Computerized Assessment of Response Bias
 (CARB)

CARB has demonstrated utility for detecting “incomplete effort”, symptom exaggeration, response bias, feigning and malingering of cognitive deficits. CARB has been effectively employed in veritable and claimed head injury, chronic fatigue syndrome, musculoskeletal injuries, DSM-IV pain disorder, and psychiatric or emotional disturbances (mostly depression). A significant overlap between cognitive malingering and the self-report of a large variety of somatic and psychiatric symptoms has been observed in populations evaluated for compensation-related claims. CARB is relatively culture fair, and is shipped with both English and Spanish Language Instructions. CARB has been entered into evidence in hundreds of state and federal court cases, and is supported by the largest research databases devoted to symptom exaggeration (1752 & 658 cases in two schema).

CARB is the most efficient forced-choice digit recognition test currently available. After studying stimulus presentation times and distraction delays in 1996, these were shortened to enable full administrations in as little as 12 minutes! Additionally, research on 1752 CARB protocols in compensation-related evaluations from 13 sites in the US and Canada resulted in the implementation of an early termination option that rewards excellent early block performance.

CARB’s “early termination feature” has a false negative rate of only 3% (on CARB itself) and can permit 67% of patients to complete the test before all blocks have been given. CARB was completely revised in 1997 with these features (reporting was updated again in May 1999):

• Rules out symptom exaggeration on CARB in as little as 5 minutes

• Full administrations in 12 minutes with the standard configuration

• References a database of 1752 clinical compensation claimants

• Multiple performance comparisons to various reference groups

• Population-based risk assessment for symptom exaggeration

• Pattern analyses for detection of unusual response sequences

• Full analysis of response time data

• Supports basic and clinical research with numerous options

CARB was first available in 1992. It is a completely self-administered test that implements the widely researched, forced choice digit recognition paradigm popularized by Binder, and Hiscock and Hiscock. The standard implementation is comprised of three blocks of 37 trials each. This number can also be reduced to only 25 trials per block. A trial begins with the presentation of a five digit number target stimulus in the center of the screen, followed by a distraction delay wherein the patient is instructed to count backwards silently from 20. The time period for this distraction increases in successive blocks, thus increasing the perceived level of difficulty.