales
Full size table
Table 6. Results of ROC analyses distinguishing credible adults with ADHD (n = 95) from simulators (n = 161)
Full size table
Table 8. Results of ROC analyses distinguishing participants with unremarkable T-Scores on the DSM Scales (n = 1103) from Overreporters (n = 191)
Full size table
Table 9. Agreement between ADHD Credibility Index and overreport on DSM scales
Full size table
Table 10. Agreement between ADHD Credibility Index and existing validity indicators
Full size table
Table 11. Agreement between ADHD Credibility Index (ACI) and CAARS Infrequency Index (CII)
Full size table