The continuous performance test (CPT) is one group of measures for the evaluation of attention as well as response inhibition (or disinhibition); Fleming, Goldberg, and Gold have described the CPT as the gold standard for measuring sustained attention. The original CPT was developed by Rosvold, Mirsky, Sarason, Bransome, and Beck in 1956 as a research tool to study vigilance. Since that time, the CPT has continued to be used in the study of attention as well as executive control, with multiple variations in the components of the task; today, the majority of CPTs are computer administered.
The basic CPT paradigm consists of rapid presentation of continuously changing stimuli with a designated “target” stimulus or pattern such that the individual is to respond (or inhibit responding) based on the stimulus presented. It requires selective attention or vigilance for an infrequently occurring target or relevant stimulus; at the same time, the duration of the task is intended to be sufficient to measure sustained attention. Despite these general similarities, there have been as many versions of the CPT available as there were clinicians who used them. Different CPTs include variations in the basic task (i.e., when to respond or inhibit), the characteristics of the target, variations in the interstimulus interval (ISI), presence or absence of distracters, modality of presentation, duration of the target presentation, duration of task, and so on. The effects of some of these possible variations and modifications to the CPT on performance have been reviewed elsewhere.
Variables of Interest
CPT variables reported include correct hits (number or percentage of correct responses to targets) and omission errors (i.e., number or percentage of targets not responded to). Both correct hits and omission errors are interpreted as indicative of selective attention or inattention. Commission errors (number or percentage of responses to stimuli other than the target) are reported as an index of response inhibition or disinhibition. In some studies, relative accuracy (percentage of correct responses) or total errors (combining omission and commission errors) may be reported.
Reaction time is another measure frequently reported with CPTs. Reaction time or response latency is believed to reflect the speed of processing as well as the speed of motor responding. For example, a child may demonstrate increased omission errors and a slower rate of responding without an associated increase in commission errors; this type of pattern may be interpreted as supporting a hypothesis of difficulty with the allocation of information-processing resources. The consistency or variability of the individual’s performance over time is also of importance; some CPT programs generate the standard deviation of the reaction time across blocks as a measure of consistency in responding and of the ability to sustain attention over time. Alternatively, some researchers report the standard error of the reaction time as an indicator of the consistency or variability of responding over time. Still others use the standard deviation of the standard error over time as an indication of consistency. Rather than using differences in the reaction time as a measure of consistency, some clinicians focus on comparisons of correct and incorrect responses over differing blocks of time within the same administration (i.e., vigilance decrement).
As an alternative to direct performance scores, some clinicians incorporate signal detection theory (SDT) in generating performance indexes for interpretation. The basic premise of SDT is that the decision to respond is based on the child’s setting a certain standard or criterion for responding. SDT variables of sensitivity (also referred to as d’ or d-prime) and response bias (also referred to as beta) are based on signal to noise (i.e., target to nontarget) ratios. Lam and Beale argued that the sensitivity and bias indexes may be more sensitive to differences in performance on the CPT than omission or commission errors and that SDT procedures may be particularly useful in neuropsychological assessment.
Ethical and professional standards demand that clinicians use measures that are technically adequate. With CPTs, as with any formal assessment measure, it is important for administration to be standardized consistent with the collection of normative data. As such, it is imperative for the CPT manuals to clearly state the conditions for standardized administration. Ballard reviewed a number of experimenter-manipulated variables that may influence CPT scores and recommended that manuals address these issues (e.g., examiner presence or absence, instructional set). The potential for these changes to affect test performance, possibly invalidating the interpretation of the results, requires adherence to standardized procedures in the administration of any CPT. In addition, given differences in task demands and parameters across CPTs, performance on any one of these CPTs can be interpreted only based on the normative data for those specific task parameters. For these reasons, although the capacity to customize the task may be beneficial for research purposes, clinical use of customized CPTs is not recommended due to the lack of normative data.
Use and Interpretation
Clinicians also need to be cautious in their interpretation of CPT results. The CPT is only one measure, and multiple sources of information and multiple measures should be used when assessing attention and impulse control problems in order to corroborate CPT findings. Taken together with direct observation, behavior rating scales, and other psychometric tests, the CPT may provide useful information. The CPT is an objective measure that is not subject to rater bias or observer drift; as such, the level of performance on CPTs may be helpful in ruling out or identifying attentional problems and for monitoring medication effectiveness. However, reliance on CPTs as a primary diagnostic tool in determining the presence of a specific disorder (e.g., attention deficit hyperactivity disorder [AD/HD]) is not warranted and will result in an unacceptably high number of false positive errors or overdiagnosis of AD/HD. Without evidence of both internal consistency and temporal stability, any conclusions related to diagnostic considerations or treatment effectiveness are questionable. Given all of the possible variations to the CPT paradigm, it is possible that some combination(s) will prove more helpful in differential diagnosis than others.
- Ballard, J. C. (1996). Computerized assessment of sustained attention: A review of factors affecting vigilance performance. Journal of Clinical and Experimental Neuropsychology, 18, 843-863.
- Bock, R. D. (1982). The role of arousal in Tourette Syndrome. Unpublished doctoral dissertation, New York University, New York.
- Carter, C. S., Krener, P., Chaderjian, M., Northcutt, C., & Wolfe, V. (1995). Abnormal processing of irrelevant information in attention deficit hyperactivity disorder. Psychiatry Research, 56, 59-70.
- Cohen, R. A. (1993). Neuropsychological assessment of attention. In R. A. Cohen (Ed.), The neuropsychology of attention (pp. 307-328). New York: Plenum.
- Fleming, K., Goldberg, T. W., & Gold, J. M. (1994). Applying working memory constructs to schizophrenic cognitive impairment. In. A. S. David & J. C. Cutting (Eds.), The neuropsychology of schizophrenia (pp. 197-213). Hillsdale, NJ: Lawrence Erlbaum.
- Lam, C. M., & Beale, I. L. (1991). Relations among sustained attention, reading performance, and teachers’ ratings of behavior problems. Remedial and Special Education (RASE), 12, 40-17.
- Levin, E. D., Conners, C. K., Sparrow, E., Hinton, S. C., Erhardt, D., Meek, W. H., et al. (1996). Nicotine effects on adults with attention-deficit/hyperactivity disorder. Psychopharmacology, 123, 55-63.
- Liu, S. K., Hwu, H., & Chen, W. J. (1997). Clinical symptom dimensions and deficits on the continuous performance test in schizophrenia. Schizophrenia Research, 25, 211-219.
- Riccio, C. A., Reynolds, C. R., & Lowe, P. A. (2001). Clinical applications of the continuous performance test: Measuring attention and impulsive responding in children and adults. New York: Wiley.
- Rosvold, H., Mirsky, A., Sarason, I., Bransome, E. D., Jr., & Beck, L. H. (1956). A continuous performance test of brain damage. Journal of Consulting Psychology, 20, 343-350.