Laboratory and questionnaire assessment of the BAS and BIS

Gray’s reinforcement sensitivity model and child psychopathology: laboratory and questionnaire assessment of the BAS and BIS

Craig R. Colder

Gray’s reinforcement sensitivity model of personality posits three orthogonal dimensions that originate from independent neuropsychological systems (Gray, 1987). The most widely studied of these systems are the Behavioral Approach System (BAS) and Behavioral Inhibition System (BIS). The BAS is thought to be a reward or approach system that responds to positive incentives by activating behavior. The BIS inhibits behavior in the presence of cues signaling that aversive consequences will follow should a response be made. Inhibition can take the form of passive avoidance in response to potential punishment or extinction in response to the nonoccurrence of expected reward. This model emphasizes reactivity to conditioned cues, and conditioning is dependent on environmental input. Socialization involves controlling contingencies (reward and punishment) to help children learn appropriate behavior, and consequently affects input from the environment to which the BIS and BAS respond. Thus, Gray’s model is of great interest to the field of developmental psychopathology because it offers an opportunity to integrate multiple levels of influence such as biologically based individual differences with contextual risk and protective factors into a theoretical framework for studying pathways to childhood psychopathology.

Quay (1988) was among the earliest proponents of applying Gray’s model to the study of child psychopathology, and several researchers have followed his lead to develop etiological theories of child behavior problems that incorporate the BAS and BIS (e.g., Beauchaine, 2001; Frick, 1998). Research suggests that BAS dominance increases risk for externalizing problems (e.g., Iaboni, Douglas, & Baker, 1995; O’Brien & Frick, 1996; Oosterlaan & Sergeant, 1998; Newman, Widom, & Nathan, 1985), whereas BIS dominance increases vulnerability for internalizing problems (Biederman, Rosenbaum, Hirshfeld, & Faraone, 1990; Caspi, 2000; Kagan, Snidman, Zentner, & Peterson, 1999; Turner, Beidel, & Epstein, 1991). Dominance of one system over the other can be attributed to several underlying mechanisms, and identifying the specific mechanism can help refine etiological theories as well as interventions. This issue is most apparent when etiological theories of childhood Attention Deficit Hyperactivity Disorder (ADHD) and Conduct Disorder are considered.

BAS dominance can occur because of a strong BAS or a weak BIS. Quay (1993, 1997) argued that ADHD is linked to a weak BIS and antisocial behavior (specifically Undersocialized Conduct Disorder) is linked to a strong BAS. Thus, the diathesis of dominant BAS is expected to vary across specific forms of externalizing behavior problems. Performance on a variety of lab tasks have been used to assess approach and inhibition, and findings suggest mixed support for Quay’s hypothesis (Daugherty & Quay, 1991; Matthys, van Goozen, de Vries, Cohen-Kettenis, & van Engeland, 1998; Oosterlaan & Sergeant, 1998; Shapiro, Quay, Hogan, & Schwartz, 1988).

Although there is mixed support for Quay’s hypothesis, the tasks used in past research may not be good measures of the BAS or BIS [this issue is discussed below in more detail]. Moreover, it may be important to take into account comorbidities between externalizing and internalizing problems when considering the role of the BIS and BAS in the development of child psychopathology. In two studies Frick and colleagues (O’Brien & Frick, 1996; O’Brien, Frick, & Lyman, 1994) found that conduct disorder was associated with approach behavior on a response perseveration task, but this was only true when Conduct Disorder occurred in the absence of an Anxiety Disorder. This suggests that BAS functioning is germane to a subgroup of antisocial children who exhibit very low levels of anxiety symptoms. Although the relation between the BIS/BAS and externalizing problems appears to be complex. Quay’s hypothesis suggests that it is not enough to just evaluate BIS or BAS dominance. It appears that it is important to determine the strength of each system when considering how the BIS/BAS are associated with externalizing problem behavior.

Different mechanisms may also account for BIS dominance and the etiology of internalizing problems. That is, BIS dominance may emerge because of a strong BIS or a weak BAS. Fowles (1988) has argued that depression may be associated with weak BAS functioning, whereas anxiety may be associated with strong BIS functioning. A notable paucity of research has examined associations between the BIS and BAS and childhood internalizing problems. However, research with adults suggests that depression in adults commonly occurs along with anxiety symptoms, and anxiety often occurs without depression (Di Nardo & Barlow, 1990; Sanderson, Di Nardo, Rapee, & Barlow, 1990). This suggests that anxiety and depression may share a common cause, but that a unique cause may contribute to the emergence of depression. Both anxiety and depression are associated with elevated levels of negative affect, whereas only depression is associated with low levels of positive affect (Clark & Watson, 1991). Negative affect may be an emotional concomitant of the BIS, whereas positive affect may be an emotional concomitant of the BAS. Moreover, some experimental research has shown that adults with dysphoria are less responsive to reward (Henriques, Glowacki, & Davidson, 1994). These findings suggest that a strong BIS may be associated with anxiety symptoms and a strong BIS in combination with a weak BAS may be associated with depression. This suggests the importance of evaluating the strength of each system. Overall, it appears that child psychopathology researchers not only need to establish whether the BIS or BAS is dominant, but they also need measures that allow them to assess the strength of each system.

Most research applying Gray’s model to child psychopathology assesses the BAS and BIS using measures not directly derived from the model. This is true of both questionnaire and laboratory measures. One common measurement approach is to select questionnaire items from instruments developed to assess childhood behavior problems. Items assessing externalizing problems are selected to assess BAS functioning, and items assessing internalizing symptoms are selected to assess BIS functioning (e.g., Masse & Tremblay, 1997; Tremblay, Pihl, Vitaro, & Dobkin, 1994; Walker et al., 1991). This is problematic in part because the BIS and BAS may be differentially related to narrowband dimensions of behavior problems and thus sampling items from the broadband dimensions may obscure relations between the BIS/BAS and behavior problems. Other studies select items from childhood temperament measures that were developed to assess individual differences in anxiety and impulsivity/activity level (e.g., Colder, Mott, & Berman, 2001; Martin & Bridger, 1999). Items from behavior problem or temperament scales do not directly assess reinforcement sensitivity, but rather represent behavioral correlates of responses to reward, and to punishment and extinction. For example, activity level is expected to be related to BAS activation, but typical items assess a preference for high energy activities (e.g., my child prefers active games involving running and jumping rather than games in which he/she must sit) or behavior that is not goal oriented (e.g., my child seems to have difficulty sitting still) rather than assessing behavioral responses to reward, or lack of responsivity to punishment or extinction. Similarly, anxiety is thought to assess BIS functioning, but items often assess cognitive aspects of anxiety (e.g., my child worries about most things), or negative emotionality, or general mood lability rather than fear or inhibition in response to potential punishment. The consequence of using items not directly derived from Gray’s model is that the core concepts of sensitivity to reward, punishment, and extinction are not adequately represented, and thus it is hard to assess the independent strength of the BIS and BAS.

Typical laboratory tasks measure behavioral control by assessing response perseveration (e.g., Daugherty & Quay, 1991; O’Brien & Frick, 1996) or passive avoidance errors during a discrimination learning task (e.g., Iaboni et al., 1995). Although performance on these tasks distinguishes groups expected to be behaviorally inhibited (BIS dominance) or disinhibited (BAS dominance), the psychological processes underlying performance remain open to different interpretations. For example, disinhibited behavior on these tasks might reflect strong BAS activation such that sensitivity to reward cues is the primary catalyst, or disinhibited behavior might stem from weak BIS activation such that cues for potential punishment do not inhibit behavior. That is, there are likely multiple influences on disinhibition (Newman & Wallace, 1993), and current measures of disinhibition do not distinguish these influences. In sum, both laboratory and questionnaire measures that assess BAS and BIS functioning in childhood are needed.

The goal of the current study was to adapt measures from the adult literature that would allow us to assess the independent strength of the BIS and BAS in children. We focused on the core concepts of sensitivity to reward, punishment, and extinction. Specifically, we report on a questionnaire instrument that utilizes caregiver reports, and two laboratory tasks that assess reactivity to reward, punishment, and extinction.

Torrubia, Avila, Molto, and Caseras (2001) reported on the Sensitivity to Punishment and Sensitivity to Reward Questionnaire. The Sensitivity to Punishment and Sensitivity to Reward Questionnaire is an adult self-report measure designed to assess core elements of the BAS and BIS. This measure is noteworthy because unlike other measures (e.g., the BIS/BAS scales, Carver & White, 1994) it was designed to conform to several predictions that follow from Gray’s reinforcement sensitivity model, including orthogonality between the BAS and BIS, expected associations between the behavior approach and inhibition systems and Eysenck’s personality dimensions of extraversion and neuroticism, and items that assess sensitivity to reward or punishment, but not both (Heubeck, Wilkinson, & Cologon, 1998; Jorm et al., 1999; Torrubia et al., 2001). Exploratory factor analysis of the items generally supported two factors corresponding to sensitivity to punishment (SP) and sensitivity to reward (SR), and the SP and SR scales were related to neuroticism and extraversion as expected. Moreover, on a laboratory task, Avila (2001) found that high levels of SP were associated with response inhibition and that high levels of SR were associated with response activation. In the current study, we adopted the Sensitivity to Punishment and Sensitivity to Reward Questionnaire to be used by caregivers to report on their children’s behavior.

We also examined how caregiver reports of SP and SR related to performance on two laboratory tasks. Avila (2001) validated a point scoring reaction time task as a measure of BAS and BIS functioning. As hypothesized, Avila (2001) found that adults who self-reported high levels of SR on the Sensitivity to Punishment and Sensitivity to Reward Questionnaire responded faster to reward cues than did participants who self-reported low levels of SR. Moreover, adults who self-reported high levels of SP on the Sensitivity to Punishment and Sensitivity to Reward Questionnaire responded more slowly when presented with punishment cues than did those who self-reported low levels of SP. We adapted this task for children and examined how performance on this task was associated with caregiver reports of SP and SR.

We also administered a continuous performance task that has previously been used with children (Beauchaine, Katkin, Strassberg, & Snarr, 2001; Iaboni, Douglas, & Ditto, 1997). This task was designed to assess physiological reactivity to reward and extinction, and thus differs from other continuous performance tasks, which typically assess attentional vigilance or attentional shifting (e.g., Epstein, Goldberg, Conners, & March, 1997). Psychophysiological research suggests that activity of the sympathetic branch of the autonomic nervous system is an index for BAS and BIS activity. In a series of studies with adults, Fowles (1983) demonstrated that heart-rate changes in response to reward incentives, and that the degree of reactivity is proportional to the amount of reward. Crone, Jennings, and van der Molen (2003) similarly found that heart rate (HR) increased with reward feedback and decreased with punishment feedback during a learning task in a sample of 6-12 year olds. Fowles has concluded that heart-rate reactivity to reward incentive is an index of BAS activity. In contrast, electrodermal reactivity to cues for punishment, nonoccurrence of expected reward, and aversive stimuli are thought to index BIS activity (Fowles, 1980). Fowles, Kochanska, and Murray (2000) found that electrodermal reactivity to a variety of stimuli was associated with children’s inhibited behavior and fearfulness, and they concluded that electrodermal reactivity reflects individual differences in BIS functioning in children. Other research with children and adults has shown that electrodermal activity increases in response to punishment cues, extinction, and aversive stimuli (Iaboni et al., 1997; Lang, Greenwald, Bradley, & Hamm, 1993; Tranel, 1983). Therefore, electrodermal reactivity to extinction or cues for punishment can be considered an index of BIS functioning. In the current study, a continuous performance task was administered to examine how physiological reactivity during the task was related to caregiver reports of SP and SR. The task was administered according to procedures of Iaboni et al. (1997), who used it with children.

Finally, caregiver reports of externalizing and internalizing symptoms were assessed so that we could evaluate if our questionnaire measure of the BIS and BAS showed a more theoretically coherent pattern of relations to our laboratory tasks compared to behavior problems. Although externalizing and internalizing symptoms are thought to develop partly from individual differences in BIS and BAS functioning, these broadband dimensions of problem behavior combine several narrowband dimensions that are thought to show different patterns to relations with the BAS and BIS. Thus, when we examined associations with our laboratory assessments, we expected to find stronger patterns of relations for our BIS/BAS questionnaire measure than for problem behavior.

In sum, we sought to measure BAS and BIS functioning in children. First, a caregiver report measure of SP and SR was adapted from an adult self-report measure, and we examined the association between this measure and performance on two laboratory tasks. We also examined how caregiver and laboratory assessments of SP and SR were associated with childhood externalizing and internalizing behavior problems. Hypotheses regarding the lab tasks are summarized in Table I. With respect to the point scoring reaction time task, general disinhibition (a high number of passive avoidance errors) was expected to be associated with low levels of parent reported SP and high levels of parent reported SR, indicating multiple influences on disinhibition. That is, disinhibition may occur because of lack of reactivity to potential punishment or strong reactivity to reward. High levels of externalizing behavior problems were also expected to be associated with a high number of passive avoidance errors. High levels of caregiver reported SP and internalizing behavior problems were expected to be associated with greater inhibition on the point scoring reaction time task (slower reaction times in response to cues previously associated with punishment), and high levels of caregiver reported SR and externalizing behavior problems were expected to be associated with less slowing of reaction times in response to reward cues in the context of mixed incentives during the point scoring reaction time task. With respect to physiological reactivity, HR was expected to increase in response to reward incentives (relative to practice and extinction), particularly for children characterized by high levels of parent reported SR and externalizing behavior problems. Moreover, electrodermal increases were expected in response to extinction, and this reactivity was expected to be greatest for children characterized by high levels of parent reported SP and internalizing problems. Finally we expected caregiver reports of externalizing and internalizing problem behavior to show similar but weaker patterns of association as our questionnaire measures of SR and SP, respectively.

METHOD

Participants

Participants were paid community volunteers. Children were eligible for the study if they were between 9 and 12 years old, without learning or physical disabilities that would preclude understanding and responding to interview questions or completing the computer tasks, not taking medication that would interfere with physiological processes (e.g., anticonvulsant medication or antidepressants), and did not have high blood pressure. Children who were taking medication for ADHD (e.g., methylphenidate) were included, but caregivers were instructed to give the last dose 12 hr prior to participating in the study (N = 2). The core sample included 63 children (mean age = 10.6 years, 42% female). The majority of the children were White (92%). Minority groups represented in the sample included Black. Hispanic, Asian-Pacific Islander, and Mixed race. Almost all of the caregivers who participated were biological mothers (88.9%). The remainder included biological fathers (7.9%) and grandmothers (3.2%). Median family income for the sample was $50,000 (range $42,000-$125,000). An additional 47 participants were included for factor analysis of the caregiver report of child SP and SR. Ages for these additional children ranged from 9 to 12 (mean age = 10.7 years, 46% female). The majority of the additional participants were White (63%), and 91% of the caregivers who participated were biological mothers. Median family income was $46,000.

Procedures

Data collection was done in one session (typically 2 hr) for the core sample. Assent/consent forms were read aloud to the caregiver and child, and after signing these forms the child was escorted to a separate room. One interviewer stayed with the caregiver and read all questions allowed. The experimenter read a script to the child, which elaborated a space academy cover story to introduce and describe the computerized tasks. To enhance motivation, children earned points on each task, which were redeemed for a prize. The tasks were administered in a fixed order (continuous performance task, followed by a 15-min break and then the point scoring reaction time task). The experimenter introduced the continuous performance task and rationale for physiological hook up. After these preliminary comments, the caregiver was brought into the room to observe attachment of electrodes, and then escorted back to the interview room. The additional 46 caregivers came into the lab for a 45-min session and answered questionnaires, which were read aloud to them by an interviewer.

Measures

Caregiver Report of Child Punishment and Reward Sensitivity

The Sensitivity to Punishment and Sensitivity to Reward Questionnaire (Torrubia et al., 2001) is an adult self-report measure that assesses two dimensions. We modified the Sensitivity to Punishment and Sensitivity to Reward Questionnaire in several ways to be used for children. First, we eliminated items that loaded on both the SR and SP factors or that loaded weakly on their respective factor based on factor analysis in adult studies (O’Connor & Colder, in press; Torrubia et al., 2001). Items inappropriate for children (e.g., “Do you often take the opportunity to pick up people you find attractive”) were also eliminated. Second, several items were reworded to be more appropriate for child behavior. For example, “When you start to play a slot machine, is it difficult for you to stop?” was rewritten to “Your child has a lot of difficulty ending a fun activity.” Finally, several items from other temperament inventories that conceptually overlap with sensitivity to punishment and reward were added (e.g., “When your child gets what they want they feel excited and energized,” “Criticism and scolding hurts your child very much”). The final measure included 34 items (16 sensitivity to punishment and 18 sensitivity to reward items). Caregivers rated their agreement with each item using a 5-point scale (1 = strongly disagree to 5 = strongly agree).

A confirmatory factor model specifying two correlated latent factors, SP and SR, did not provide an adequate fit to the data, [chi square](526, N = 110) = 999.84, p < .01, root means square error of approximation = .09, comparative fit index = .57. Examination of the modification indices did not suggest that this model was easily modifiable to produce an adequate fit to the data. Consequently, exploratory factor analysis was used (principal factor extraction with varimax and promax rotation) to identify the factor structure of the 34 SP and SR items. Eigenvalues from the full correlation matrix were used to determine the number of factors to extract. There were 10 eigenvalues greater than 1, which was judged to be too many factors. The scree test suggested a 4- to 6-factor solution. The 4-, 5-, and 6-factor solutions were estimated, and the 4-factor solution provided the most interpretable solution and was retained. One item loaded weakly on its factor (standardized regression coefficient = .33) with a low item-total correlation (r = .05). This item ("Generally, your child pays more attention to threats than to pleasant events") was deleted and the solution was rerun with 33 items. The final 4-factor solution is presented in Table II. (3) The standardized regression coefficients suggested minimal cross-loading. The factor pattern suggests that one factor represents SP, whereas the remaining factors could be considered facets of SR, including impulsivity/fun seeking, drive, and reward responsiveness. This factor solution is similar to the one identified by Carver and White (1994) who developed a measure to assess adult individual differences in BAS and BIS functioning. These authors also identified three facets of BAS functioning that directly correspond to our SR scales.

The internal consistencies were acceptable for the four scales (standardized Cronbach’s Alpha = .87 for SP, .76 for impulsivity/fun seeking, .73 for drive, and .69 for reward responsiveness). Factor scores using the regression method were output and used in subsequent analyses. Descriptive statistics for the 4-factor scores are presented in Table III. SP was uncorrelated with the SR scales (r = .06 for impulsivity/fun seeking, r = -.01 for drive, and r = .15 for reward responsiveness, all ps > .25). Drive was significantly correlated with impulsivity/fun seeking (r = .35, p < .01) and with reward responsiveness (r = .51, p < .05). Impulsivity/fun seeking and reward responsiveness were unrelated (r = .16, p < .19).

Laboratory Assessment of Sensitivity to Punishment and Sensitivity to Reward

Laboratory tasks were completed in a sound and light attenuated room and were programmed using E-Prime Version 1.0 (Beta 5.0/Service Pack 2) on a Dell Optiplex 133 MHz Pentium III PC. The color monitor was a 17″ Dell E700. Children responded using Psychology Software Tools. Inc. Deluxe Serial Response Box and a response box built using the custom expansion kit (specifications are described below). They were instructed to press the response button using their dominant hand.

Point Scoring Reaction Time Task. Children completed the point scoring reaction time task (Avila, 2001, experiment 1). The task involved 20 practice trials and 3 experimental blocks, each of which included 100 3-s trials. The experimental blocks were presented in a fixed order (prepunishment, punishment, and postpunishment). The stimuli were the same across each block (a colored circled presented above a two-digit number), and the participant’s task was to discriminate odd and even numbers. Correct discriminations were rewarded by earning a variable number of points, which depended on reaction time (earned points = 635/reaction time in ms.). Incorrect discriminations were punished with a loss of 3 points. Children were told to ignore the colored circles during the prepunishment block. Before beginning the punishment block, they were told that responding when a red circle appeared would result in a loss of 50% of their accumulated points. All responses (whether they were correct or incorrect discriminations) when a red circle appeared resulted in point loss. Thus, a red circle became a cue for potential punishment. Of the 100 trials, 8 included a red circle (punishment trials). Prior to initiating the postpunishment block, children were told that a red circle would not cause a loss of points, and that they should respond during these trials. The point scoring reaction time task as described above included two changes from the original task used by Avila (2001). First, the trial length was increased from 2 to 3 s to allow children more time to make the distinction between odd and even numbers. Initial piloting suggested that some children were unable to make the distinction in 2 s. Second, the number of points lost for incorrect discriminations was reduced from 5 to 3 points so that children did not lose too many points. Substantial point loss might reduce motivation during the task. Other than these two changes, the task procedures follow those of Avila (2001, experiment 1). Before beginning the practice block, experimenters tested the children to be certain that they could distinguish odd and even numbers. If not, they were given brief instructions and then tested again. If during the practice block the experimenter determined that the child did not understand the distinction, the experiment was stopped and the child was again instructed about the distinction between odd and even numbers, and the practice block was readministered. This occurred with one participant. (4)

Reaction times (RTs) are the primary measure of interest, and were calculated using RTs from trials that included both correct and incorrect responses. During the punishment block reward and punishment cues are presented. Trials in which a non-red circle is presented are considered reward trials because they signal the potential for reward, whereas trials in which a red circle is presented are considered punishment trials because they signal the potential for punishment. Non-red circles during the prepunishment block similarly signal the potential for reward. How responses to reward cues (trials in which non-red circles are presented) change during the punishment versus prepunishment block indicates SR and BAS activation. Slower responding is expected during the punishment block because of the possibility of increased punishment, and less slowing suggests greater SR and strong BAS activation. It is notable that SR is assessed during the punishment block when either a reward or punishment cue could appear, and thus BAS activation is assessed in the context of mixed incentives. Past research has shown that disinhibited behavior is most apparent in the context of mixed incentives (Patterson & Newman, 1993). During the postpunishment block, inhibition is expected in response to red circles because these cues were previously associated with punishment. A comparison of reaction times during non-red circle trials (trials in which a reward cue is presented) to red circle trials (trials in which a cue previously associated with punishment is presented) during the postpunishment block is an index of BIS activation. Greater slowing of reaction times during red circle trials is indicative of high SP and strong BIS activation. In addition to these comparisons, we also considered the number of passive avoidance errors made (responding during red circle trials in the punishment block) to assess disinhibition.

Physiological Reactivity. Children were fitted with three disposable ECG electrodes and two skin conductance cuffs, ECG data were collected using Electrocap International (ECI) E28 alligator clip electrode leads and E27 disposal electrodes placed axially on the left and right rib at approximately the elevation of the heart. The reference electrode was placed on the collar bone. A pair of UFI 1081FG skin conductance electrodes were used with UFI 1090 skin conductance gel and attached with UFI 1080B adhesive collars. Skin conductance electrodes were attached to the middle phalange of the child’s nondominant index and middle fingers. Respiration bellows were placed at the height of the zyphoid process to measure inspiration and expiration. Once stable recording levels are obtained, baseline physiological measures were collected while children sat quietly for 5 min, and then the experimenter instructed the child on the continuous performance task originally developed by Fowles (1983). We administered the task according to the procedures of Iaboni et al. (1997).

The task involves a response box with five lights arranged in a semicircle with each light accompanied by a button which turns the light off when pressed (the Psychology Software Tools. Inc. Custom Response Box Expansion Kit was used to construct the response box). Participants press the button adjacent to the illuminated light and then press a central button, which randomly turns on another light. Thus, the participant responds continuously by pressing the button adjacent to the illuminated light and then the central button until the block is over. Points are earned when a light is turned off. The task consists of seven 2-min blocks with a rest period of 2.5 min preceding each block. Thirty seconds prior to the end of each rest period, a warning frame (“get ready”) appears on the computer monitor. Children were instructed to sit quietly and not to talk during rest periods. Points earned for each block are displayed on a computer screen during the rest periods, and the screen is blank during the blocks. The first block is a practice block in which no reward is administered. Blocks 2, 3, and 4 are reward blocks during which the child earns a point each time they turn off a light. Turning off a light is accompanied by a 200-MHz tone signaling reward. Reward continues for the first 30 s of block 5, followed by 90 s of extinction. Extinction is introduced at this time rather than after the rest period so that children experience reward immediately prior to extinction, thus enhancing the expectation of reward. During extinction the tones are no longer delivered and no points are earned. In block 6, reward was reinstated for the entire 2 min. After extinction, reward expectancies are presumably weakened, and a full reward block was administered after the first exposure to extinction so that strong expectancies of reward were reinstated. This allowed a second examination of extinction. Extinction is reintroduced for the first 90 s of block 7 followed by 30 s of reward so that participants do not finish the task in the extinction condition. We wanted to avoid disappointment that might follow extinction so that participants did not leave with a negative feeling about participating in the experiment. The children are told at the beginning of the task that tones signal a point earned and that they will hear tones on most but not all trials. This is necessary so that children do not assume that the equipment is malfunctioning.

Physiological data were recorded on-line directly into a data acquisition computer using a five-channel Bioamp (model # CRC-04BA. SA Instrumentation Company. San Diego, CA), SNAPMASTER software (HEM Data Corporation, 2000), and Daqbook A/D board. ECG was sampled at 10 Hz. A constant voltage of 0.5 V was used to record skin conductance level (SCL). IBI Analysis Software (James Long Company, 1998) was used to compute heart rate (HR). Comparison of HR reactivity during the practice (when no reward can be earned) and reward blocks is a measure of reactivity to reward incentives and an index of BAS activation. Comparison of SCL reactivity during practice and extinction blocks is a measure of reactivity to the nonoccurrence of expected reward and is an index of BIS activation.

Child Behavior Problems

Caregivers reported on their child’s problem behavior using the Child Behavior Checklist (CBCL. Achenbach, 1991). This measure has a long history and has been shown to be both a valid and reliable measure of children’s behavior problems (see Achenbach, 1991). Standardized Cronbach Alphas were high for both internalizing and externalizing scales (.87 and .91, respectively). Items were averaged for analysis. (5) Internalizing and externalizing scale scores were correlated (r = .61, p < .01), such that children reported to exhibit high levels of internalizing symptoms also exhibited high levels of externalizing symptoms. Descriptive statistics for internalizing and externalizing behavior problems are presented in Table III.

RESULTS

Results are organized to address two main questions. First, how does caregiver-report of SP and SR relate to laboratory assessments? Second, how do caregiver-report and laboratory assessments of SP and SR relate to behavior problems? For all analyses child age and sex were included as covariates, and for analysis of laboratory tasks, age and sex interactions with condition were also tested. There were no sex effects, and age effects are only described below when they were significant. When significant interactions with continuous variables were found, we conditioned our model on high (1 SD above the mean) and low (1 SD below the mean) levels of the predictor to test simple effects (Aiken & West, 1991). To facilitate interpretation of the interactions, the model implied means are reported in figures. Continuous predictor variables were standardized to eliminate nonessential collinearity that often occurs when higher-order terms are tested. Sample size varied somewhat from analysis to analysis (N ranged from 53 to 63) because of equipment failure or other difficulties (e.g., electrodes slipping off). Multilevel models estimated in Proc Mixed in SAS version 8.02 (Littell, Milliken. Stroup, & Wolfinger, 1996) were used to analyze physiological data and reaction times from the point scoring reaction time task. Multilevel modeling is appropriate where repeated measures are nested within a random sample of participants, as it permits estimation of variability due to subject (Bryk & Raudenbush, 1992). When the SP and the three SR scales were considered as predictors, they were tested simultaneously in the same model, which allowed us to consider the unique effects of SP and SR. Similarly, internalizing and externalizing symptoms were included in the same model so that we could evaluate the unique relations.

Associations With Laboratory Assessments

Point Scoring Reaction Time Task

First, we examined the number of passive avoidance errors made. The number of responses made during punishment trials (trials that included a red circle) during the punishment block was regressed on SP and SR scales. When the dependent measure is a count variable, the data often follow a poisson distribution, so poisson regressions were used for this analysis. We expected more passive avoidance errors among children characterized by low levels of SP than those characterized by high levels of SR. Results suggested that age (B = -.46, p < .01), and both SP (B = -.40, p < .05) and impulsivity/fun seeking (B = .41, p < .05) predicted passive avoidance errors. (6) Low levels of SP and high levels of impulsivity/fun seeking were associated with a high number of passive avoidance errors. Older children made fewer passive avoidance errors relative to younger children. When problem behavior was considered, passive avoidance errors were predicted by externalizing symptoms (B = .35, p < .04) but not internalizing symptoms (B = -.20, p < .41). A high number of passive avoidance errors were associated with increasing levels of externalizing behavior problems.

Next, reactivity to punishment cues were examined by comparing RTs during trials with a red circle and trials without a red circle (reward trials) from the postpunishment block. During the postpunishment block, RTs are expected to be slower during red circle trials because red circles were previously associated with punishment, and therefore, we refer to these red circle trials as punishment trials even though in the postpunishment block red circles were associated with reward. Children characterized by high SP were expected to show more slowing during punishment trials because they were more likely to retain the association between red circles and punishment. Trial type (punishment vs. reward), SP and SR scales, and the interaction between trial type and the SP and SR scales were included as effects in the model. Results suggested a significant effect of age (t value = -4.00, p < .01) and trial type (t value = 4.27, p < .01). Older children responded more quickly than younger children, and as expected, RTs were slower during the punishment compared to reward trials. The trial effect was qualified by a marginal SP X trial type interaction (t value = 1.87, p = .06). The trial effect was evident at both high (t value = 4.30, p .28). Results for problem behavior suggested that neither the internalizing by trial (t value = 1.64, p < .11) nor the externalizing by trial (t value = -1.50, p < .14) interaction term was statistically significant. However, results suggested that the pattern of findings for internalizing problems were in the same direction as for SP (High internalizing: t value = 4.00, p < .01; Low internalizing: t value = 1.47, p < .15).

[FIGURE 1 OMITTED]

Finally, RTs during reward trials were compared across prepunishment and punishment blocks. RTs are generally expected to slow down during the punishment block compared to the prepunishment block, however, children characterized by high levels of SR were expected to show less slowing. Block (prepunishment vs. punishment). SR and SP scales, and the interaction between block and the SR and SP scales were included as effects in the model. Results suggested significant effects of block (t value = 9.78, p < .01) and age (t value = -4.59, p < .01). RTs slowed during the punishment block compared to the prepunishment block, and older children responded more quickly than younger children. Age by block (t value = 2.30, p < .05) and impulsivity/fun seeking by block (t value = -1.78, p .14). Older children showed greater slowing of reaction time than did younger children. Although children characterized by high (t value = 5.40, p < .01) and low (t value = 7.76, p < .01) levels of impulsivity/fun seeking showed significant increases in reaction times during the punishment block, this increase was smaller for the former group as hypothesized (see Panel B in Fig. 1). Results for problem behavior suggested that the externalizing by block (t value = -1.77, p < .09) interaction term was marginally significant while the internalizing by block (t value = 0.96, p < .34) interaction term was not statistically significant. As hypothesized, the pattern of findings at high (t value = 4.57, p < .01) and low (t value = 7.13, p < .01) externalizing behavior problems were in the same direction as found at high and low levels of impulsivity/fun seeking.

Physiological Reactivity

SCL data were not normally distributed, and so the data were log transformed for analysis. Preliminary analysis suggested that participants showed physiological reactivity to the “get ready” screen, and that the pattern of reactivity was similar across all six rest periods. Preliminary analyses also suggested that HR and SCL did not change during the 30 s prior to the “get ready” screen, and this was true across all six rest periods. Accordingly, a baseline measure for each experimental block was taken by averaging physiological measures during the 30 s prior to the “get ready” screen. Averages were also taken within each experimental block (the practice, reward, and extinction blocks), and then change scores were calculated by subtracting the baseline measure from each experimental block measure so that positive scores indicated an increase from baseline to experimental block. Change scores were used in all subsequent analyses of physiological data.

In our first set of analyses we examined the potential effect of reward and extinction on physiological reactivity, and whether caregiver reports of SP, SR, and problem behavior moderated these effects. The impact of reward was examined by contrasting physiological reactivity during the first three reward blocks with reactivity during the practice block. The fourth reward block was only 30 s long and was not included in these analyses. These models included condition (Reward 1, Reward 2, Reward 3, or practice), SP and SR scale scores, and the interaction between condition and the SP and SR scale scores. HR was expected to increase more during reward blocks indicating BAS engagement, and this increase was expected to be greater among children characterized by high levels of SR. Results for HR suggested a significant overall condition effect (F value = 5.22, p < .01), and contrasts suggested that HR increased more during each reward block relative to the practice block (t values = 2.68 to 3.74, all ps .17). When problem behavior was considered, neither the internalizing by condition nor externalizing by condition interaction term was statistically significant (F values = 0.19 and 1.29, ps > .28).

SCL reactivity was not expected to differ across the reward and practice blocks because SCL is thought to index sensitivity to punishment and extinction, but not to reward. Moreover, the SP and SR scales were not expected to moderate the condition effect. Contrary to expectation, results for SCL suggested an overall condition effect (F value = 16.12, p < .01) and contrasts suggested that SCL reactivity declined during each reward block relative to the practice condition (t values = -2.76 to -6.66, ps < .01). Age also predicted SCL reactivity (F value = 5.05, p .13), but the SP by condition interaction term was statistically significant (F value = 3.30, p < .05). Post hoc contrasts suggested that the SP by condition interaction was driven by a much more rapid decline of SCL reactivity during the first reward block relative to the practice block for children characterized by low SP (t value = 2.94, p < .01) compared to children characterized by high SP (t value = 0.25, p < .80) (see Panel A in Fig. 2). Results for problem behavior suggested that the internalizing problems by condition interaction term was not statistically significant (F value = -1.89, p < .13), but that the externalizing behavior problems by condition interaction term was statistically significant (F value = 2.53, p < .05). Post hoc contrasts suggested that only the comparison between the third reward block and practice was moderated by externalizing behavior problems (t value = 2.00, p < .05). Children characterized by high (t value = -5.77, p < .01) and low levels (t value = -2.50, p < .05) of externalizing behavior problems both showed declines in SCL reactivity during the third reward block, but this effect was greater for the former group.

[FIGURE 2 OMITTED]

Next the impact of extinction was examined by contrasting physiological reactivity during extinction to reactivity during practice. The first extinction period occurred after 30 s of reward (the last 90 s of the fourth experimental block), whereas the second extinction period occurred during the first 90 s of the sixth experimental block. Accordingly, we used the last 90 s of the practice block as a comparison condition for the first extinction block, and the first 90 s of the practice block as a comparison condition for the second extinction block. HR reactivity was expected to be similar during extinction and practice because HR is thought to be responsive to reward, but not punishment or extinction. No difference would suggest similar levels of BAS activity across the blocks. SP and SR scales were not expected to moderate the condition effect. Results for the first extinction block suggested no difference across extinction and practice (t value = -0.99, p < .32). Only impulsivity/fun seeking moderated this condition effect (t value, -2.69, p < .01). HR reactivity increased during extinction relative to practice for children characterized by low levels of impulsivity/fun seeking (t value = 2.87, p < .01) and did not change for children characterized by high levels of impulsivity/fun seeking (t value = -1.34, p .73). Results for problem behavior suggested that the externalizing by condition (t value = -2.45, p < .05) but not the internalizing by condition (t value = 0.79, p < .43) interaction term was statistically significant. The findings for high (t value = -1.23, p < .22) and low (t value = 2.58, p < .05) externalizing behavior problems were consistent with those for high and low impulsivity/fun seeking.

When the second extinction block was considered for HR, the condition effect was not significant (t value = 0.80, p .21). Similarly, neither the externalizing by condition (t value = -1.18, p < .24) nor the internalizing by condition (t value = 0.34, p < .74) interaction term was statistically significant when the second extinction block was considered. For SCL we expected greater reactivity during extinction relative to practice, indicating greater BIS engagement during the nonoccurrence of expected reward. However, contrary to expectation, results suggested that SCL reactivity decreased during both the first and second extinction blocks compared to practice (t values = -5.48 and -5.71, ps .14). Similarly, neither the externalizing by condition nor the internalizing by condition interaction term was statistically significant for both of the extinction blocks (t values = -0.94 to 1.38, ps > .17).

Finally, we compared physiological reactivity during extinction and reward. For these comparisons we considered reward block 3 as the comparison condition because this block was the last full reward block to be administered before the introduction of extinction. Similar to our analysis comparing extinction and practice, we contrasted the last 90 s of reward block 3 with the first extinction block, and contrasted the first 90 s of reward block 3 with the second extinction block. We expected greater HR reactivity during reward relative to extinction, suggesting greater BAS engagement during reward, and this effect was expected to be stronger for children characterized by high levels of SR. Results suggested greater HR reactivity during reward relative to both extinction blocks (t values = -2.72 and -2.79, ps < .01), and this effect was moderated by impulsivity/fun seeking, but only for the first extinction block (t value = -2.00, p < .06). As hypothesized, greater HR reactivity during reward was evident for children characterized by high (t value = -3.41, p < .01), but not low (t value = -0.56, p .20). However, results suggested that HR reactivity was greater during reward versus the first extinction block for children characterized by high levels of externalizing behavior problems (t value = -2.80, p < .01), but not for children characterized by low levels of externalizing behavior problems (t value = -0.72, p < .47).

[FIGURE 3 OMITTED]

Greater SCL reactivity was expected during extinction, indicating greater BIS activation, and this effect was expected to be most evident at high levels of SP. Results suggested that SCL reactivity was similar across extinction and reward, and this was true for both extinction blocks (t values = -1.15 to 0.32, ps > .25). The SP by condition interaction term was marginally significant when the second (t value = 1.80, p < .08), but not the first (t value = 0.72, p < .48) extinction block was considered (see Panel B in Fig. 2). Post hoc contrasts suggested that SCL reactivity was similar across the second extinction and reward blocks at high levels of SP (t value = 0.63, p < .54), but that SCL reactivity declined during the second extinction block at low levels of SP (t value = -2.05, p .13). When problem behavior was considered, the internalizing by condition (t value = -1.92, p < .07) but not the externalizing by condition (t value = 1.16, p .10). Neither the internalizing by condition (t value = 0.57, p < .57) nor the externalizing by condition (t value = -0.94, p < .35) interaction term was statistically significant when the second extinction block was compared to reward.

The Relation Between Caregiver Reports of SP and SR, and Behavior Problems

Internalizing and externalizing scale scores were regressed on SP and SR. Externalizing behavior was included as a predictor of internalizing behavior, and internalizing behavior was included as a predictor of externalizing behavior because of the high correlation between these two scales. This allowed us to examine symptoms that were unique to the externalizing and internalizing behavior problems. Results of the regression models suggested that impulsivity/fun seeking (B = .51, t value = 5.60, p .22). SP significantly predicted internalizing behavior problems (B = .40, t values = 3.95, p .46).

DISCUSSION

Caregiver Reports of BAS and BIS

We adopted a questionnaire measure from the adult literature (Torrubia et al., 2001) to be used by caregivers to report on their child’s sensitivity to reward (the BAS) and sensitivity to punishment (the BIS). Our final measure included 33 items that represented SP and three facets of SR, including impulsivity/fun seeking, drive, and reward responsiveness. The SP scale was uncorrelated with the SR scales, and this is consistent with Gray’s theory and with findings from adult studies (O’Connor & Colder, in press: Torrubia et al., 2001). Our subscales map directly onto the BIS/BAS scales developed by Carver and White (1994). The BIS/BAS scales are comprised of four scales, including BIS, fun seeking, drive, and reward responsiveness (Carver & White, 1994; Heubeck et al., 1998; Jorm et al., 1999). One difference between our child measure and the adult BIS/BAS scales is that the reward responsiveness and the BIS scales are not independent in the Carver and White (1994) measure, but they are independent in our measure. This is notable because Gray’s theory posits independence between the BIS and BAS.

Caregiver Reports of SP and SR and Laboratory Assessments of BAS and BIS

When we examined the association between caregiver reports of SP and SR and behavioral responses to reward and punishment cues on the point scoring reaction time task, results were consistent with our hypotheses. There was evidence that both low levels of caregiver reported SP and high levels of caregiver reported SR were associated with behavioral disinhibition on this task (passive avoidance errors), which supports the notion that there are multiple influences on disinhibited behavior (Avila, 2001; Newman & Wallace, 1993). These findings suggest that measures of general behavioral disinhibition do not distinguish BIS and BAS activation. High levels of caregiver reported SP were associated with greater inhibition in response to punishment cues, and high levels of caregiver reported impulsivity/fun seeking were associated with greater BAS activation in response to reward cues in the context of mixed incentives. These findings suggest convergent validity of these measures. It is notable that caregiver report of SP was not associated with responsiveness to reward cues, and caregiver report of SR was not associated with responsiveness to punishment cues, which supports the discriminant validity of these measures. It is also notable that impulsivity/fun seeking was the only SR scale to correlate with performance on the point scoring reaction time task. Impulsivity/fun seeking may represent a more pure behavioral expression of BAS activation than the other SR scales.

With respect to the continuous performance task, we found support for BAS engagement as indexed by heart rate (HR) reactivity. HR reactivity was greater during reward relative to practice and extinction. Fowles (1988) posited that cardiac reactivity in response to reward incentive is a physiological index of BAS activation, and our findings are consistent with this view. High cardiac activity is also associated with less focused and sustained attention (Jennings, 1992; Porges, 1991). HR reactivity during reward may not only suggest engagement of the BAS, but also less focused attention during the reward blocks. This would be consistent with the notion that engagement in approach behavior reduces the likelihood of allocating attentional resources to reflect on potential changing situational circumstances (Patterson & Newman, 1993).

There was limited evidence for an association between caregiver reports on the SR scales and HR reactivity in response to reward. High levels of impulsivity/fun seeking were associated with greater HR reactivity during reward compared to the first exposure to extinction. This effect seemed to be driven by high HR reactivity during extinction for children characterized by low levels of impulsivity/fun seeking. HR reactivity was not expected to change in response to extinction when compared with practice because extinction involves the nonoccurrence of an expected reward, which should engage the BIS, not the BAS. Overall, the pattern of cardiac reactivity suggests that children characterized by low impulsivity/fun seeking experienced an increase in HR reactivity during the first exposure to extinction, which was not apparent during the second exposure to extinction. These data are consistent with an extinction burst, which was not apparent during the second exposure to extinction, perhaps because after the first exposure to extinction, the children learned that their behavior was not effective in producing reward. Why this would occur only for low impulsive/fun seekers is unclear. However, since interpretation is post hoc and these effects were not expected, they need to be replicated.

Contrary to expectation, SCL reactivity declined during extinction relative to practice. It is possible that habituation might account for this pattern of SCL reactivity. SCL reactivity was lower during reward and extinction relative to practice, and the reward and extinction blocks occurred later in the experiment compared to the practice block. SCL reactivity may represent arousal in response to a variety of features of the task (pushing a response button, lights, tone) that were novel during the practice block, and this reactivity may have habituated to repeated exposure to the task so that the highest levels of reactivity were observed during practice. Also contrary to expectation, SCL reactivity was similar during extinction and reward. This was surprising because reward tones were not produced during extinction, which should have produced BIS activation. Orienting responses might also have been expected when the tone was not sounded as expected, and orienting responses typically produce increases in SCL (Raskin, 1973). It is possible that the termination of tones was not particularly salient to most of the children. They were told at the beginning of the task that tones signaled reward, and that sometimes pushing the buttons would not produce tones and earn reward. By the fourth experimental block, when extinction was first introduced, this information may not have been a prominent feature of the task for most children.

SP was related to SCL reactivity, but not as expected. SCL reactivity did not decline during the initial exposure to reward for children characterized by high levels of SP, but it did decline for the overall sample and for children characterized by low levels of SP. There was also evidence that SCL reactivity did not change across extinction and reward for children characterized by high levels of SP, but it was lower during extinction for children characterized by low levels of SP. Perhaps the BIS was more likely to be engaged (higher levels of SCL reactivity) for children characterized by high levels of caregiver reported SP in some contexts. Alternatively. SCL reactivity may habituate more slowly with repeated exposure to the task for children characterized by high levels of SP.

The Relation Between SR and SP and Behavior Problems

Caregiver reports of SP and SR were related to problem behavior as expected. High levels of impulsivity/fun seeking (a strong BAS) were associated with increasing levels of externalizing behavior problems, and high levels of SP (a strong BIS) were associated with increasing levels of internalizing behavior problems. It is notable that impulsivity/fun seeking was the only dimension of the BAS that was associated with externalizing behavior problems. Prior research has also found that impulsive sensation seeking is associated with externalizing behavior problems (Farrington & West, 1993; Russo et al., 1991). Impulsivity/fun seeking may represent a more risky temperament than reward responsiveness or drive because it increases the likelihood of frequent coercive social transactions, which is known to be an important process in the development of disruptive behavior problems (Patterson, Reid, & Dishion, 1992).

In general, problem behavior showed a similar but weaker pattern of relations to our laboratory measures compared to caregiver reports of SP and SR. This is not surprising because although the BIS and BAS are expected to be associated with problem behavior, many other risk factors are important for the development of problem behavior (e.g., family context, community risk, etc.). In contrast, our SP and SR (particularly impulsivity/fun seeking) scales were developed to assess BAS and BIS functioning, and in this regard, they represent better measures of BAS and BIS functioning than behavior problem items.

Conclusion

Overall, the psychophysiological findings do not consistently support a BIS/BAS interpretation. However, performance during the point scoring reaction time task and questionnaire data are consistent with a growing body of evidence (e.g., Biederman et al., 1990; Iaboni et al., 1995; Kagan et al., 1999; Matthys et al., 1998; Newman et al., 1985; O’Brien & Frick, 1996; Oosterlaan & Sergeant, 1998) that suggests that strong BAS places children at risk for disinhibited behavior and externalizing behavior problems such as aggression and delinquency, whereas a strong BIS places children at risk for inhibition and internalizing symptoms. Our findings are notable in that these relations are consistent across two methods of assessing BIS and BAS functioning.

Limitations and Future Directions

In closing, the limitations of the current study should be noted. First, we found that SP and SR were related to childhood problem behavior using a community sample. It will be important for future research to examine clinical groups using multiple assessments of the BAS and BIS. In doing so, it may be important to assess specific forms of psychopathology, rather than global internalizing and externalizing because different configurations of BAS and BIS functioning are posited to relate to specific forms of psychopathology. Second, the current findings are based on a community sample that was somewhat homogeneous, and ranging from middle to upper middle class, and thus they should be generalized across contexts and samples with caution.

Third, some of the associations we report may be inflated because of contamination of items assessing behavior problems and caregiver reports of SP and SR. Moreover, contamination of our SP and SR scales with the behavior problem scales may account for the close correspondence we found between these scales and the laboratory tasks. A useful direction for future research would be to decontaminate our SP and SR scales, and measures of problem behavior. A combination of expert ratings and confirmatory factor analysis have been used in past research to this end (Lengua, West, & Sandler, 1998), but there is considerable overlap between the two approaches in terms of the final item set they identify, and confirmatory factor analysis may be a simpler approach to use (Lengua, personal communication, February 19, 2003). Our small sample size precluded us from estimating a factor model with SP, SR, and problem behavior items. Moreover, our sample size likely limited our ability to detect potential interactions with age and sex.

Finally, the design of the continuous performance task made it difficult to interpret reactivity to extinction, and this was particularly true with respect to SCL findings because SCL reactivity could be attributed to habituation. Future research is necessary to refine this task so that it better assesses physiological reactivity to reward and extinction.

Despite these limitations, the current findings suggest measures that are useful for separating BIS and BAS influences on behavior in children. Such measures are important given growing interest in integrating these systems with models of child psychopathology and the potential utility of integrating reinforcement sensitivity with socialization models of risk. Good measures of the BAS and BIS will be important for etiological models of psychopathology to advance. Continued refinement of the questionnaire and tasks presented in this paper, and development of new measures to assess additional domains (e.g., attention) are important directions for future research.

Table I. Hypotheses That Involve Laboratory Tasks

Point scoring reaction time task

High number of passive avoidance errors associated with high levels of

SR and externalizing behavior, and with low levels of SP

RTs in presence of cues previously associated with punishment > reward

cues, particularly at high levels of SP and internalizing behavior

RTs in presence of reward cues during mixed cues > reward only cues,

less so at high levels of SR and externalizing behavior

Continuous performance

Heart rate reactivity during:

Reward > practice, particularly at high levels of SR and externalizing

behavior

Extinction = practice, and no difference expected across levels of SR

and SP, and across levels of internalizing and externalizing behavior

Reward > extinction, particularly at high levels of SR and

externalizing behavior

Skin conductance level reactivity during:

Reward = practice, and no difference expected across levels of SR and

SP, and across levels of internalizing and externalizing behavior

Extinction > practice, particularly at high levels of SP and

internalizing behavior

Reward < extinction, particularly at high levels of SP and

internalizing behavior

Note. SP = sensitivity to punishment; SR = sensitivity to reward.

Table II. Standardized Factor Pattern and Communalities ([h.sup.2]) for

Sensitivity to Punishment and Sensitivity to Reward Items

Sensitivity Impulsivity/

to punishment fun seeking Drive

2. Your child prefers not to ask .35 -.14 -.07

for something when they are not

sure they will obtain it

4. Your child is often afraid of .65 -.17 -.04

new or unexpected situations

6. Your child is a shy person .74 -.19 -.09

8. Whenever possible, your child .60 .29 -.26

avoids demonstrating their

skills for fear of being

embarrassed

10. When in a group, your child has .45 .15 -.05

difficulty thinking of something

to say

12. Whenever they can, your child .48 -.13 .04

avoids going to unfamiliar

places

14. Your child often worries about .55 -.07 .32

things they said or did

16. It is difficult for your child .57 -.17 -.08

to talk with someone they do not

know

18. Your child generally tries to .67 .01 -.16

avoid speaking in groups

20. Your child could do more things .72 .17 -.01

if it were not for their fear

22. Your child is afraid of many .72 .08 .02

things compared to other

children their age

24. Your child often refrains from .55 .27 .18

doing something they like in

order not to be rejected or

disapproved of by others

26. Your child often refrains from .69 -.01 .02

doing something because of fear

of being embarrassed

28. If your child thinks that .46 .08 .24

something unpleasant is going to

happen, they get pretty worked

up.

30. Criticism or scolding hurts your .33 -.07 -.07

child very much

1. The good prospect of obtaining a .23 -.09 -.10

reward motivates your child

strongly to do some things

3. Your child often does things to .21 -.23 .14

be praised

5. Your child enjoys being the -.10 .32 .02

center of attention

7. When your child is in a group, .02 .28 .20

they try to stand out as the

smartest or the funniest

9. When your child gets something -.06 -.10 -.22

they want, they feel excited and

energized

11. Your child does a lot of things .21 .04 .14

for approval

15. Does your child generally prefer -.02 .49 -.22

activities that involve

immediate reward

13. The possibility of obtaining .25 .59 .19

social status moves your child

to action, even if this involves

not playing fair

17. Your child often has trouble .04 .72 -.02

resisting the temptation of

doing forbidden things

21. Your child has a lot of -.13 .53 .06

difficulty ending a fun activity

23. Your child sometimes does things -.14 .57 -.07

for quick reward

25. Your child has difficulty -.05 .71 -.25

staying focused on their school

work in the presence of an

attractive alternative

27. Your child engages in risky .01 .53 .31

behavior to obtain a reward

33. Your child craves excitement and -.19 .35 .24

new sensations

19. Your child likes to compete and .05 .07 .63

do everything they can to win

29. Your child likes competitive -.10 -.10 .70

activities

31. Your child would like to be a -.13 .13 .52

socially powerful person

32. Your child likes displaying -.03 .01 .63

their physical abilities even

though it may involve danger

Percent variance 18% 11% 9%

Reward

responsivity [h.sup.2]

2. Your child prefers not to ask .13 .15

for something when they are not

sure they will obtain it

4. Your child is often afraid of .07 .45

new or unexpected situations

6. Your child is a shy person -.04 .62

8. Whenever possible, your child -.15 .54

avoids demonstrating their

skills for fear of being

embarrassed

10. When in a group, your child has -.33 .34

difficulty thinking of something

to say

12. Whenever they can, your child .01 .24

avoids going to unfamiliar

places

14. Your child often worries about .14 .41

things they said or did

16. It is difficult for your child -.06 .40

to talk with someone they do not

know

18. Your child generally tries to -.07 .50

avoid speaking in groups

20. Your child could do more things .14 .56

if it were not for their fear

22. Your child is afraid of many .05 .53

things compared to other

children their age

24. Your child often refrains from .01 .43

doing something they like in

order not to be rejected or

disapproved of by others

26. Your child often refrains from .12 .48

doing something because of fear

of being embarrassed

28. If your child thinks that .07 .28

something unpleasant is going to

happen, they get pretty worked

up.

30. Criticism or scolding hurts your .15 .13

child very much

1. The good prospect of obtaining a .50 .25

reward motivates your child

strongly to do some things

3. Your child often does things to .52 .36

be praised

5. Your child enjoys being the .54 .50

center of attention

7. When your child is in a group, .38 .41

they try to stand out as the

smartest or the funniest

9. When your child gets something .39 .14

they want, they feel excited and

energized

11. Your child does a lot of things .48 .36

for approval

15. Does your child generally prefer .48 .49

activities that involve

immediate reward

13. The possibility of obtaining -.04 .47

social status moves your child

to action, even if this involves

not playing fair

17. Your child often has trouble -.11 .49

resisting the temptation of

doing forbidden things

21. Your child has a lot of -.01 .32

difficulty ending a fun activity

23. Your child sometimes does things .41 .58

for quick reward

25. Your child has difficulty -.11 .47

staying focused on their school

work in the presence of an

attractive alternative

27. Your child engages in risky -.30 .40

behavior to obtain a reward

33. Your child craves excitement and .04 .29

new sensations

19. Your child likes to compete and .15 .54

do everything they can to win

29. Your child likes competitive -.15 .42

activities

31. Your child would like to be a .27 .58

socially powerful person

32. Your child likes displaying -.18 .34

their physical abilities even

though it may involve danger

Percent variance 8%

Note. Highest factor loading for each item is in bolded print.

Table III. Descriptive Statistics for the SP and SR Scales and for

Internalizing and Externalizing Behavior Problems

M SD Min Max

Sensitivity to Punishment and Sensitivity

to Reward Questionnaire (N = 59)

SP -.00 1.01 -1.63 2.58

SR-impulsivity/fun seeking .08 1.00 -2.18 2.96

SR-drive .06 1.03 -2.63 2.16

SR-reward responsiveness .11 0.96 -2.74 2.41

Child Behavior Checklist (N = 63)

Internalizing behavior problems .29 0.22 0 0.87

Externalizing behavior problems .31 0.26 0 1.27

ACKNOWLEDGMENTS

Thanks to Cesar Avila and Don Fowles for sharing details about their tasks, and to Larry Hawk for advice in analyzing physiological data. We also acknowledge Paula Fite and Patrick Nowlin for help in scheduling and running participants. This study was supported by grant R03DA14386 from the National Institute on Drug Abuse.

Received September 17, 2002; revision received December 12, 2003; accepted February 9, 2004

(3) Recovery of population factors based on sample data is influenced by several issues including sample size, ratio of observed to latent variables, and communalities. Based on the simulation work of MacCallum, Widaman, Zhang, and Hong (1999), with a sample size of 109, communalities that ranged from .13 to .62, and a ratio of 33 observed variables to four factors, our EFA likely provided “good” recovery of the population factors.

(4) Analyses were run with and without this child and there was no substantive change in the results, and so results are reported with this child included.

(5) There is some indication in the literature that conduct disorder maybe more strongly linked to a strong BAS, whereas ADHD maybe more strongly linked to a weak BIS. Accordingly, we formed an attention problems scale and ran all of our analyses with this measure. The pattern of findings was similar to those conducted with externalizing behavior problems, although relationships were generally weaker for the attention problems variable. There was no evidence of a relationship between a weak BIS and attention problems, however, it should be noted that attention problems as measured on the CBCL is a limited measure of ADHD symptoms.

(6) We also examined the potential interaction between sensitivity to punishment and impulsivity/fun seeking predicting passive avoidance errors, and this interaction was not statistically significant (p > .90).

REFERENCES

Achenbach, T. M. (1991). Manual for the child behavior checklist/4-18 and 1991 profile. Burlington, VT: University of Vermont Department of Psychiatry.

Alken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage.

Avila, C. (2001). Distinguishing BIS-mediated and BAS-mediated disinhibition mechanisms: A comparison of disinhibition models of Gray (1981, 1987) and of Patterson and Newman (1993). Journal of Personality and Social Psychology, 80, 311-324.

Beauchaine, T. P. (2001). Vagal tone, development, and Gray’s motivational theory: Toward an integrated model of autonomic nervous system functioning in psychopathology. Development and Psychopathology, 13, 183-214.

Beauchaine, T. P., Katkin, E. S., Strassberg, Z., & Snarr, J. (2001). Disinhibitory psychopathology in male adolescents: Discriminating conduct disorder from attention-deficit/hyperactivity disorder through concurrent assessment of multiple autonomic states. Journal of Abnormal Psychology, 110, 610-624.

Biederman, J., Rosenbaum, J. F., Hirshfeld, D. R., & Faraone, S. V., et al. (1990). Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders. Archives General Psychiatry, 47, 21-26.

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and data analysis methods. Thousand Oaks, CA: Sage.

Carver, S. C., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psychology, 67, 319-333.

Caspi, A. (2000). The child is father of the man: Personality continuities from childhood to adulthood. Journal of Personality and Social Psychology, 78, 158-172.

Clark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100, 316-336.

Colder, C. R., Mott, J. A., & Berman, A. S. (2001). The interactive effect of infant activity level and fear on growth trajectories of early childhood symptomatology. Development and Psychopathology, 14, 1-23.

Crone, E. A., Jennings, J. R., & van der Molen, M. W. (2003). Sensitivity to interference and response contingencies in attentiondeficit/hyperactivity disorder. Journal of Child Psychology and Psychiatry and Allied Disciplines, 44, 214-226.

Daugherty, T. K., & Quay, H. C. (1991). Response perseveration and delayed responding in childhood behavior disorders. Journal of Child Psychology and Psychiatry, 32, 453-461.

Di Nardo, P A., & Barlow, D. H. (1990). Syndrome and symptom cooccurrence in the anxiety disorders. In J. D. Maser & R. C. Cloninger (Eds.), Comorbidity of mood and anxiety disorders (pp. 205-230). Washington, DC: American Psychiatric Press.

Epstein, J. N., Goldberg, N. A., Conners, C. K., & March, J. S. (1997). The effects of anxiety on continuous performance test functioning in an ADHD clinic sample. Journal of Attention Disorders, 2, 45-52.

Farrington, D. P., & West, D. J. (1993). Criminal, penal and life histories of chronic offenders: Risk and protective factors and early identification. Criminal Behavior and Mental Health, 3, 492-523.

Fowles, D. C. (1980). The three arousal model: Implications of Gray’s two-factor learning theory for heart rate, electrodermal activity, and psychopathy. Psychophysiology, 17, 87-104.

Fowles, D. C. (1983). Appetitive motivational influences on heart rate. Personality and Individual Differences, 4, 393-401.

Fowles, D. C. (1988). Psychophysiology and psychopathology: A motivational approach. Psychophysiology, 25, 373-391.

Fowles, D. C., Kochanska, G., & Murray, K. T. (2000). Electrodermal activity and temperament in preschool children. Psychophysiology, 37, 777-787.

Frick, P. J. (1998). Conduct disorders and severe antisocial behavior. New York: Plenum.

Gray, J. A. (1987). The psychology fear and stress (2nd ed.). Cambridge, UK: Cambridge University Press.

Henriques, J. B., Glowacki, J. M., & Davidson, R. J. (1994). Reward fails to alter response bias in depression. Journal of Abnormal Psychology, 103, 460-466.

Heubeck, B. G., Wilkinson, R. B., & Cologon, J. (1998). A second look at Carver and White’s (1994) BIS/BAS scales. Personality and Individual Differences, 25, 785-800.

Iaboni, F., Douglas, V. I., & Baker, A. G. (1995). Effects of reward and response costs on inhibition in ADHD children. Journal of Abnormal Psychology, 104, 232-240.

Iaboni, F., Douglas, V. I., & Ditto, B. (1997). Psychophysiological response of ADHD children to reward and extinction. Psychophysiology, 34, 116-123.

Jennings, J. R. (1992). Is it important that the mind is in the body? Inhibition and the heart. Psychophysiology, 29, 369-383.

Jorm, A. F., Christensen, H., Henderson, A. S., Jacomb, P. A., Korten, A. E., & Rodgers, B. (1999). Using the BIS/BAS scales to measure behavioural inhibition and behavioural activation: Factor structure, validity and norms in a large community sample. Personality and Individual Differences, 26, 49-58.

Kagan, J., Snidman, N., Zentner, M., & Peterson, E. (1999). Infant temperament and anxious symptoms in school age children. Development and Psychopathology, 11, 209-224.

Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm, A. O. (1993). Looking at pictures: Affective, facial, visceral, and behavioral reactions. Psychophysiology, 30, 261-273.

Lengua, L. J., West, S. G., & Sandler, I. N. (1998). Temperament as a predictor of symptomatology in children: Addressing contamination of measures. Child Development, 69, 164-181.

Littell, R. C., Milliken, G. A., Stroup. W. W., & Wolfinger, R. D. (1996). SAS system for mixed models. Cary, NC: SAS Institute.

Long, J. (1998). IBI analysis system: Reference guide. Caroga Lake, NY: James Long.

MacCallum, R. C., Widaman, K. F., Zhang, S., & Hong, S. (1999). Sample size in factor analysis. Psychological Methods, 4, 84-99.

Martin, R. P., & Bridger, R. C. (1999). Temperament Assessment Battery for Children-Revised: A tool for the assessment of temperamental traits and types of young children. Unpublished manual.

Masse, L. C., & Tremblay, R. E. (1997). Behavior of boys in kindergarten and the onset of substance use during adolescence. Archives of General Psychiatry, 54, 62-68.

Matthys, W., van Goozen, S. H. M., de Vries, H., Cohen-Kettenis, P. T., & van Engeland, H. (1998). The dominance of behavioural activation over behavioural inhibition in conduct disordered boys with or without attention deficit hyperactivity disorder. Journal of Child Psychology and Psychiatry, 39, 643-651.

Newman, J. P., & Wallace, J. F. (1993). Diverse pathways to deficient self-regulation: implications for disinhibitory psychopathology in children. Clinical Psychology Review, 13, 699-720.

Newman, J. P., Widom, C. S., & Nathan, S. (1985). Passive avoidance in syndromes of disinhibition: Psychopathy and extraversion. Journal of Personality and Social Psychology, 48, 1316-1327.

O’Brien, B. S., & Frick, P. J. (1996). Reward dominance: Associations with anxiety, conduct problems, and psychopathy in children. Journal of Abnormal Child Psychology, 24, 223-240.

O’Brien, B. S., Frick, P. J., & Lyman, R. D. (1994). Reward dominance among children with disruptive behavior disorders. Journal of Psychopathology and Behavioral Assessment, 16, 131-145.

O’Connor, R. M., & Colder, C. R. (in press). Confirmatory factor analysis of the sensitivity to punishment and sensitivity to reward questionnaire. Personality and Individual Differences.

Oosterlaan, J., & Sergeant, J. A. (1998). Effects of reward and response cost on response inhibition in AD/HD, disruptive, anxious, and normal children. Journal of Abnormal Child Psychology, 26, 161-174.

Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys. Malden, MA: Blackwell.

Patterson, M. C., & Newman, J. P. (1993). Reflectivity and learning from aversive events: toward a psychological mechanism for the syndromes of disinhibition. Psychological Review, 100, 716-736.

Porges, S. W. (1991). Vagal tone: An automatic mediator of affect. In J. Garber & K. A. Dodge (Eds.). The development of emotion regulation and dysregulation (pp. 111-128). Cambridge, UK: Cambridge University Press.

Quay, H. C. (1988). The behavioral reward and inhibition system in childhood behavior disorder. In H. C. Quay (Ed.), Attention deficit disorder, Vol. 3: New research in attention, treatment, and psychopharmacology (pp. 176-186). Oxford, England: Pergamon Press.

Quay, H. C. (1993). The psychobiology of undersocialized aggressive conduct disorder: A theoretical perspective. Development and Psychopathology, 5, 165-180.

Quay, H. C. (1997). Inhibition and attention deficit hyperactivity disorder. Journal of Abnormal Child Psychology, 25, 7-13.

Raskin, D. C. (1973). Attention and arousal. In W. F. Prokasy & D. C. Raskin (Eds.), Electrodermal activity in psychological research (pp. 125-156). London: Academic Press.

Russo, M. F., Lahey, B. B., Christ, M. A., Frick, P. J., McBurnett, K., & Walker, J. L., et al. (1991). Preliminary development of a sensation seeking scale for children. Personality and Individual Differences, 12, 399-405.

Sanderson, W. C., Di Nardo, P. A., Rapee, R. M., & Barlow, D. H. (1990). Syndrome comorbidity in patients diagnosed with a DSM-III-R anxiety disorder. Journal of Abnormal Psychology, 99, 308-312.

Shapiro, S. K., Quay, H. C., Hogan, A. E., & Schwartz, K. P. (1988). Response perseveration and delayed responding in undersocialized aggressive conduct disorder. Journal of Abnormal Psychology, 97, 371-373.

Torrubia, R., Avila, C., Molto, J., & Caseras, X. (2001). The Sensitivity to Punishment and Sensitivity to Reward Questionnaire (SPSRQ) as a measure of Gray’s anxiety and impulsivity dimensions. Personality and Individual Differences, 31, 837-862.

Tranel, D. T. (1983). The effects of monetary incentive and frustrative nonreward on heart rate and electrodermal activity. Psychophysiology, 20, 652-657.

Tremblay, R. E., Pihl, R. O., Vitaro, F., & Dobkin, P. L. (1994). Predicting early onset of male antisocial behavior from preschool behavior. Archives General Psychiatry, 51, 732-739.

Turner, S. M., Beidel, D. C., & Epstein, L. (1991). Vulnerability and risk for anxiety disorders. Journal of Anxiety Disorders, 5, 151-166.

Walker, J. L., Lahey, B. B., Russo, M. F., Frick, P. J., Christ, M. A. G., & McBurnette, K., et al. (1991). Anxiety, inhibition, and conduct disorder in children: I. Relations to social impairment. Journal of the American Academy of Child and Adolescent Psychiatry, 30, 187-191.

Craig R. Colder (1,2) and Roisin M. O’Connor (1)

(1) State University of New York at Buffalo, Buffalo, New York.

(2) Address all correspondence to Craig Colder. Department of Psychology, Box 604110, State University of New York at Buffalo. Buffalo, New York 14260-4110; e-mail: ccolder@buffalo.edu.

COPYRIGHT 2004 Plenum Publishing Corporation

COPYRIGHT 2004 Gale Group