A COMPARISON BETWEEN ADULTS WITH CONDUCT DISORDER AND NORMAL CONTROLS ON A CONTINUOUS PERFORMANCE TEST: DIFFERENCES IN IMPULSIVE RESPONSE CHARACTERISTICS.
A number of different approaches have been used to study impulsive responding in the laboratory. One of the most popular methods is the operant model, which defines impulsive responding as a choice for a smaller immediate reinforcer over a larger delayed reinforcer. The advantages of this approach are its rich history from which comparisons can be made and the clearly defined responses that are amenable to statistical and theoretical analyses (see Anslie, 1975, for a review). As a result, the operant conditioning perspective remains a popular model for investigating impulsive choice behavior in both humans and nonhumans (e.g., Green & Rachlin, 1996; Rachlin, Raineri, & Cross, 1991). It is important to recognize, however, that impulsivity measured by operant methodology may be only one type of impulsive behavior. Impulsivity is multidimensional, in that it involves several aspects of behavior. For example, Barratt and colleagues (Barratt, 1985; Patton, Stanford, & Barratt, 1995) have proposed a three-factor m odel of impulsivity. Their theory views impulsiveness as a combination of three distinct factors: (a) motor impulsiveness, which is acting on the spur of the moment and having an inconsistent lifestyle; (b) attentional impulsiveness, which is an inability to focus on the task at hand and a propensity for racing and interrupted thoughts; and (c) nonplanning impulsiveness, which is the inability to plan and think carefully, along with a lack of enjoyment of challenging mental tasks. All three factors include dimensions of arousal, information processing, and social learning. Support for the three-factor theory has been provided by a factor analyses of the Barratt Impulsiveness Scale-11 (Patton et al., 1995), a psychometric instrument of impulsivity.
The operant approach to the study of impulsiveness falls primarily into Barratt's (1985) nonplanning factor of impulsiveness. However, in order to study impulsiveness from a different perspective, we used a modified version of the Continuous Performance Test (CPT; Rosvold, Mirsky, Sarason, Bransome, & Beck, 1956), a popular measure of attentional capacity. There have been some studies indicating that attentional deficits, as measured by the CPT, may be related to impulsive behaviors (see discussion below). This approach differs from the operant approach in that it has characteristics that would seem to involve both the attentional and/or motor factors of the three-factor theory of impulsiveness.
The CPT typically requires participants to respond selectively to a series of stimuli (e.g., abstract shapes, letters, or numbers) that are presented briefly and rapidly (usually presentations and delays of less than 500 ms). Since its inception, the CPT has been modified many times, but usually the subject responds to identify a single target stimulus (e.g., "O") or a series of target stimuli (e.g., 'A" followed by "X") which remains constant throughout a testing session. These CPT procedures have been used to identify and characterize attentional deficits in a variety of subject populations including attention deficit disorder in children and adults (Halperin, Wolf, Greenblatt, & Young, 1991), schizophrenics and persons at-risk for schizophrenia (see reviews by Erlenmeyer-Kimling & Cornblatt, 1987; Nuechterlein & Dawson, 1984), as well as persons with learning disabilities (Dykman, Ackerman, & Oglesby, 1979; Swanson, 1981).
Despite the number of variations, the CPT yields several key data, each parameter believed to evaluate different components of attentional processing. Three primary measures are used. First, omission errors (or misses) are failures to respond to a target stimulus. Most researchers agree that these errors represent deficits in sustained attention or vigilance. Second, commission errors (or false alarms) are responses made to stimuli other than target stimuli. Researchers have varied in both the way they have defined commission errors (in various paradigms) and in the interpretation of these errors. Some have defined commission errors as any response to a nontarget stimulus and have suggested elevated frequency of these errors represents impulsive responding (O'Dougherty, Nuechterlein, & Drew, 1984; Sykes, Douglas, Weis, & Minde, 1971). However, others disagree with the breadth of this interpretation (Halperin et al., 1988, 1991; Sostek, Buchsbaum, & Rapoport, 1980; Sykes, Douglas, & Morgenstern, 1973; Wohlber g & Kornetsky, 1973) and define impulsive-type commission errors as a subset of all errors in which incomplete processing of a stimulus similar to the target leads to a rapid, but incorrect response. These more narrowly defined commission errors may be more indicative of impulsive behavior. And third, latencies are the delay between the onset of a stimulus and the participant's response. Latencies provide information about the temporal requirements for processing, indicate how difficult a discrimination is in a particular CPT task, and may also give an indication about how different populations process stimuli before responding (Dougherty et al., 1998; Halperin et al., 1988, 1991).
The present study was primarily concerned with two of the above measures, commission errors and response latencies, because they may be related to impulsivity. Although some authors may disagree with this interpretation, at least part of the skepticism surrounding this interpretation stems from the fact that little effort has been made to externally validate these measures with populations having impulse control problems. The use of commission errors and latencies as measures of impulsivity are consistent with the "fast-guess" model of impulsivity, whereby judgment or accuracy is found to be compromised for the sake of speed (Ollman, 1966; Yellot, 1971). According to this model, making decisions takes a certain amount of processing time and decisions (or responses) made before processing has been completed are impulsive (Sergeant & Sholten, 1985; Sperling, 1984; Yellot, 1971). In other words, commission errors result from failure to withhold responses until a stimulus identity could be completely processed, and are therefore impulsive. When this model is applied to the commission errors and response latencies of the CPT, it follows that impulsive responding would be characterized by shorter latencies and more commission errors (because stimuli have not been completely processed).
CPT research with adults with histories of conduct disorder has been scant. However, other data exist which suggest that men with conduct disorder histories may represent a population likely to evidence CPT performance deficits. Behavioral disinhibition has been extensively researched in the context of EEG responses evoked during go/no-go tasks. Like the CPT, these tasks require subjects to respond to a target stimulus and to withhold responses to an inhibitory stimulus. In particular, in humans the neural processes involved in withholding responses to rapidly presented nontarget stimuli generated a substantial P300 EEG wave compared to brain activity when processing a target stimulus (discussed in Roberts, Rau, Lutzenberger, & Birbaumer, 1994). Interestingly, men with a history of childhood conduct disorder (CD) have been shown to have blunted frontal cortex electroencephalographic (P300) activity to nontarget stimuli (Bauer, 1997). Moreover, in these men, the number of conduct disorder symptoms prior to ag e 15 correlated negatively with frontal P300 amplitude during a task similar to the CPT used in the present study. Additionally, P300 amplitude has been shown to be inversely proportional to Barratt Impulsiveness Scale scores (Barratt, 1965; Branchey, Buydens-Branchey, & Horvath, 1993). This literature suggests that persons with histories of DSM-III-R-defined conduct disorder represent an ideal population in which to determine the external validity of CPT commission errors and latency as measuring impulsiveness.
Some studies have found increased commission error rates and shorter latencies in other impulsive populations. For example, Halperin et al. (1991) found certain types of commission errors on the CPT ("A followed by X" version) to be more frequently made by children with attention deficit disorder (compared to normal controls), and that reaction times (latencies) were shorter in duration. Commission error rates are also elevated in nonalcoholic subjects having a family (parental) history of alcohol-related problems (Miller, 1984; unpublished dissertation) or history of DWI arrests (Koch & Morguet, 1985). These studies indirectly support the idea that commission error rates and latencies are perhaps related to impulsive or disinhibited behaviors/traits. To investigate this possibility, we used a recently developed OPT procedure (The Immediate Memory Task and Delayed Memory Task, IMT/DMT; Dougherty et al., 1998).
Specifically, the main purpose of this study was to compare CPT performance between 15 controls and 15 adults with a self-reported childhood/adolescent history of behavior that would have met criteria for conduct disorder (CD). We chose CD individuals because impulsiveness has been an important factor in explaining this common childhood condition (Loeber, 1990; Quay, 1988; Schachar, Tannock, & Logan, 1993). Gorenstein and Newman (1980) included CD in a constellation of "syndromes of disinhibition" which they proposed to have a common foundation of disinhibited, impulsive behaviors. Because externalized disinhibition is a main feature of CD (Windle & Windle, 1993), we hypothesized that adults with a history of childhood CD would have maintained several impulsive traits, and that these would be evidenced in the IMT/DMT performance in the following ways: (1) higher rates of false alarm commission errors, (2) shorter response latencies, (3) lower stimulus discriminability, and (4) liberal responding strategy (me asured by the signal detection parameter Beta). Furthermore, we hypothesized that these measures would be related to scores on a self-reported measure of impulsivity, the Barratt Impulsivity Scale (BIS-11; Patton et al., 1995).
The conduct disorder group (n = 15) consisted of persons who met criteria for childhood/adolescent conduct disorder (CD) with onset before age 15. Diagnosis was made using the Antisocial Personality Disorder (ASPD) module of the Structured Clinical Interview for DSM-III-R Axis II Disorders (SCID-II; Spitzer, Williams, Gibbon, & First, 1990; at the time of the study the SCID-II for DSM-IV was not yet available). In the CD group, 2 subjects were Caucasian, 1 was Hispanic, and 12 were African-American. The mean age and years of education for this group were 30.5 [pm] SD 7.4, and 11.0 [pm] SD 1.36, respectively. Subjects in the control group (n = 15) were recruited to demographically resemble the CD group with respect to age, education, and race, with the highest matching priorities placed on age and education. In the control group, 6 were Caucasian, 3 were Hispanic, and 6 were African-American. The mean age and years of education for this group were 27.6 [pm] SD 6.4 and 13.6 [pm] SD 1.9, respectively.
Both groups were recruited from the community using newspaper advertisements. Prospective participants were screened in a brief telephone interview and callers who reported no psychiatric history were invited to a subsequent on-site interview. The on-site interview consisted of a medical history interview and a Structured Clinical Interview for DSMIV Axis I disorders (First, Spitzer, Gibbon, & Williams, 1996), as well as the SCID-II for DSM-III-R (Spitzer, Williams, Gibbon, & First, 1990). Applicants were not admitted for study if the interviews indicated a significant current medical condition or the presence of current or past Axis I disorder other than substance abuse.
For both study groups, participation in the research study was voluntary and informed consent was obtained. All methods and consent forms were approved by the University of Texas-Houston Institutional Review Board. Subjects were told that the study was concerned with memory and motor performance; no mention was made of impulsivity, either verbally or in the consent forms. After completing the last testing session and other psychometric measures subjects were paid approximately $45.
Testing schedule. Each subject participated throughout 1 day. Subjects began testing at 0800 hrs, at which time they submitted breath and urine samples for tests of current alcohol or other drug use; all subject samples tested negative. Subjects then completed a total of six CPT sessions (using the Immediate Memory Task / Delayed Memory Task, described below), with each of these sessions beginning at 0830, 0930, 1030, 1130, 0130, and 0330. Lunch was provided at noon.
Physical apparatus. Behavioral testing was performed in 1.8-m x 1.8m sound-insulated chamber equipped with an IBM-compatible color monitor, a ventilation fan (which provided masking noise), and a two-button computer mouse. A computer located in an adjacent room controlled and monitored experimental events.
Continuous Performance Test
A modification of the original CPT, the Immediate Memory Task / Delayed Memory Task (IMT/DMT; Dougherty et al., 1998) was used for this investigation. In each testing session, the two task conditions (IMT and DMT) alternated in 5-mm testing blocks with a 30-s rest period preceding each block. The IMT was always first and alternated with the DMT, with each presented twice. As a result, testing sessions lasted exactly 22 minutes. Both conditions are described below.
Immediate Memory Task (IMT). This task was designed to measure brief attentional capacity. A series of 5-digit numbers (e.g., 73021) were displayed on the monitor for 0.5 s and separated by a 0.5-s blackout period. Each of the digits measured 2.0 cm wide X 3.3 cm high, and the numbers were presented on the computer monitor in black on a white background.
There were several distinct types of stimuli presented and types of responses that could be made. Subjects were instructed to respond on the computer's left mouse button when a 5-digit number (the target stimulus) appeared that was exactly like the preceding stimulus. The probability of a target stimulus was 33%. A response made while a target stimulus appeared on the monitor, or made before the next stimulus appeared (1.0 s total), was recorded as a correct detection (hit). A failure to respond to a target stimulus was recorded as an omission error (miss). In addition to target stimuli, there was a 33% probability that a catch stimulus would appear. A catch stimulus was a number that differed from the preceding number by one of the five digits (its position and value was determined randomly). Responses (errors) made to catch stimuli were considered commission errors (false alarms). The remaining stimuli, novel stimuli (numbers), which were not either target or catch trials were called filler stimuli and res ponses made to these stimuli were categorized as "other errors." [Note: A novel stimulus always followed a target or catch trial.]
Delayed Memory Task (DMT). This task was designed to measure a subject's ability to retain a stimulus and identify it after a longer delay (compared to the IMT above). The primary difference between this task and the IMT was that all stimuli (including target, catch, and filler) were separated by the number "12345:" which was repeated three times at the same rate and duration as all other stimuli. For example, one possible sequence involving a target stimulus would be: 39863, 12345, 12345, 12345, 39863. Subjects were instructed to ignore the "12345" and only to remember and identify stimuli separated by the series of "12345" numbers. These stimuli, "12345," were designated as distracter stimuli. Presenting these distracter stimuli allowed us to increase task difficulty, to control for rates of visual stimulus presentation, and to improve task sensitivity.
In addition to the number of responses made to each type of stimulus, response latencies for correct detections and commission errors were recorded. The time between onset of stimulus presentation and a response was recorded in milliseconds.
Barratt Impulsiveness Scale (BIS). The BIS (Version 11, Patton et al., 1995) is a self-completed, 30-item questionnaire on which participants rate their frequency of several common impulsive behaviors and traits (e.g., "I do things without thinking" ) or nonimpulsive behaviors/traits (scored in reverse; e.g., "I am self-controlled") on a scale from 1 ("rarely/never") to 4 ("almost always/always"). In order to examine the relationship between impulsivity and CPT performance, several correlations were made between BIS scores and parameters of the CPT.
Wisconsin Card Sorting Test (WCST). The WCST (Computer Version-2; Curtiss & Tuttle, 1993; Heaton, Curtiss, Tuttle, & PAR Staff, 1993) is a neuropsychological test that purportedly measures deficits in frontal lobe or "executive functioning." This test involves memory, abstraction, and sensitivity to performance feedback. A subject is instructed to match a card to one of a set of four stimulus cards based on an unknown sorting dimension (color, shape, or number of symbols) which changes periodically during testing. To determine whether our two groups differed in WCST performance, we compared percentage of errors, percentage of perseverative responses (continuing to match by an incorrect sorting principle), and number of categories completed.
Beck Depression and Anxiety Inventories. As part of our screening procedure subjects completed the Beck Depression (BDI; Beck, Rush, Shaw, & Emery, 1979) and the Beck Anxiety Inventories (BAI; Beck, Brown, Epstein, & Steer, 1988). These instruments were only used to quickly identify persons with possible psychiatric problems and to compare subclinical symptoms between our two groups.
The BDI is a commonly used self-report instrument which provides a numerical index of symptomatology associated with depression. On the BDI, the rater selects from a series of numerically graded statements pertaining to mood, self-worth, suicidal thoughts, and somatic symptomatology during the previous week. On the BAI, the respondent assigns a rating from 0-3 to the severity of symptoms of anxiety, such as fear of dying, breathing difficulty, and unsteadiness.
Data Analysis. Response latencies to target and catch stimuli, percentages of correct detections (responses to target stimuli), percentages of commission errors (responses to similar but not identical stimuli), discriminability (d'), and Beta values (description of d' and Beta parameters appears below in results) were analyzed in repeated-measures analyses of variance in a 2 x 6 x 2 design, with task type (IMT vs. DMT) and session (1-6) as within-subject factors, and group (CD vs. control) as the between-subject factor. In the latency analysis, stimulus type (target vs. catch) was also included as a within-subject factor. Finally, response latencies, commission errors (false alarms), correct detections, and BIS scores were correlated using Spearman correlation. Other between-group differences in demographic variables and psychometric instruments were assessed using independent t tests or chi-square tests (where appropriate). In previous studies we have failed to detect a within-session effect of block (time) on any parameter of IMT and DMT performance (e.g., Dougherty et al., 1998). For this reason, data from the two blocks within each testing session were averaged. Additionally, because of the near absence of responses to random and/or distracter "12345" stimuli, statistical analysis of rates of these types of errors was not appropriate, and brief descriptive data are presented instead.
All correlations were conservatively calculated using Spearman's nonparametric tests because of the number of correlations being done.
Independent t tests and chi-square tests indicated that the two subject groups did not significantly differ (p [greater than] .05) in age, gender, or racial composition. The control group averaged 2.6 years more education than the CD group, which was statistically different [t(28) = 4.28, p [less than] .001]. A Pearson's chi-square test showed the incidence of past substance abuse/dependence between the CD (n = 7) and control groups (n = 3) was not significantly different (p[greater than] .05). Independent t tests showed commission error rates and latencies for IMT and DMT conditions did not differ (p [greater than] .05) between subjects with or without a history of substance abuse/dependence.
BAI and BDI scores indicated no significant between-group differences in subclinical symptoms of anxiety or depression. Additionally, using independent t tests we found no group differences in any component of WCST performance (p [greater than] .40 for % errors, % perseverative responses, and number of categories completed). Psychometric measures (WCST, BIS, BDI, and BAI) were not obtained from 2 subjects.
Correct detections (hits). Subjects made responses to the majority of target stimuli presented in both the IMT and DMT, making a few more correct detections on the IMT than on the DMT [main effect of task type F(1, 28) = 7.504, p = .0111. These data appear in Figure 1. The main effect of session showed a slight, but significant, increase in correct detections across sessions [F(5, 140) = 2.632, p = .026]. The main effect of group was not significant, indicating that the groups did not differ in the rates of correct detections [F(1, 28) = 0.718, p = .404].
All higher-order interaction effects were not significant: task type x subject group [F(1, 28) = 0.013, p = .909], session x subject group [F(5, 140) = 0.70, p = .622], task type x session [F(5, 140) = 0.88, p = .496], and task type x session x subject group [F(5, 140) = 0.278, p = .924].
Commission errors (false alarms). The CD subjects made approximately twice as many commission errors (responses made to similar but not identical stimuli) than controls in both the IMT and DMT [F(1, 28) = 28.915, p [less than] .0001]. These data appear in Figure 2. The main effects of task type [F(1, 28) = 1.485, p = .233] and session [F(5, 140) = 1.066, p = .381] on commission error rates were not significant.
All other effects were not significant: task type x subject group [F(1, 28) = 0.102, p = .752], session x subject group [F(5, 140) = 0.820, p = .537], task type x session [F(5, 140) = 1.765, p .124], and task type x session x subject group [F(5, 140) = 0.856, p = .512].
Discriminability (d). The signal detection parameter d', the measure of discriminability between signal and noise (i.e., between target and catch stimuli), was calculated and appears in the top panels of Figure 3. For d', higher values indicate better discriminability. These values were calculated using standardized formulas taken from Gescheider (1985, p. 97). The main effect of memory task type was significant, indicating that discrimination was more difficult in the DMT compared to the IMT [F(1, 28) = 31 .950, p [less than] .0001]. Also, CD subjects had d' values that were significantly lower than controls for both tasks [main effect of group was F(1, 28) = 36.418, p [less than] .0001].
All other effects and interactions were not significant: session [F(5, 140) = 0.521, p = .760], session x subject group [F(5, 140) = 0.431, p = .83], session x memory task type [F(5, 140) = 0.497, p = .778], subject group x memory task type [F(1, 28) = 1.863, p = .183], and session x memory task type x subject group [F(5, 140) = 0.472, p = .797].
Beta. An identical analysis was performed using the signal detection parameter Beta. These data appear in the bottom panels of Figure 3. Beta scores give an indication of how subjects' response criteria are distributed (e.g., either conservative or liberal). Higher scores indicate a more conservative response strategy. These values were calculated using standardized formulas taken from Gescheider (1985, p. 97). Subjects with histories of CD did not respond more liberally than controls to CPT stimuli in general. Beta was not significantly higher in controls compared to CD subjects [F(1, 28) = 1.059, p = .312]. All subjects were more conservative in responding to DMT stimuli (higher beta values) than IMT stimuli [memory task type main effect F(1,28) = 23.824, p [less than] .0001].
All other effects were not significant: session [F(5,1 40) = 1.303, p = .266], session x subject group [F(5,140) 1.127, p = .349], session x memory task type [F(5, 140) = 0.964, p = .442], subject group x memory task type [F(1 ,28) = 1.31, p = .261], and session x memory task type x subject group [F(5,140) = 1.578, p= .170].
Response latencies. As shown in Figure 4, response latencies were shorter for the CD subjects compared to controls across both memory tasks [main effect of group F(1, 28) = 4.211, p = .049]. Across all subjects, latencies were shorter in the MT than the DMT [main effect of memory task type [F(1, 28) = 56.893, p [less than].0001]. The memory task type x group interaction was not significant [F(1, 28) = 0.349, p = .559]. In addition, there was also a main effect of session [F(5, 140) = 2.651, p = .025], with latencies becoming shorter in succeeding sessions. This effect was not specific to group or to task [session x group F(5, 140) = 1.262, p = .283; session x task F(5, 140) = 0.813, p = .542].
There was no main effect of stimulus type [target vs. catch; F(1, 28) = 2.582, p = .119] on latency, but the data suggested higher order interactions of stimulus by task type. For example, although response latencies to catch and target stimuli in the IMT were strikingly similar, there was a trend for longer latencies for catch stimuli compared to target stimuli in the DMT [stimulus x task type F(1, 28) = 3.581, p = .068]. Moreover, the session by stimulus interaction [F(5, 140) = 3.171, p = .009] indicated that subjects took longer to respond to catch stimuli across sessions, and the session x stimulus x memory task type interaction revealed a trend for this to be specific to the DMT [F(5, 140) = 2.181, p = .069]. With respect to between-group differences, a significant stimulus x group interaction [F(1, 28) = 4.750, p = .037] suggested that CD subjects (but not controls) had longer latencies to catch stimuli than target stimuli. No other higher order interactions were significant.
Commission errors corresponded to shorter response times. IMT commission error rates negatively correlated with mean IMT latencies (r= -.620, p [less than] .0002), as did DMT commission error rates with mean DMT latencies (r = -.740, p [less than] .0001). These Spearman correlations appear in the top panels of Figure 5. To determine whether or not the differences in latencies could be responsible for the differences in commission error rates between our two groups, we reanalyzed the between-group differences with latency added as a covariate. In this analysis we found that adjusted mean commission error rates between the two groups were still significantly different in the IMT [group main effect F(1, 27) = 22.044, p = .0001] and DMT [F(1, 27) = 24.186, p [less than] .0001].
Other errors. Responding to random (novel) or distracter stimuli ("12345" in the DMT) was rare in both groups, suggesting that all subjects could easily discriminate stimuli which differed noticeably from target stimuli. These error responses were made to less than 1% of these stimuli. Collectively, rates of these sporadic responses were not suitable for an analysis of variance.
Psychometric measures. Barratt Impulsiveness Scale (BIS-11; Patton et al., 1995) total scores were significantly higher for CD subjects (Mean = 67.3, SD = 7.4) than for controls [M = 57.1, SD = 13.9; t(26) = 2.43, p = .0225]. The BIS nonplanning subscale scores were also higher for the CD subjects (M = 30.9, SD = 6.9) than for controls [M = 23.07, SD = 6.6; 1(26) = 3.081, p = .0048]. Neither the motor nor the attentional BIS subsoales were significantly different between groups.
BIS scores were correlated with IMT/DMT commission error rates (averaged across the six sessions; Figure 5) to determine how impulsivity in the IMT/DMT (conceptualized here as high commission error rates and shorter latencies) related to an external measure of impulsivity. Across all subjects, 615 scores correlated positively with IMT commission error rates (r= .42, p = .025) but not with DMT commission error rates (r = .19, p = .329). Correlations between the BIS and response latencies (target and catch stimuli latencies each averaged across sessions) showed no significant relationship for either the IMT (r = -.19, p = .341) or the DMT(r = .01, p= .926).
We hypothesized that adults whose childhood/adolescent behavior met the Structured Clinical Interview for DSM-III-R (Spilzer et al., 1990) criteria for Conduct Disorder (CD) would perform in a manner consistent with what previous CPT researchers have proposed as impulsive response characteristics. Compared to control subjects, CD subjects had significantly (a) higher rates of commission errors (i.e., more false alarms) on both the IMT and DMT; (b) shorter response latencies (approximately 100 ins); and (c) lower discriminability (d') between target and catch stimuli.
Not only were significant differences observed in IMT/DMT behavior between the two groups, but the 815 total and nonplanning subtrait scores for the CD group were significantly higher than the control group. The 815 total score also significantly correlated with IMT, but not DMT commission error rates, which is very likely a reflection of fundamental differences in the tasks. The IMT requires fast, repetitive discriminations and is more closely related to previous CPT tasks, while the DMT is a more difficult task with the inclusion of delays and distracter stimuli. The greater difficulty of the DMT has probably contributed to an increase in variability which reduces the likelihood of finding a meaningful relationship with the BIS scores. However, overall, the data seem to provide further support for the contention that CPT commission errors may be related to impulsivity.
Our findings are especially interesting given that there exists data indicating that electrophysiological responses to commission errors (on tasks similar to IMT/DMT) differ between various populations. For example, men with histories of conduct disorder prior to age 15 have been shown to have blunted frontal cortex electroencephalographic (P300) activity in a task similar to the one used in our study (Bauer, 1997). The P300 wave is generally considered an indication of attention or interest. Interestingly, Bauer reported that P300 amplitude decrements were not evident when subjects were deciding whether to respond to target stimuli, but rather the P300 deficit was evident only when subjects were presented with stimuli analogous to the CPT catch stimuli of this study. Moreover, these P300 decrements were only found in men with conduct disorder history prior to age 15. Additionally, P300 amplitude has been shown to be inversely proportional to Barratt Impulsiveness Scale scores (Barratt, 1965; Branchey et al. , 1993). This neurophysiological data, together with our present findings, suggest that conduct disorder prior to age 15 may predict discrimination and response inhibition deficits as an adult.
Although the studies described above collectively support the contention that commission errors may be an indicator of impulsive behavior, some caution should be exercised. Only a handful of studies (mostly with children) have examined the relationship between commission errors and impulsivity. Other interpretations of the data are plausible. For example, group differences in latencies and commission error rates may be the result of attentional, memory, or perhaps sensory deficits of adults with a history of conduct disorder. These factors could explain why the CD group showed longer commission error latencies compared to the control group. While commission error rates and the overall CD group latencies were significantly different than controls for both IMT and DMT conditions as we hypothesized, the specific commission error latency group difference did not support our hypothesis.
Some of the parameters of the IMT/DMT procedures may have maximized differences in commission error rates and are worth noting. First, previous studies have often used liberal definitions of commission errors, such as defining these errors as any response made to some stimulus other than the target stimulus. We suggest that defining commission error responses in this manner limits the interpretation of the data. It cannot be determined whether the errors are made because the subject cannot discriminate between signal (target) and noise (other stimuli), or whether the errors are a result of a population's elevated arousal levels (e.g., ADHD children). Procedures that more conservatively define commission errors allow for a comparison of two types of error rates between stimuli which are both similar to, and different from, the target stimuli. Differential responding to these two error types would provide clues to whether responses were due to discriminability or to higher arousal levels.
Second, increasing the task difficulty compared to earlier CPT versions provided more distinctive differences between groups. The magnitude of the differences in commission error rates between groups was large. Previous studies which used easier OPT procedures engendered relatively low rates of commission errors, and so differences between normal and patient populations were small. Generally, commission error rates were less than 10%, with differences between populations being less than 5%. Previous research indicates that detecting subtle differences in attention in adult populations appears to be dependent on the difficulty of the OPT procedure used; easier tasks are not as sensitive as difficult tasks (for a discussion see Cornblatt & Keilp, 1994). Additionally, in order to better discriminate between populations (especially between adult populations) the task should be of sufficient difficulty to produce sufficient numbers of errors to allow for a better separation between populations. A higher commission error rate may be more sensitive to the introduction of independent variable manipulations.
There are a number of obvious routes that future research, examining the relationship between commission errors and impulsivity, should take. First, the effects of drugs of abuse and medications on commission errors should be studied. This can be approached from two different perspectives: (a) studying the effects of medications (e.g., stimulants) on commission error rates, and (b) how drugs of abuse (e.g., alcohol) may increase commission error rates. The expectation, from an impulsivity approach, would be that alcohol would increase impulsive responding, and stimulants (at therapeutic doses) would decrease impulsive responding. In fact, a recent study with 18 social alcohol drinkers (Dougherty et al., 1999) found that small doses of alcohol (achieving BACs of approximately 0.035) increased commission error rates, and changes in performance were specific to these errors. Second, commission error rates should be studied by systematically examining the effects of contingencies on performance to determine whet her or not differences between populations can be minimized. And third, other neuropsychological measures should be taken in an attempt to determine the nature of the deficits between populations and to explore whether other interpretations of data could account for these observed differences.
This research was supported by grants from the National Institute on Alcoholism and Alcohol Abuse (AA-10095 & AA-10828) and the National Institute on Drug Abuse (DA-08425).
ANSLIE, G. (1975). Specious reward: A behavioral theory of impulsiveness and impulse control. Psychological Bulletin, 82(4), 463-496.
BARRATT, E. S. (1965). Factor analysis of some psychometric measures of impulsiveness and anxiety. Psychological Reports, 16, 547-554.
BARRATT, E. S. (1985). Impulsiveness subtraits: Arousal and information processing. In J. T. Spence & C. E. Izard (Eds.), Motivation, Emotion, and Personality (pp. 137146). North-Holland, NY: Elsevier Science Publishers B.V.
BAUER, L 0. (1997). Frontal P300 decrements, childhood conduct disorder, family history, and the prediction of relapse among abstinent cocaine abusers. Drug and Alcohol Dependence, 44,1-10.
BECK, A. T., BROWN, G., EPSTEIN, N., & STEER, R. S. (1988). An inventory for measuring clinical anxiety: Psychometric properties. Journal of Consulting and Clinical Psychology, 56, 893-897.
BECK, A. T., RUSH, A. J., SHAW, B. F, & EMERY, G. (1979). Cognitive theory of depression. New York: The Guilford Press.
BRANCHEY, M. H., BUYDENS-BRANCHEY, L, & HORVATH, T.B.(1993). Event related potentials in substance abusing individuals after long term abstinence. American Journal of Addiction, 2, 141-148.
CORNBLATT, B. A., & KEILP, J. G. (1994). Impaired attention, genetics, and the pathophysiology of schizophrenia. Schizphrenia Bulletin, 20, 31-46.
CURTISS, G., & TUTTLE, K. (1993). Wisconsin card sorting test: Computer version 2 [Program Manual]. Odessa, FL. Psychological Assessment Resources, Inc..
DOUGHERTY, D. M., MOELLER, F. G., STEINBERG, J. L., MARSH, D. M., HINES, S. E., & BJORK, J. M. (1999). Alcohol increases commission error rates for a continuous performance test. Alcoholism: Clinical and Experimental Research, 23(8), 1342-1351.
DOUGHERTY, D. M., STEINBERG, J., WASSEF, A. A., MEDEARIS, D., CHEREK, D. R., & MOELLER, F. G. (1998). Immediate versus delayed visual memory task performance among schizophrenic patients and normal controls. Psychiatry Research, 79,255-265.
DYKMAN, R. A., ACKERMAN, P. T., & OGLESBY, D. M. (1979). Selective and sustained attention in hyperactive, learning-disabled and normal boys. Journal of Nervous and Mental Disease, 167, 288-297.
ERLENMEYER-KIMLING, L., & CORNBLATT, B. (1987). High-risk research in schizophrenia: A summary of what has been learned. Journal of Psychiatric Research, 21,401-411.
FIRST, M. B., SPITZER, R. L., GIBBON, M., & WILLIAMS, J. B. W. (1996). Structured Clinical Interview for DSM-IV Axis I Disorders: Non-patient Edition (SCID-NP). Biometrics Research Department, New York State Psychiatric Institute, NY.
GESCHEIDER, G. A. (1985). Psychophysics: Method theory, and application (p. 97). Hillsdale, NJ: Lawrence Erlbaum.
GORENSTEIN, E. E., & NEWMAN, J. P. (1980). Disinhibitory psychopathology: A new perspective and a model for research. Psychological Review, 87, 301-315.
GREEN, L., & RACHLIN, H. (1996). Commitment using punishment. Journal of the Experimental Analysis of Behavior, 65(3): 593-601.
HALPERIN, J. M., WOLF, L., GREENBLATT, E. R., & YOUNG, G. (1991). Subtype analysis of commission errors on the continuous performance test in children. Developmental Neuropsychology, 7, 207-217.
HALPERIN, J. M., WOLF, L. E., PASCUALVACA, D. M., NEWCORN, J. H., HEALEY, J. M., O'BRIEN, J. D., MORGANSTEIN, A., & YOUNG, J. G. (1988). Differential assessment of attention and impulsivity in children. Journal of American Academy of Child and Adolescent Psychiatry, 27, 326-329.
HEATON, R. K., CURTISS, G., TUTTLE, K., & PAR STAFF (1993). Wisconsin Card Sorting Test: Computer, (Version 2) [Computer Software]. Odessa, Fl: Psychological Assessment Resources, Inc..
KOCH, M., & MORGUET, M. (1985). Investigation in the use of a vigilance test in psycho-medical assessment. Blutalkohol, 22, 391-396.
LOEBER, R. (1990). Development and risk factors of juvenile antisocial behavior and delinquency. Clinical Psychology Review, 10, 1-41.
MILLER, R. C. (1984). The effect of low doses of alcohol on human behavior. Unpublished doctoral thesis, University of Texas at Austin.
NUECHTERLEIN, K. H., & DAWSON, M. E. (1984). Information processing and attentional functioning in the developmental course of schizophrenic disorders. Schizophrenia Bulletin, 10, 160-203.
OLLMAN, R. (1966). Fast guess in choice reaction time. Psychonomic Sciences, 6, 155-156.
O'DOUGHERTY, M., NUECHTERLEIN, K. H., & DREW, B. (1984). Hyperactive and hypoxic children: Signal detection, sustained attention, and behavior. Journal of Abnormal Psychology, 93, 178-191.
PATTON, J. M., STANFORD, M. S., & BARRATT, E. S. (1995). Factor structure of the Barratt Impulsiveness Scale. Journal of Clinical Psychology, 51, 768-774.
QUAY, H. C. (1988). The behavioral reward and inhibition system in childhood behavior disorder. In L. M. Bloomingdale (Ed.), Attention Deficit Disorder (Vol. 3, pp. 176-186). New York: Spectrum.
RACHLIN, H., RAINERI, A., & CROSS, D. (1991). Subjective probability and delay. Journal of the Experimental Analysis of Behavior, 55(2), 233-244.
ROBERTS, L E., RAU, H., LUTZENBERGER, W., & BIRBAUMER, N. (1994). Mapping P300 waves onto inhibition: Go/No-Go discrimination. Electroencephalography and Clinical Neurophysiology, 92, 44-55.
ROSVOLD, H. E., MJRSKY, 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.
SCHACHAR, R. J., TANNOCK, R., & LOGAN, G. (1993). Inhibitory control, impulsiveness, and attention deficit hyperactivity disorder. Clinical Psychology Review 13, 721-739.
SERGEANT J. A., & SHOLTON, C. A. (1985). On resources strategy limitations in hyperactivity: Cognitive impulsivity reconsidered. Journal of Child Psychology and Psychiatty 25, 809-819.
SOSTEK, A. J., BUCHSBAUM, M. S., & RAPOPORT, J. L. (1980). Effects of amphetamine on vigilance performance in normal and hyperactive children. Journal of Abnormal Child Psychology, 8, 491-500.
SPERLING, G. (1984). A unified theory of attention and signal detection. In R. Parasuraman & D. R. Davies (Eds.), Varieties of attention (pp. 103-181). New York: Academic.
SPITZER, R. L., WILLIAMS, J. B. W, GIBBON, M., & FIRST, M. B. (1990). Structured clinical interview for DSM-III-R - personality disorders (SCID-II, Version 1.0). Washington, DC: American Psychiatric Press.
SWANSON, L (1981). Vigilance deficit in learning disabled children: A signal detection analysis. Journal of Child Psychology and Psychiatry, 22, 393-399.
SYKES, D. H., DOUGLAS, V. I., & MORGENSTERN, G. (1973). Sustained attention in hyperactive children. Journal of Child Psychology and Psychiatty 14, 213-220.
SYKES, D. H., DOUGLAS, V. I., WEISS, G., & MINDE, K K. (1971). Attention in hyperactive children and the effect of methyiphenidate (Ritalin). Journal of Child Psychology and Psychiatry, 12, 129-139.
WINDLE, M., & WIN DLE, R. C. (1993). The continuity of behavioral expression among disinhibited and inhibited childhood subtypes. Clinical Psychology Review, 13, 741-761.
WOHLBERG, G. W., & KORNETSKY, C. (1973). Sustained attention in remitted schizophrenics. Archives of General Psychiatty, 28, 533-537.
YELLOT, J. I. (1971). Corrections for fast guessing and speed-accuracy-tradeoff in choice reaction time. Journal of Mathematical Psychology 8,159-199.
|Printer friendly Cite/link Email Feedback|
|Author:||DOUGHERTY, DONALD M.; BJORK, JAMES M.; MARSH, DAWN M.; MOELLER, F. GERARD|
|Publication:||The Psychological Record|
|Date:||Mar 22, 2000|
|Previous Article:||"A Guide to Treatments That Work".|
|Next Article:||RISK AVERSION IN HUMAN SUBJECTS UNDER CONDITIONS OF PROBABILISTIC REWARD.|