Functional impulsivity and reinforcement sensitivity theory.

Abstract

In this article, we attempt to integrate Dickman's (1990) descriptive concept of Functional Impulsivity (FI) with Gray's (1970, 1991) Reinforcement Sensitivity Theory (RST). Specifically, we consider that FI bears great conceptual similarity to Gray's concept of reward-reactivity, which is thought to be caused by the combined effects of a Behavioral Activation System (BAS) and Behavioral Inhibition System (BIS). In our first study, we examine the construct validity and structural correlates of FI. Results indicate that FI is related positively to measures of BAS and Extraversion, negatively to measures of BIS and Neuroticism, and is separate from Psychoticism and typical trait Impulsivity, which Dickman calls Dysfunctional Impulsivity (DI). In our second study, we use a go/no-go discrimination task to examine the relationship between FI and response bias under conditions of rewarding and punishing feedback. Results indicate that FI, along with two measures of BAS, predicted the development of a response bias for the rewarded alternative. In comparison, high DI appeared to reflect indifference toward either reward or punishment. We consider how these findings might reconcile the perspectives of Gray and Dickman and help clarify the broader understanding of Impulsivity.

Full text

Page 1

Functional Impulsivity and Reinforcement
Sensitivity Theory
Luke D. Smillie and Chris J. Jackson
University of Queensland, Australia
ABSTRACT
descriptive concept of Functional Impulsivity (FI) with Gray’s (1970,
1991) Reinforcement Sensitivity Theory (RST). Specifically, we consider
that FI bears great conceptual similarity to Gray’s concept of reward-
reactivity, which is thought to be caused by the combined effects of a
Behavioral Activation System (BAS) and Behavioral Inhibition System
(BIS). In our first study, we examine the construct validity and structural
correlates of FI. Results indicate that FI is related positively to measures
of BAS and Extraversion, negatively to measures of BIS and Neurotic-
ism, and is separate from Psychoticism and typical trait Impulsivity,
which Dickman calls Dysfunctional Impulsivity (DI). In our second study,
we use a go/no-go discrimination task to examine the relationship be-
tween FI and response bias under conditions of rewarding and punishing
feedback. Results indicate that FI, along with two measures of BAS,
predicted the development of a response bias for the rewarded alternative.
In comparison, high DI appeared to reflect indifference toward either re-
ward or punishment. We consider how these findings might reconcile the
perspectives of Gray and Dickman and help clarify the broader under-
standing of Impulsivity.
In this article, we attempt to integrate Dickman’s (1990)
We are indebted to Dr. Alan Pickering of Goldsmiths College, The University of
London, for his many helpful comments on an earlier version of this article. We also
thank Paul Jackson, who wrote the experimental task used in Study 2.
Luke Smillie and Chris Jackson are at the School of Psychology, The University of
Queensland. All correspondence to Luke Smillie, School of Psychology, University of
Queensland, Brisbane Q4072, Australia. E-mail: luke@psy.uq.edu.au.
Journal of Personality 74:1, February 2006
r 2005, Copyright the Authors
Journal compilation r 2005, Blackwell Publishing, Inc.
DOI: 10.1111/j.1467-6494.2005.00369.x

Page 2

Impulsivity is widely regarded as having various negative or dys-
functional implications, being associated with rash actions, irrespon-
sibility, a failure to consider the consequences of one’s actions, and
even as having a key role in criminal behavior (Barratt, 1985; Claes,
Vertommen & Braspenning, 2000; Coscina, 1997; Dickman, 1990;
Doob, 1990; Eysenck & Gudjonsson, 1989; Levine & Jackson, 2004).
For instance, Mathias and Stanford (2003) define Impulsivity as ‘‘a
tendency towards acting without forethought, making quick cogni-
tive decisions and failing to appreciate circumstances beyond the
here-and-now’’ (p. 355). Similarly, within the broader domains of
personality, Impulsivity tends to be aligned with less adaptive traits
such as Psychoticism (Eysenck, Barrett, Wilson, & Jackson, 1992;
Zuckerman, 1979) or low Conscientiousness (Costa & McCrae,
1992). Considering that there is in fact very little consensus about
what actually constitutes Impulsivity (Claes et al., 2000; Miller,
Joseph & Tudway, 2004; Parker & Bagby, 1997), it is interesting that
the negative view of this trait is so pervasive.
Dickman (1990) directly challenged this traditional view of Imp-
ulsivity and developed two scales to measure what he terms Func-
tional Impulsivity (FI) and Dysfunctional Impulsivity (DI).
Conceptually, Dickman argued that, while some forms of Imp-
ulsivity have negative or dysfunctional consequences (e.g., failing to
‘‘look before you leap’’), other forms have clearly positive or func-
tional consequences (e.g., the tendency to ‘‘seize the moment’’). Ac-
cording to this conceptualization, Dickman suggests that DI reflects
the more usual understanding of this trait, capturing reckless or rash
behavior with a lack of forethought (e.g., Item 8: I frequently make
appointments without thinking about whether I will be able to keep
them). Conversely, FI is conceptually and statistically distinct from
DI (Dickman, 1990, 2000) and concerns rapid responding to situa-
tional demands in order to optimize one’s circumstances (e.g., Item
15: I like sports and games in which you have to choose your next move
very quickly). In the broad spaces of personality, FI is most strongly
related to Extraversion, while DI is most strongly related to Psy-
choticism (Chico, 2000; Chico, Tous, Lorenzo-Seva, & Vigil-Colet,
2003). To support the adaptiveness of FI, Dickman demonstrates
that high scorers report greater benefits from their behavior than do
dysfunctional impulsives (Dickman & Meyer, 1988) and exhibit rela-
tively higher performance on basic cognitive and motor tasks (Dick-
man, 1990). The important characteristic of FI, which sets it apart
2
Smillie & Jackson

Page 3

from almost all other varieties of Impulsivity, is that the impulsive
behavior ultimately serves some positive outcome.
While Dickman’s work offers an interesting and potentially useful
organizing framework for what has become a complex cluster of
Impulsivity traits, the nature and causes of FI, in particular, remain
unknown. Not surprisingly, the literature concerning the bases of
Impulsivity has been chiefly influenced by the more negative view of
this trait and, as such, has only attempted to identify casual subst-
rates relating to maladaptive manifestations of Impulsivity (Coscina,
1997; Dickman, 2000). For example, substantial research has exam-
ined the role of dopamine and serotonin as regulators of impulsive
behavior. Processes involving serotonin have been linked to aggres-
sion (for a review, see Young, Phil, & Ervin, 1988), substance abuse
(e.g., Moeller et al., 1994), and antisocial behavior in criminal pop-
ulations (Dolan & Anderson, 2003). Similarly, dopamine processes
have been implicated in a number of theoretical frameworks as a basis
for Impulsivity, typically in relation to substance abuse (e.g., Peter-
son, Wolf, & White, 2003), disorders of disinhibition such as Atten-
tion Deficit/Hyperactivity Disorder (ADHD; e.g., Davids, Zhang,
Tarazi, & Baldessarini, 2003), and antisocial behavior (Goldman &
Fishbein, 2000). However, as has been noted elsewhere (Coscina,
1997), none of this research to date seems to indicate a basis for more
functional aspects of Impulsivity.
Recently, Dickman (2000) compared four arousal hypotheses of
Impulsivity and found reasonable support for his ‘‘attentional-fixi-
ty’’ theory (Dickman, 1993). This posits that Impulsivity is associ-
ated with an inability to attend to key features of a decision-making
situation, thus explaining the tendency for highly impulsive individ-
uals to act with little forethought. This hypothesis is not dissimilar
from other cognitive explanations of Impulsivity (and psychopathy,
e.g., Abramowitz, Kosson, & Seidenberg, 2004), and, like all four
theories examined by Dickman, was implicitly related only to typical
(i.e., dysfunctional) Impulsivity. However, Dickman (2000) noted a
modest correlation between FI and measures of energetic arousal
and speculated that this may improve the ability of functional im-
pulsives to attend more effectively to the decision-making properties
of the task at hand. An alternative interpretation of this data,
however, is that FI is related to motivation and emotion processes.
As has been noted elsewhere (e.g., Revelle, 1993), Thayer’s (1978,
1989) concept of ‘‘energetic arousal’’ is closely linked with positive
Functional Impulsivity and RST
3

Page 4

affect (e.g., Watson & Tellegen, 1985) and behavioral approach
(Fowles, 1980; Gray, 1987a). If Dickman’s observation was reliable,
therefore, it could indicate that FI has some similarity to traits that
have been related to positive incentive motivation (e.g., as reviewed
by Depue & Collins, 1999).
Functional Impulsivity and Behavioral Activation
Jeffrey Gray’s (1970, 1981, 1987a) Reinforcement Sensitivity Theory
(RST) is arguably the most influential motivation-based theory of
personality. Furthermore, other than Dickman’s concept of FI, RST
is one of the few personality models that does not demand a strictly
negative view of Impulsivity. Gray suggested that Impulsivity is
caused by reward motivation, regulated by the Behavioral Activa-
tion System (BAS; Fowles, 1980). While, originally, it was thought
that the BAS was responsive only to conditioned stimuli (Gray,
1987b), the scope of this system has now been broadened (Gray &
McNaughton, 2000), such that the BAS is responsive to all positively
valenced stimuli (presentation of reward or termination/omission of
punishment). Activation of the BAS by reward is thought to produce
an output to the motor cortex and an increase in arousal, enabling
the individual to approach the source of reward. BAS-Impulsive
individuals are thought to be more reactive to and more strongly
motivated by positive incentive stimuli, such that they are disposed
toward approaching situations likely to bring reward (and/or relief
from punishment). Gray’s perspective on Impulsivity has some sim-
ilarity with mainstream views, such as the involvement of do-
paminergic processes (Pickering & Gray, 2001) and the suggestion
that Impulsivity has links with psychopathology (Gray, Feldon,
Rawlins, Hemsley, & Smith, 1991). However, the concept of reward-
reactivity is clearly distinct from more common conceptualizations
of Impulsivity. In fact, while the typical view of Impulsivity empha-
sizes a failure to consider the outcomes of one’s actions, the RST
view emphasizes the specific direction of behavior towards certain
outcomes, namely, rewarding or positive stimuli. In this sense, BAS-
mediated Impulsivity is conceptually quite similar to Dickman’s
notion of FI.
It is pertinent to the subject of this article to note that the kind of
Impulsivity related to reactivity of the BAS has never been entirely
clear (Diaz & Pickering, 1993; Pickering & Gray, 2001). This is to be
4
Smillie & Jackson

Page 5

expected given the considerable heterogeneity among common def-
initions of this trait in the literature. In addition, Impulsivity was
specified as a potential manifestation of BAS reactivity, but initial
research concerning the BAS did not serve the exclusive purpose of
deriving an explanation of trait Impulsivity (Gray, Owen, Davis &
Tsaltas, 1983). Rather, the BAS was linked to Impulsivity in an ad
hoc manner (Diaz & Pickering, 1993), and many have suggested that
‘‘Impulsivity’’ as referred to by Gray (1981) is not the same trait
referred to more generally by that name (e.g., Dawe & Loxton, 2004;
Depue & Collins, 1999; Quilty & Oakman, 2004; Smillie & Jackson,
2005; Zelenski & Larsen, 1999). Therefore, just as FI lacks a theo-
retical basis, the motivational processes comprising RST are not
clearly delineated at the trait level ( Jackson, 2002, 2003). The con-
ceptual similarity and potential theoretical convergence of the views
of Gray and Dickman is therefore very interesting.
Functional Impulsivity and Behavioral Inhibition
In addition to the BAS, the other central component of RST is the
Behavioral Inhibition System (BIS), which is understood to provide
the causal basis of Anxiety. Initially, the BIS was thought to be en-
gaged by conditioned signals of punishment, resulting in the inter-
ruption of ongoing behavior and simultaneous direction of attention
and arousal toward the potential threat. According to recent revi-
sions of the model (Gray & McNaughton, 2000), it is now specifi-
cally seen as being responsive to conflicting goals.1An example of
such conflict might include the simultaneous goals to approach and
avoid an appetitive stimulus previously associated with punishment,
such as food paired with electric shock in a classic animal learning
experiment. This stimulus would be expected to trigger behavioral
inhibition, enabling the animal to assess the situation further and
resolve the source of conflict (e.g., either continuing to approach the
stimulus or avoiding it altogether). In personality terms, it is posited
that BIS-reactive individuals are higher in trait Anxiety and will
1. It is important not to confuse Gray and McNaughton’s (2000) use of the word
‘‘goal’’ with the typical cognitive conceptualisation of this term within various
domains of psychology (e.g., ‘‘goal setting,’’ ‘‘goal orientation’’). Gray and
McNaughton define a ‘‘goal’’ as the conflation of a stimulus and a behavioral
response (p. 23).
Functional Impulsivity and RST
5

Page 6

therefore react more readily to potential sources of conflict, thus
disposing them towards cautious behavior and risk assessment.
Interestingly, FI appears to have some conceptual similarity to
BIS functioning as well as to the BAS. A critical feature of FI con-
cerns rapid responding to ‘‘situational demands’’ in order to opt-
imize ones’ circumstances (Claes et al., 2000; Dickman, 1990). An
example of a situational demand is ‘‘a car suddenly braking on a
crowded freeway’’ (Claes et al., 2000, p. 28), which may require im-
mediate action by the driver of the following car. Of importance here
is the potential for either action or inaction to produce a better or
worse result, and, to this extent, the term ‘‘situational demands’’
appears very similar to Gray and McNaughton’s concept of goal
conflict. In RST, it is the BIS that responds to conflict by inaction,
thus enabling further assessment of the situation. FI on the other
hand seems to concern rapid action to seize a window of opportunity.
To borrow Claes and colleagues’ example of a car suddenly braking
on a crowded freeway, this conflict may trigger behavioral inhibition
in a BIS-reactive (highly anxious) individual traveling in the car be-
hind, with the result that he or she might not respond in time to
avoid a collision. In contrast, a prototypical functionally impulsive
individual might not be expected to experience behavioral inhibition,
but rather to act spontaneously (e.g., by swerving) in order to avoid
a crash. In other words, the opportunistic nature of the functional
impulsive suggests that such individuals will be disposed toward less
behavioral inhibition. This being the case, FI might be expected to
negatively correlate with measures of BIS, in addition to the positive
association we have suggested it may have with measures of BAS.
Now, at this point, we should note that the BIS and BAS are
thought to be physiologically independent, and, as such, their man-
ifestations, namely, Anxiety and Impulsivity, have been traditionally
viewed as orthogonal constructs (Gray, 1981, 1987a; Gray et al.,
1983). Therefore, it might be argued from the RST perspective that if
FI is related to both the BIS and BAS, then it is a confounded
measure. Nevertheless, while the BIS and BAS are thought to be
physically independent, their neural and motor output is unlikely to
be orthogonal. In fact, Gray and Smith (1969) proposed mutual in-
hibitory pathways between the BIS and BAS, such that the func-
tioning of one system would attenuate the output of the other. From
this, some have proposed (Corr, 2001, 2002; Pickering, 1997) that the
BIS and BAS will combine to influence behavior. Specifically, re-
6
Smillie & Jackson

Page 7

ward-reactivity will be relatively stronger in BAS reactive individu-
als, but strongest in BAS1/BIS- individuals. Corr (2004) reviews
evidence for this ‘‘Joint Subsystems Hypothesis,’’ whereby reward-
reactivity is related to measures of BAS (positively) and BIS (neg-
atively). Based upon the conceptual similarity of FI to both BIS (1)
and BAS (?) processes, it might be suggested that FI reflects Gray’s
notion of reward-reactivity and could therefore have an underlying
basis in the structures of RST. This is a novel suggestion, which the
present research will endeavor to evaluate.
To summarize, we suggest that Dickman’s (1990) thesis of Imp-
ulsivity may offer a useful means to distinguish between two funda-
mental variations of this trait. Unfortunately, FI, in particular, is in
need of further construct validation and theoretical explanation.
Based upon the conceptual overlap, we identify between FI and the
motivational processes with which RST is concerned and we theorize
that FI may represent (or be highly similar to) Gray’s notion of re-
ward-reactivity. As such, we hypothesize that FI will be related to
measures of BAS (1) and BIS (?) and should also predict behavi-
oral reactions to reward. In Study 1, we investigate bivariate and
structural relationships of FI and DI with specific measures of RST
and broader scales of personality. In Study 2, we use a go/no-go task
to examine the ability of FI and DI, along with purpose-built meas-
ures of BAS, to predict response bias during a decision-making task
employing schedules of rewarding or punishing feedback. This re-
search aims to clarify both the conceptual and substantive nature of
FI and may also help to clarify the kind of Impulsivity to which RST
is relevant.
STUDY 1
To date, surprisingly few studies have examined the construct
validity of FI and DI or the location of Dickman’s dimensions with-
in the factor spaces of personality. Basic correlational studies suggest
that FI is most strongly related to Extraversion, while DI appears
most strongly related to Psychoticism (Chico, 2000; Chico et al.,
2003). This supports the conceptualization of DI as representative of
typical trait Impulsivity, which tends to be viewed as a component of
Psychoticism (Eysenck et al., 1992). It is also consistent with our
suggestion that FI has some similarity with the BAS, which Gray
Functional Impulsivity and RST
7

Page 8

positioned at a 30orotation from Extraversion (Pickering, Corr, &
Gray, 1999) and which is often measured using Extraversion (e.g.,
Gupta, 1990; Pickering, 2004). Nevertheless, very little research has
examined correlates of FI using specific measures of RST. In a com-
parative psychometric study, Jackson and Smillie (2004) investigated
relationships among a number of Impulsivity and BAS scales, in-
cluding the Dickman Impulsivity Inventory (DII; Dickman, 1990).
FI was strongly related to BAS measures but also negatively related
to BIS measures. Somewhat similar findings were observed by Go-
mez, Cooper, McOrmond, and Tatlow (2004), whereby FI was as-
sociated positively with the processing of pleasant stimuli (putatively
BAS-mediated) but negatively with processing of unpleasant stimuli
(putatively BIS-mediated). These findings suggest that there may in-
deed be some overlap between FI and reward-reactivity (BAS1/
BIS?).
Our first study seeks to build upon the dearth of research into the
construct validity and factor structure of FI and DI. Our specific
focus will be to examine the overlap of FI with measures of RST and
the extent to which it can be considered independent from typical
Impulsivity (e.g., DI). We predict that (1) FI will relate positively to
measures of BAS and negatively to measures of BIS; and (2) DI will
be related to Psychoticism and other trait Impulsivity measures, but
will have a less clear relationship with measures of RST. Although
these predictions derive from our theoretical position that FI may
have a causal basis in RST, a direction of causality is not tested
because all measures were taken concurrently.
Method
Participants
A total of 299 undergraduate psychology students, enrolled at the Uni-
versity of Queensland, participated in Study 1. Some completed ques-
tionnaires as part of their practical assessment in a psychology course,
while others were recruited via a first-year psychology research partici-
pation pool. As questionnaires were administered over a period of weeks,
sample size varied between scales due to differing attendance, and not all
participants completed all scales. The results section indicates sample size
resulting from pair-wise deletion of cases, giving well over 100 data points
for most analyses conducted. The mean age of the sample was 23.1 years
(SD53.16), and 65% were female.
8
Smillie & Jackson

Page 9

Questionnaires
Functional and dysfunctional Impulsivity.
Dickman Impulsivity Inventory (DII: Dickman, 1990) using a 6-point
scale (15completely disagree, 65completely agree). While this measure
is typically responded to using a True/False format, we had wanted to
maximize variance for the purpose of this exploratory study. The FI scale
consisted of 11 items concerning opportunistic or directed impulsive be-
havior (e.g., Item 4: ‘‘I would enjoy working at a job that required me to
make a lot of split-second decisions’’). The DI scale consisted of 12 items
concerning reckless or undirected impulsive behavior (e.g., Item 2: ‘‘I of-
ten get into trouble because I don’t think before I act’’). Both scales have
been found to have adequate internal consistency (alphas 4.80 are typ-
ically reported, e.g., Chico et al., 2003).
Most participants also completed all or some of the following person-
ality questionnaires:
All participants completed the
RST measures.
2004) is a recently developed 20-item measure of BAS/reward-reactivity.
Psychometric and experimental investigation suggests that the measure
has good general psychometric properties ( Jackson & Smillie, 2004) and
predicts behavioral reactions to reward (Smillie & Jackson, 2005). The
Sensitivity to Reward and Punishment Questionnaire (SPSRQ; Torrubia,
Avila, Molto, & Caseras, 2001) consists of 25 items to index Sensitivity to
Reward (BAS) and 25 items to index Sensitivity to Punishment (BIS), and
shows good reliability and validity (e.g., Avila, 2001). The BIS/BAS Scales
(Carver & White, 1994) is a widely used RST questionnaire, consisting of
one 7-item BIS scale (CWBIS) and three 4- to 5-item BAS scales. Due to
the low (o.70) alphas that are commonly observed for some of the
CWBAS scales (e.g., Smillie & Jackson, 2005), we use the sum total of the
three scales as is common practice. (In no analysis did this result in a
substantive change of interpretation as, overall, the three scales did not
diverge in their prediction of relevant criteria.) The Gray-Wilson Person-
ality Questionnaire (GWPQ; Wilson, Gray, & Barrett, 1990) is the earliest
purpose-built measure of RST, consisting of six 20-item scales to assess
BIS (passive avoidance and extinction), BAS (approach and active avoid-
ance), and Fight/Flight (fight and flight). The GWPQ has consistently
poor psychometric properties ( Jackson, 2002, 2003), but it continues
to hold great theoretical interest due to the fact that the items are direct
human analogues of the animal behavior upon which RST was first based.
The Appetitive Motivation scale ( Jackson & Smillie,
Three- and five-factor personality measures.
Questionnaire Revised (EPQ-R: Eysenck & Eysenck, 1991) is a widely
The Eysenck Personality
Functional Impulsivity and RST
9

Page 10

used, broad-focus measure of personality, considered to provide the
benchmark measures of Extraversion, Neuroticism, and Psychoticism.
It is also of particular relevance to the present study due to the fact that
Extraversion is a common proxy measure of the BAS and Neuroticism is
a common proxy measure of the BIS (e.g., Gupta, 1990; Pickering, 2004).
The Eysenck Personality Profiler (EPP; Eysenck et al., 1992) is a less
commonly used measure of Eysenck’s model and divides each factor into
seven narrow-focus scales, each consisting of 20 items. The NEO-FFS
(Costa & McCrae, 1992), and the Big Five Inventory ( John, Donahue, &
Kentle, 1991) are two widely used inventories for assessing the five factors
of Extraversion, Neuroticism, Conscientiousness, Agreeableness, and
Openness.
Statistical Analyses
Our first strategy was to identify scales from the questionnaires that cor-
related significantly with either FI or DI. These correlation patterns were
used to identify the overlap in content between Dickman’s scales and
other personality dimensions. We then used Exploratory Factor Analysis
(via Principal Components Analysis) to investigate the alignment of FI
and DI within the factor spaces of personality. It was expected that FI
would be aligned between BIS (low scores on purpose-built BIS measures
and EPQ Neuroticism) and BAS (high scores on purpose-built BAS
measures and EPQ Extraversion), while DI would be aligned with typical
Impulsivity (high scores on trait Impulsivity measures and EPQ Psycho-
ticism).
Results and Discussion
Preliminary Analyses
Cronbach’s alpha, means, and standard deviations are shown for all
scales in Table 1, and correlations between Dickman’s scales and the
other personality scales are shown in Table 2. FI shows high internal
consistency (alpha5.85) and was not correlated with DI, r5.06, ns.
Correlations With Functional Impulsivity
Consistent with our predictions and the findings of Jackson and
Smillie (2004), FI correlated with many indices of Gray’s BAS.
Significant loadings included Appetitive Motivation; Sensitivity to
Reward; EPP Impulsiveness; CWBAS; plus EPQ, EPP, and NEO
Extraversion. Relationships were also observed with measures that
10
Smillie & Jackson

Page 11

Table1
Means, Standard Deviations, and Alphas for Personality Scales
Nalpha MeanSD
(a) Dickman’s Impulsivity Inventory
DII Functional
DII Dysfunctional
(b) RST Measures
CWBAS
CWBIS
Appetitive Motivation Scale
Sensitivity to Reward
Sensitivity to Punishment
GWPQ Approach
GWPQ Active Avoidance
GWPQ Passive Avoidance
GWPQ Extinction
GWPQ Fight
GWPQ Flight
(c) High bandwidth personality
EPQ Psychoticism
EPQ Extraversion
EPQ Neuroticism
EPP Psychoticism
EPP Extraversion
EPP Neuroticism
NEO Openness
NEO Conscientiousness
NEO Extraversion
NEO Agreeableness
NEO Neuroticism
BFI Openness
BFI Conscientiousness
BFI Extraversion
BFI Agreeableness
BFI Neuroticism
(d) Low bandwidth personality
(EPP subscales)
Extraversion
EPP Activity
EPP Sociability
EPP Expressiveness
299
299
.85
.83
41.29
35.64
8.13
10.21
256
256
273
104
104
201
202
187
193
197
189
.77
.82
.74
.82
.73
.63
.45
.64
.64
.76
.76
39.67
20.86
14.61
11.08
11.55
18.48
22.39
19.06
20.55
16.96
18.33
4.71
3.87
3.24
3.23
4.99
6.06
5.63
5.90
5.61
6.50
7.79
202
201
201
127
127
127
111
57
119
113
109
175
175
175
174
173
.69
.82
.85
6.59
15.57
14.34
126.94
131.32
70.74
32.31
31.65
33.38
31.35
24.56
33.03
30.39
27.57
31.91
23.39
3.95
4.69
5.83
29.33
30.61
33.69
7.71
8.11
7.12
5.34
10.53
6.32
5.46
6.49
5.92
6.85
.78
.83
.86
.83
.88
.78
.81
.86
.77
.87
127
127
127
20.22
23.24
17.99
8.20
9.37
6.01
(Continued)
Functional Impulsivity and RST
11

Page 12

could be expected to associate with BAS activity (e.g., EPP primary
scales of Extraversion including Assertiveness, Ambition, Expres-
siveness, Activity, and Sociability). These patterns are consistent
with the notion that FI has some similarity with Gray’s (1987a)
conceptualization of the BAS. Also consistent with our hypothesis
were a number of strong inverse correlations between FI and various
benchmark measures of BIS. Specifically, significant (negative)
N alpha MeanSD
EPP Assertiveness
EPP Ambition
EPP Dogmatic
EPP Aggression
Neuroticism
EPP Low self-esteem
EPP Unhappiness
EPP Anxiety
EPP Dependence
EPP Hypochondria
EPP Guilt
EPP Obsessiveness
Psychoticism
EPP Risk taking
EPP Impulsiveness
EPP Irresponsibility
EPP Manipulativeness
EPP Sensation seeking
EPP Tough mindedness
EPP Practical
127
126
127
127
19.37
19.73
14.90
14.07
7.18
7.16
4.75
5.70
127
126
127
127
127
126
127
15.08
10.44
15.82
7.46
4.51
8.04
11.72
9.82
8.87
8.46
5.55
4.25
5.69
6.31
126
127
126
127
201
127
127
18.75
17.66
20.88
15.48
20.57
15.23
16.40
7.44
8.61
6.52
5.88
7.66
7.03
8.24
nAlpha coefficients were not computed for EPP scales as only total scores were
computed during data collection.
Note: DII5Dickman Impulsivity Inventory (Dickman, 1990); CWBIS and
CWBAS are described by Carver & White (1994); GWPQ5Gray-Wilson Person-
ality Questionnaire (Wilson et al., 1990); the Appetitive Motivation scale is de-
scribed in Jackson & Smillie (2004); EPQ5Eysenck Personality Questionnaire
Revised (EPQ-R: Eysenck & Eysenck, 1991); EPP5Eysenck Personality Profiler
(Eysenck & Wilson, 1992); Sensitivity to Reward and Punishment scales are from
the Sensitivity to Reward and Punishment Questionnaire (SPSRQ: Torrubia et al.,
2001); NEO5NEO-FFS (NEO-FFS: Costa & McCrae, 1992); BFI5Big Five
Inventory ( John et al., 1991).
Table1
(Continued)
12
Smillie & Jackson

Page 13

Table2
Correlations Between Dickman’s Scales of Impulsivity and Other
Personality Questionnaires
FunctionalDysfunctionalN
(a) Significant correlations with
Functional Impulsivity
CWBIS
Sensitivity to punishment
EPP Assertiveness
EPP Dependence
Appetitive Motivation
EPP Low self-esteem
EPP Anxiety
BFI Neuroticism
NEO Extraversion
NEO Neuroticism
GWPQ Passive Avoidance
EPQ Neuroticism
EPP Ambition
EPP Neuroticism
EPP Sociability
EPP Guilt
GWPQ Extinction
GWPQ Flight
(b) Significant correlations with
both types of Impulsivity
EPQ Extraversion
EPP Extraversion
BFI Extraversion
EPP Unhappiness
EPP Activity
EPP Impulsiveness
EPP Expressiveness
CWBAS
Sensitivity to reward
EPP Risk taking
GWPQ Active Avoidance
(c) Significant correlations with
Dysfunctional Impulsivity
EPP Psychoticism
BFI Conscientiousness
?.47nn
?.43nn
.43nn
?.43nn
.43nn
?.42nn
?.41nn
?.41nn
.39nn
?.37nn
?.37
?.36nn
.34nn
?.32nn
.28n
?.25n
?.24nn
?.23nn
.06 104
104
81
81
145
81
81
148
81
104
163
169
80
104
81
80
164
161
?.17
.22
?.09
.14
?.17
?.04
.01
.08
?.13
?.01
.07
?.07
?.15
.22
?.12
?.02
.05
.49nn
.48nn
.45nn
?.44nn
.40nn
.34nn
.34nn
.31nn
.25n
.24n
?.15n
.29nn
.25n
.37n
?.26n
.28n
.59nn
.40nn
.23n
.35nn
.35nn
?.16n
168
105
151
80
81
81
81
104
104
80
172
.13
.03
.47nn
?.47nn
104
150
(Continued)
Functional Impulsivity and RST
13

Page 14

loadings included GWPQ passive avoidance and extinction; EPP
Anxiety; Sensitivity to Punishment; CWBIS; plus the Neuroticism
scales of EPP, EPQ, and the NEO. Moreover, similar relationships
were observed with BIS related measures (e.g., EPP primary scales of
Unhappiness, Dependence, and Low Self-Esteem). Therefore, FI
also has some descriptive similarity with Gray’s notion of BIS.
Correlations With Dysfunctional Impulsivity
Correlations generally support Dickman’s (1990, 2000) claim that DI
reflects typical trait Impulsivity and findings by Chico et al. (2003)
that this scale is aligned with Psychoticism. Significant correlations
included EPQ and EPP Psychoticism; EPP Sensation Seeking; and
both (low) Conscientiousness and (low) Agreeableness from the
NEO and Big Five inventories. DI was not associated with any
measures of BIS but was significantly correlated with a number of
BAS measures (although these tended to be weaker than those
between FI and measures of BAS).
Factor Analysis
A factor analysis (via Principal Components Analysis) with varimax
rotation was conducted to determine the alignment of FI and DI
within orthogonal factor spaces formed from measures of BAS (and
Extraversion), BIS (and Neuroticism) and trait Impulsivity (and
Psychoticism). Specific scales selected from our battery of measures
FunctionalDysfunctionalN
EPP Irresponsibility
EPQ Psychoticism
NEO Conscientiousness
NEO Agreeableness
EPP Obsessiveness
EPP Sensation Seeking
GWPQ Approach
.20
.16
.44nn
.43nn
?.36n
?.30nn
?.38nn
.24n
.22nn
81
169
44
83
81
104
175
?.08
?.16
?.15
.18
.07
npo.05.nnpo.01.
Note: Only scales significantly correlated with at least one of Dickman’s scales are
represented.
Table2
(Continued)
14
Smillie & Jackson

Page 15

were DII Functional and Dysfunctional Impulsivity; EPQ Extraver-
sion, Neuroticism, and Psychoticism; SPSRQ Sensitivity to Reward
and Sensitivity to Punishment; Appetitive Motivation; CWBIS and
CWBAS; and EPP Sensation Seeking (whose definition in the EPP
manual is very similar to the notion of typical Impulsivity we have
considered in this article). Recalling that not all of our 299 partic-
ipants completed all measures, listwise deletion resulted in a total of
104 participants who completed all 11 scales (52% female).
The first three components extracted had eigenvalues greater than
1 (3.32, 2.60, 1.30), and together accounted for 65.3% of the total
variance. It can be seen from the rotated component matrix pre-
sented in Table 3 that the three factors clearly resemble the broad
dimensions we had expected. The first factor consists of CWBIS,
EPQ Neuroticism, Sensitivity to Punishment, and FI (?) and
yields an (unweighted) internal consistency of .80. As indicated by
our suggested labels, this factor seems to capture the conceptual di-
mension of the BIS. The second factor consists of CWBAS,
Appetitive Motivation, EPQ Extraversion, FI, and Sensitivity to
Reward and yields an (unweighted) internal consistency of .77. It
seems likely that the conceptual communality among these scales is
synonymous with the BAS. The third factor is composed of EPQ
Psychoticism, EPP Sensation Seeking, DI, and Sensitivity to Re-
ward. This factor might reflect typical Impulsivity or some key facet
of Psychoticism, and, for this reason, it is unexpected that Sensitivity
to Reward was also an important loading variable.2The lower (un-
weighted) internal consistency of this factor (.66) indicates that it
is less homogenous, perhaps reflecting the multifaceted and
unclear nature of both trait Impulsivity and Psychoticism (Eysenck
et al., 1992).
These findings contribute significantly toward the clarification of
Impulsivity. First, results provide some support for our hypotheses
concerning the structure of FI and DI and the possible bases we have
suggested for FI in particular. If factors 1 and 2 from our Principal
2. Note that results of recent factor analytic investigations using the Sensitivity to
Reward scale tend to be similar to ours in this respect (Caseras, Avila, & Torrubia,
2003). A possible explanation for this relates to the scale’s heterogeneous content,
perhaps making it a sufficiently broad-focus measure to encompass both BAS/
Reward Responsiveness and Impulsivity/Sensation-Seeking (Torrubia, personal
communication).
Functional Impulsivity and RST
15