Journal of Research in Personality 45 (2011) 655–661
Contents lists available at SciVerse ScienceDirect
Journal of Research in Personality
journal homepage: www.elsevier.com/locate/jrp
Appetitive and impulsive components in the Appetitive Motivation Scale
Luigi Leone a,⇑, Fridanna Maricchiolo b, Fabio Presaghi a
a
b
Department of Developmental and Social Psychology, ‘‘Sapienza’’ University of Rome, Italy
Department of Cultural and Educational Studies, Experimental Psychology Laboratory, University of Roma Tre, Italy
a r t i c l e
i n f o
Article history:
Available online 25 September 2011
Keywords:
BAS
Impulsivity
Reinforcement Sensitivity Theory
Structural equation modeling
Personality
a b s t r a c t
Measures of the Behavioral Activation System (BAS) often conflate reward-reactivity (the core of the BAS)
and impulsivity-related variance. The Appetitive Motivation Scale (AMS), a recently-developed BAS measure, has been proposed as a one-dimensional scale which deals exclusively with core reward-reactivity.
However, we show here how a two-dimensional interpretation of the scale reveals components which
are linked with both the core of the BAS and impulsivity. A clear pattern of associations and dissociations
with other BAS- and impulsivity-related measures provides support for such a two-dimensional interpretation. The results are discussed, highlighting the parallels between the AMS structure and other BASrelated measures, and emphasizing the theoretical challenges which are inherent in developing reliable
BAS-related personality measures.
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
Gray’s (e.g., 1982) Reinforcement Sensitivity Theory (RST) is a
widely influential theory of personality, linking individual differences in behavioral activation and inhibition to underlying biological structures. The RST conceptualizes three main neurobiological
systems which are allegedly linked to individual differences in personality but here we will focus solely on the Behavioral Activation
System (BAS), which is activated by positive stimuli and mediates
approach responses towards rewards. The BAS is a particularly relevant system for personality research because of its connections
with action, goal-pursuit and positive emotionality (Carver &
White, 1994). In addition, the system is theoretically appealing in
clinical research, because it has been suggested that an unchecked
hyperactive BAS may be conducive to various reckless behaviors
(O’Connor, Stewart, & Watt, 2009; Voigt et al., 2009; Zisserson &
Palfai, 2007). Finally, the BAS has been interpreted as the key motivational foundation of trait impulsivity and impulsive action (e.g.,
Pickering & Gray, 1999) and even as the basis of the impulsive
component in different forms of psychopathology (Newman, MacCoon, Vaughn, & Sadeh, 2005; Urosevic, Abramson, Harmon-Jones,
& Alloy, 2008).
However, measuring BAS reactivity has proven to be a difficult
and theoretically challenging task.
A first attempt to design a purpose-built RST measure (Wilson,
Gray, & Barrett, 1990) was hampered by serious psychometric
shortcomings (e.g., Carver & White, 1994). At the end of the
⇑ Corresponding author. Address: Dipartimento di Psicologia dei Processi di
Sviluppo e di Socializzazione, Via dei Marsi, 78, 00185 Roma, Italy.
E-mail address: luigi.leone@uniroma1.it (L. Leone).
0092-6566/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jrp.2011.09.003
1990s, the BAS scales developed by Carver and White (1994;
CW-BAS) provided the first purpose-built BAS measure to gain
widespread popularity, and to display satisfactory and replicable
psychometric properties (e.g., Jorm et al., 1999; Leone, Perugini,
Bagozzi, Pierro, & Mannetti, 2001). The CW-BAS comprises three
scales which are related to different functions and definitions of
the BAS (Carver & White, 1994). In spite of its success (it is indexed
as being the most frequently used BAS-related measure), the CWBAS has been criticized for failing to adequately represent the BAS:
it has been argued that the complex three-factor structure of the
CW-BAS sets it apart from the theoretical tenets of RST, which conceptualizes the BAS as a tighter one-dimensional construct (e.g.,
Caseras, Avila, & Torrubia, 2003; Jackson, 2009). Furthermore, the
ability of the CW-BAS scales to reflect the motivational properties
of the BAS has been questioned (e.g., Jackson & Smillie, 2004).
The intensity of the debate regarding how the BAS could and
should be measured reflects a real conceptual problem, one that
goes beyond psychometric quarrels. Indeed, there are theoretical
difficulties in building a trait-level measure of the BAS (Smillie,
Pickering, & Jackson, 2006). The BAS’ anatomical and neuro-chemical foundations are complex and not yet definitively identified,
and its emotional, motivational and behavioral consequences are
very broad and not necessarily homogeneous (Pickering & Smillie,
2008). In brief, it has proven to be difficult to translate the complex
motivational processes regulated by the BAS into a trait measure,
let alone a one-dimensional one (Jackson, 2003; Smillie & Jackson,
2006).
A recent purpose-built BAS-related measure, the Appetitive
Motivation Scale (AMS), was an attempt to directly tackle this difficulty (Jackson & Smillie, 2004). The authors intended to develop
indicators of the construct by dealing exclusively with the motiva-
656
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
tional role of appetitive stimuli in triggering approach behaviors
aimed at obtaining rewarding experiences, thereby clearly defining
the boundaries of the construct (Jackson & Smillie, 2004). Unlike
the CW-BAS scales, the AMS was designed as a one-dimensional
measure, thereby avoiding the possibility of the divergent concurrent and predictive validity results that are common when the
three CW-BAS scales are considered separately (Carver, 2004;
Smillie, Jackson, & Dalgleish, 2006). The AMS was also intended
to avoid the conflation between BAS-related and impulsivityrelated indicators that is typical of other popular BAS measures
(Cooper, Smillie, & Jackson, 2008; Smillie & Jackson, 2006), such
as the Sensitivity to Reward (SR) scale (Caseras et al., 2003). This
is a crucial point because a deep revision of the RST (r-RST; Gray
& McNaughton, 2000) clarifies that BAS should be highly related
with extraversion (e.g., Smillie, Pickering & Jackson, 2006) and
more closely linked with functional forms of impulsivity (Dickman,
1990) rather than with dysfunctional or rash impulsivity (e.g., Jackson, 2009; Smillie & Jackson, 2006).1 Recent measures explicitly
developed under the framework of the r-RST (Jackson, 2009) satisfy
such requirements; however, the popularity of measures developed
under the original RST (o-RST) among researchers makes it relevant
to investigate whether such measures could be refined to adequately
capture the theoretical definitions put forward by the r-RST (e.g.,
Heym et al., 2008; Smillie, Jackson et al., 2006).
A number of articles concerning the AMS scale validation and
validity have appeared (Cooper et al., 2008; Jackson & Smillie,
2004; Smillie & Jackson, 2005) providing some support for the
one-dimensionality of the scale – although the evidence is admittedly not decisive (Cooper et al., 2008) – and its predictive and construct validity. Nonetheless, the authors lamented a relative lack of
replication of their results by independent researchers. We will attempt in this study to accomplish this goal, by adopting the same
aims that were embraced by the AMS authors, shaping a BAS-related measure that: (1) addresses the motivational processes
which are supervised by the BAS and (2) avoids any confusion
and content-conflation with trait impulsivity, as required by the
r-RST. We maintain that a further revision of the structure of the
AMS will help to satisfy these two important theoretical and measurement criteria (Smillie, Pickering et al., 2006).
1.1. The AMS: theory, structure, validity
Table 1 presents the 20 items comprised in the AMS (Jackson &
Smillie, 2004). The authors stated that they intended to assess the
BAS with the aim of ‘‘. . . impulsively approaching appetitive stimuli
with the goal of obtaining a rewarding experience’’ (Jackson &
Smillie, 2004, p. 1628, emphasis added), a definition consistent
with the original RST. This emphasis on impulsivity faded away
in the subsequent articles about the scale, as the authors recognized the need to differentiate impulsivity, at least in its dysfunctional form, from the BAS (e.g., Smillie & Jackson, 2005; Smillie &
Jackson, 2006), more consistently with the r-RST (Smillie, Pickering
et al., 2006). Nevertheless, the early inclusion of impulsivity as a
defining feature of BAS-regulated behaviors substantially influenced the composition of the scale. Items 7, 10, 13, 16 and 18
are reminiscent of the typical sensation-seeking and lack of planning indicators which are found in several impulsivity measures.
A different cluster of items (# 1, 5, 8) instead reflects a reluctance
to adhere to norms and rules, a component that appears to be
linked with psychoticism, or with the more general socially unde1
The revised RST proposes crucial changes in the definitions and functions of the
Behavioral Inhibition System and of the Fight/Flight System (Gray & McNaughton,
2000; Heym, Ferguson, & Lawrence, 2008; Jackson, 2009; Smillie, Pickering et al.,
2006). Here we focus only on predictions involving the BAS as conceptualized within
the r-RST.
Table 1
Items of AMS by Jackson and Smillie (2004) and
items retained (⁄) in the Cooper et al. (2008)
revision.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Believe rules are stifling
Put plans into action
Like to be busy
Like to see how things work
Choose which rules to follow
Like things new and different⁄
Like to do things spontaneously⁄
Like to do things my way
Do several things at same time⁄
Important to enjoy moment
Actively look for new experiences⁄
Feel for how things work⁄
Look for new sensations⁄
Excited by what is new in my field⁄
Often lots of ideas⁄
Prefer not to plan too much
Like to be rewarded⁄
Here and now is important
New ideas all the time⁄
Enjoy new projects⁄
sirable consequences of impulsive behaviors aimed at obtaining
personal advantage, rather than with BAS activity per se.
In spite of the presence of distinct clusters of items, the authors
opted for a one-dimensional structure for the scale. However, the
presence of impulsivity- and psychoticism-related indicators
shaped the pattern of associations with concurrent validity criteria.
The AMS correlated with the Impulsivity scale of the Eysenck Personality Profiler (EPP; Eysenck, 1993), and with the Eysenck Personality Questionnaire (EPQ) Psychoticism scale. The AMS was
also associated with EPQ-Extraversion, as BAS measures are expected to do from a theoretical standpoint (e.g., Smillie, Pickering
et al., 2006). However, the magnitude of the AMS-Extraversion
association was almost identical to the AMS-Psychoticism correlation, which appears problematic (e.g., Jackson, 2009). Concurrent
associations with the CW-BAS scales also revealed the coexistence
in the AMS of both impulsivity- and BAS-related indicators. Recent
research has shown that the Drive and Reward-Responsiveness
scales of the CW-BAS are more closely linked to the core BAS-concept, whilst the Fun scale is more closely linked to rash impulsivity
(Heym & Lawrence, 2010; Leone, 2009; Smillie, Jackson et al.,
2006). Thus, the AMS should be more closely linked to Drive and
Reward-Responsiveness than to Fun scores. Instead, the AMS correlated slightly more strongly with Fun (Jackson & Smillie, 2004),
and therefore to impulsivity-related manifestations, than with
the other scales that more effectively represent the core functions
of the BAS in sustaining efforts to attain appetitive outcomes (Smillie, Jackson et al., 2006). As such associations were computed using
a small sample (N = 59), Smillie and Jackson (2005) sought replication with a bigger sample (N = 144). As it turned out, the AMS still
correlated more strongly with the CW-BAS Fun scale than with the
CW-BAS Reward-Responsiveness and Drive scales.
Smillie and Jackson (2005) clarified an important theoretical
point regarding the connections between the BAS and impulsivity.
In contrast to Jackson and Smillie (2004), Smillie and Jackson
(2005) clearly stated that the association between the BAS and trait
impulsivity may be deemed arbitrary, at best. The descriptive correspondence between BAS and impulsivity may stem from the simple fact that being motivated by reward may prompt – when the
right circumstances occur – quick impulsive action to seize the
incentive (Corr, 2008). However, it may be argued that this kind
of BAS-related impulsivity differs from the classical dysfunctional
strain of impulsivity. The relatively under-researched forms of
functional impulsivity (i.e., appropriate behavior intended to seize
the reinforcer – Dickman, 1990) may reflect BAS sensitivity more
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
directly (Smillie & Jackson, 2006). Recent evidence confirms the
similarities between functional impulsivity and the BAS (Leone &
Russo, 2009; Smillie & Jackson, 2006), and the clear distinction between the core BAS measures and rash and dysfunctional impulsivity (e.g., Smillie, Jackson et al., 2006; Leone, 2009). The divergent
associations of BAS measures with functional and dysfunctional
impulsivity may be interpreted as a crucial test in developing
and testing measures consistent with the revised RST (Jackson,
2009). In accordance with this view, the AMS was more strongly
connected with Functional Impulsivity than dysfunctional
impulsivity scores (Jackson & Smillie, 2004); however, such an
association is likely to have been biased by the classical impulsivity-related indicators included in the AMS. Unsurprisingly, the
presence of classical impulsivity indicators leads to sizeable associations of the AMS with impulsivity scales, as well as with the Fun
scale of the CW-BAS (Smillie & Jackson, 2005) that appears to better reflect dysfunctional impulsivity (Leone & Russo, 2009).
The body of evidence regarding the original 20-item form of the
AMS may be deemed mixed. The AMS was aligned with BAS criteria such as extraversion, the CW-BAS scales and functional impulsivity, but also with psychoticism and other dysfunctional
impulsivity measures. Cooper et al. (2008) attempted to revise
the scale to render it more internally coherent and more independent of the elusive concept of classical ‘‘dysfunctional’’ or ‘‘rash’’
impulsivity. Using a sizeable sample, Cooper et al. (2008) refined
the AMS, eliminating 9 out of the 20 original items. The remaining
11 items (Table 1) were assumed to be a reflection of a single onedimensional factor, and yielded satisfactory internal consistency
estimates. The items which were eliminated included most of the
typical impulsivity-related indicators, as well as those referring
to discomfort with adhering to the rules. In spite of the changes
in the scale’s composition, validity correlations were noticeably
similar for both the original and the revised scales (Cooper et al.,
2008, Table 6). The correlations of the AMS (both in its original
and revised form) with Impulsivity, Psychoticism and the Impulsivity-leaning CW-BAS Fun scale were only slightly weaker than
the correlations between the AMS and the ‘‘core’’ BAS measures
(Functional Impulsivity, CW-BAS Drive scale). It appears that a
one-dimensional interpretation of both the original and the revised
AMS scales still delivers a fuzzy set of relationships with core BAS
and impulsivity measures.
657
dysfunctional outcomes (Corr, 2008). We will show that such
components can be found in the AMS and that double dissociation patterns best represent the linkages between these two
components and the criteria representing the core BAS and
impulsivity, respectively. Our hypotheses could be more formally
summarized as follows:
Hypothesis 1. We anticipate that two dimensions reflecting the
core BAS (appetitive motivation and impulsivity-related variance)
would account more satisfactorily for variation among the AMS
items, compared with the original one-factor structure.
Hypothesis 2. Appetitive motivation will be related to extraversion (e.g., Jackson, 2009) and conscientiousness, as is expected of
a BAS measure reflecting sensitivity to appetitive stimuli and persistence and conscientiousness in pursuing reward contingencies
(Heym et al., 2008). The AMS–Impulsivity factor will be more
weakly related to extraversion and significantly associated with
classical impulsivity.
Hypothesis 3. Appetitive motivation will more strongly predict
CW-BAS Drive and Reward-Responsiveness scores than CW-BAS
Fun scores. We expect the opposite pattern to materialize for
AMS-Impulsivity.
Hypothesis 4. Appetitive motivation will be linked with functional
impulsivity, but not with dysfunctional impulsivity (e.g., Jackson,
2009; Smillie & Jackson, 2006). The reverse pattern is expected
for the AMS-Impulsivity factor.
Support for the hypotheses listed above will be interpreted as
implying the conceptual and empirical value of recognizing the
bidimensionality of the AMS. Results supportive of our hypotheses
would be consistent with r-RST definition of the BAS (e.g., Jackson,
2009; Smillie, Pickering et al., 2006) and with empirical and conceptual evidence on the structure and function of the BAS (e.g.,
Corr, 2008; Leone, 2009; Smillie, Jackson et al., 2006) obtained
using different measures (the AMS instead than the CW-BAS) and
in different cultural contexts.
3. Method
2. Aims and hypotheses
3.1. Participants and procedure
Although Cooper et al.’s (2008) revision enhanced the consistency and content validity of the AMS, the revised scale still did
not attain the desired pattern of strong associations with other
BAS measures and weak or null associations with rash impulsivity measures. We suspect that some conflation of impulsivity-related and BAS-related content still exists in the revised AMS. The
authors’ preference for a one-dimensional structure for the AMS
may have overshadowed the presence of a distinct impulsivityrelated component. Relaxing the one-factor-only assumption
may allow us to disentangle the impulsivity and core-BAS components that could be present in the AMS, and that have been
detected in other BAS measures (e.g., Smillie, Jackson et al.,
2006). We anticipate that the AMS may include two components: a dimension which is closely linked with conscientiousness in pursuing reward contingencies, persistence in working
towards future positive outcomes and a future-oriented focus
(e.g., Heym et al., 2008) regulated by the BAS, and a dimension
which clings to the impulsive behavioral components that in
some instances (proximity to the reward) may be considered
to serve BAS-motivated goal-pursuit, but that under different circumstances (when the reward is not yet in reach) may lead to
The study was conducted in Italy. Students at a large University
were asked to recruit participants from among their acquaintances.
A group of 388 individuals agreed to participate (males = 179; females = 207; two missing values). Participants varied considerably
in age (M = 35, SD = 13.35, range 18–76) and occupation. Importantly, psychology students were not included in the sample. The
participants anonymously filled in booklets including the personality measures described below. In order to avoid systematic sequence and fatigue effects, the order of the measures in the
booklets was pseudo-randomized across subjects.
3.2. Measures
An Italian version of the AMS (Jackson & Smillie, 2004) was
used. The scale includes 20 items and a five-point (disagree/agree)
response scale (Cooper et al., 2008). AMS items have been translated by the first author and then back-translated by a native English speaker. The original and back-translated versions closely
matched.
Impulsivity was measured using the Italian version (Russo,
Leone, & De Pascalis, 2011) of the I7 scale (Eysenck, Pearson,
658
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
Easting, & Allsopp, 1985), which is a widely-used 19-item (YES/NO
response) measure of trait impulsivity (a = .80).
The Big Five Questionnaire (short version, Caprara, Schwartz,
Capanna, Vecchione, & Barbaranelli, 2006) is a 60-item five-point
response scale of the Big Five traits. Due to the objective of this
study (see Hypothesis 2), only Energy (E, 12 items, a = .73) and
Conscientiousness (C, 12 items, a = .82) scores were used for the
analyses.
The CW-BAS (Carver & White, 1994; Italian version Leone,
Pierro, & Mannetti, 2002) includes 13 BAS-related items (5-point
disagree/agree scale). Three facets are measured: positive responses to the occurrence or anticipation of rewards or RewardResponsiveness (five items, a = .78); persistence in the pursuit of
reward or Drive (five items, a = .75); and a desire for novel rewards
and a tendency to act quickly in the pursuit of appetitive goals or
Fun (four items, a = .79).
An Italian version (Leone et al., 2002) of the Dickman Impulsivity Inventory (Dickman, 1990) was used. Twelve items measured
dysfunctional impulsivity (a = .83), and 11 items measured functional impulsivity (a = 71). Five-point disagree/agree scales were
used.
4. Results
Prior to performing a factor analysis of the AMS, a formal test
(Shapiro–Wilk test) of the normality of the distribution of the 20
items was performed. We found significant departures from normal distributions, as has already been reported (Cooper et al.,
2008). As a first step towards testing the tenability of a two-factor
structure for the AMS (Hypothesis 1), exploratory factor analysis
(EFA) was conducted. The full information maximum likelihood
method, assuming multivariate ordinal data, was used as method
of estimation (Jöreskog & Moustaki, 2006). We used Parallel analysis (PA; Horn, 1965) and Velicer’s MAP test (Velicer, 1976) to
determine the number of factors to retain (Velicer, Eaton, & Fava,
2000; Zwick & Velicer, 1986). The results of PA and Velicer’s MAP
test indicated that the number of factors ranged from a minimum
of two (Velicer’s MAP) to a maximum of seven (as suggested by
PA). Alternative solutions to the one-factor model put forward by
Cooper et al. (2008) do exist and can be further explored.
The first EFA model investigated includes all 20 AMS items. A
one-factor solution was inspected but several items loaded relatively poorly (<.35). A two-factor solution was then investigated.
Two interpretable factors emerged after Promax rotation. Items
3, 2 and 4 marked the first factor. These items reflect the subjects’
approach in terms of action orientation; the second extracted factor was marked by items 8, 10, 5 and 18, which reflect a mixture of
present orientation, fun-seeking and discomfort with external
rules. Finally, a three-factor solution was considered. In this solution, the first factor replicated its counterpart in the two-factor
solution, while the second and third factors were a split of the second factor in the previous solution. The first ‘‘slice’’ of the split was
fairly close to the ‘‘present orientation and sensation-seeking’’
(items 10, 11 and 18) interpretation suggested for the two-factor
solution, whilst the other ‘‘slice’’ (items 5, 8 and 1) reflected items
pertaining to adherence to the norm and a ‘‘personal way of doing
things’’. It is doubtful whether the latter component fits conceptually within the BAS construct. Therefore, the items marking this
component were eliminated from further analyses. Of the remaining items, those with unsatisfactory factor loadings (i.e., <.35) were
recursively eliminated from the analysis and different solutions
(i.e., the number of factors) were compared at each step. At the
end of this process, three items were eliminated, leading to a 14item set (Table 2) and a two-factor correlated solution (r = .39,
p < .01) that outperformed competing structures. Cronbach’s alpha
values for the ordered category items (Zumbo, Gadermann, & Zeisser, 2007) were satisfactory (Factor 1 = .83; Factor 2 = .79).
Confirmatory factor analyses (CFA) were then performed in order to provide a further test of the 14-item two-factor solution. CFA
models were estimated using a polychoric correlation matrix and
the robust maximum likelihood estimation function (Jöreskog,
2005). In order to obtain correct standard errors and to test the statistics under non-normality, the Satorra-Bentler chi-square statistic (1988) was used. Following Hu and Bentler’s (1999)
suggestions, values of root mean square error of approximation
(RMSEA) and of the standardized mean square residual
(SRMR) 6 .08 (better if 6.05) and comparative fit index (CFI) and
non-normed fit index (NNFI) values above P.95 are interpreted
as indicative of good fit. Indices for the two-factor solution were
then compared with alternative one-dimensional structures: a
20-item one-dimensional model (Jackson & Smillie, 2004), and
the revised 11-item one-dimensional model (Cooper et al., 2008).
The 20-item one-factor model fitted unsatisfactorily [v2(170) =
1272.61, p < .01; SB-v2(170) = 1155.74, p < .01; RMSEA = .093;
SRMR = .092; CFI = .90; NNFI = .89], replicating results from Cooper
et al. (2008), with all indices exceeding the acceptable cut-offs.
Even though the 11-item revised AMS showed a consistent
improvement in terms of CFI and NNFI (.94 and .92), the RMSEA
still indicated an unacceptable fit (.10) [v2(44) = 417.74, p < .01;
SB-v2(44) = 220.41, p < .01; SRMR = .074]. Instead, the two-factor
model comprising 14 items presented a very encouraging fit, with
all indices within acceptable thresholds [v2(76) = 486.10, p < .01;
SB-v2(76) = 258.44, p < .01; RMSEA = .079; SRMR = .077; CFI = .95;
NNFI = .95].
It appears that the results of both EFA and CFA converge in suggesting the presence of two related yet distinct facets of appetitive
motivation on the AMS. The first facet (items 3, 2, 4, 20, 14, 19, 15
and 12, see Table 2) refers to a goal-committed appetitive motivation, possibly regulating complex approach behaviors and entailing
some degree of future oriented focus, planning and perseverance.
Tentatively, we prefer to refer to this factor using the original label
‘‘Appetitive Motivation’’. Items on the second facet (11, 10, 6, 13,
18 and 16) refer instead to behavioral tendencies targeting the
immediate enjoyment of sensations and emotions without much
planning; tentatively, we have simply labeled this factor ‘‘Impulsivity’’ or ‘‘Impulsive hedonism’’.2
Hypotheses 2-4 concern various issues of convergent and discriminant validity. We adopted a structural equation approach to
test for the convergence, divergence and discriminant hypotheses
on the associations between approach motivation and general
and more specific personality measures. In the three subsequent
models the two AMS factors were defined by their respective items
(8 and 6 items, respectively), whilst the criteria were defined by a
single indicator each (average scale-score). In the first model we
regressed extraversion, conscientiousness and impulsivity (I7) on
the two AMS factors. Impulsivity is perhaps misleadingly (e.g.,
Depue & Collins, 1999; Smillie & Jackson, 2006) connected with
BAS sensitivity; BAS measures should be aligned with extraversion
instead (e.g., Jackson, 2009); therefore, we expected stronger linkages between appetitive motivation and extraversion, compared
with the association between extraversion and impulsive hedonism. The r-RST conceptualization of the BAS point instead to a
2
We are aware that different labels could have been proposed. For instance, it
could be argued that the Appetitive Motivation factor also reflects sensation-seeking
plus action orientation and/or quick refocusing of attention (Dickman, 1990), plus
possibly future oriented motivations (Corr, 2008; Heym et al., 2008). Therefore, a
different label may have been chosen. Similarly, the ‘‘Impulsivity’’ factor could have
been more narrowly labeled in order to refer to present orientation and a lack of
planning. Nonetheless, these alternative interpretations are still consistent with
reward sensitivity and therefore with the core BAS on the one hand, and to
impulsivity-related manifestations on the other.
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
Table 2
Factor loadings (after Promax rotation) of the two factors solution.
2. Put plans into action
3. Like to be busy
4. Like to see how things work
12. Feel for how things work
14. Excited by what is new in my field
15. Often lots of ideas
19. New ideas all the time
20. Enjoy new projects
6. Like things new and different
10. Important to enjoy moment
11. Actively look for new experiences
13. Look for new sensations
16. Prefer not to plan too much
18. Here and now is important
Factor 1
Factor 2
Uniqueness
0.689
0.803
0.632
0.423
0.536
0.473
0.496
0.584
0.325
0.009
0.212
0.220
0.174
0.137
0.188
0.026
0.053
0.185
0.310
0.281
0.314
0.327
0.598
0.600
0.674
0.569
0.534
0.563
0.591
0.371
0.624
0.726
0.487
0.593
0.534
0.403
0.386
0.636
0.389
0.530
0.757
0.724
Note: factor loadings exceeding 0.35 are presented in boldface.
looser association with impulsivity. Therefore, appetitive motivation should be more weakly linked with impulsivity, whilst the
Impulsive component of the AMS was expected to be closely related to classic impulsivity. We partialled out impulsivity effects
on extraversion and conscientiousness in order to avoid spurious
associations between the AMS factors and extraversion and conscientiousness due to their association with impulsivity. Considering
the overall information stemming from the different fit indices, the
model fitted satisfactorily: v2(112) = 365.52, p < .001,
RMSEA = .076 (90% CI for RMSEA: .068–.085), NNFI = .92,
CFI = .93. As expected, the two AMS factors show a complex dissociation pattern with the criteria. Appetitive motivation was posip < .01)
and
tively
related
to
extraversion
(c = .443,
conscientiousness (c = .44, p < .01) and negatively to impulsivity
(c = .34, p < .01). Instead, the impulsive hedonism facet was positively related to impulsivity (c = .52, p < .01) and to a lesser extent
to extraversion (c = .20, p < .05), and was non-significantly associated with conscientiousness (c = .02, ns). Clearly, the two facets relate very differently to extraversion, conscientiousness and
impulsivity. In addition, it appears that only the Appetitive Motivation facet behaved as a core BAS measure as defined within r-RST,
whilst the impulsive hedonism facet was aligned with impulsivity,
and therefore it is doubtful whether it should be regarded as an indicator of core BAS sensitivity (e.g., Smillie & Jackson, 2006).
In order to gain further information regarding which AMS component best reflects core BAS sensitivity, we regressed in a Structural Equation Model the three CW-BAS components on the two
AMS factors. Hypothesis 3 anticipated a closer association between
the Approach Motivation facet of AMS and Drive and RewardResponsiveness, the two CW-BAS scales which most closely
approximate BAS sensitivity (Leone, 2009; Smillie, Jackson et al.,
2006). The fit can be considered acceptable: v2 (112) = 351.18,
p < .001, RMSEA = .074 (90% CI for RMSEA: .066–.083), NNFI = .93,
CFI = .94. As anticipated, a clear dissociation pattern was found:
the Appetitive Motivation factor was connected to Drive (c = .37,
p < .01) and Reward-Responsiveness (c = .33, p < .01), but not to
Fun (c = .08, ns), whilst the Impulsive facet of the AMS was closely aligned with Fun (c = .63, p < . 01), but not connected to Drive
(c = .11, ns) or Reward-Responsiveness (c = .07, ns) (the CW-BAS
facets considered to be markers of the core BAS; Smillie, Jackson
et al., 2006).
Previous evidence (Leone & Russo, 2009) has shown that the
CW-BAS facets of Drive and Fun are related to functional and dysfunctional impulsivity in different ways, suggesting that only drive
is aligned with the functional impulsive functions supervised by
3
Completely standardized parameters are reported.
659
the BAS (Smillie & Jackson, 2006). In an analogous fashion, Hypothesis 4 hypothesized that only the Appetitive Motivation factor of
the AMS would lead to functional impulsivity (Jackson, 2009),
whilst the Impulsive facet of the AMS should instead be uniquely
associated with dysfunctional impulsivity. In order to control for
the association between the two forms of impulsivity, dysfunctional effects on functional impulsivity were partialled out. When
functional and dysfunctional impulsivity were regressed on the
two AMS factors, a barely acceptable fit was found: v2
(100) = 343.85, p < .001, RMSEA = .079 (90% CI for RMSEA: .070–
.089), NNFI = .91, CFI = .93. As anticipated, appetitive motivation
was suitably aligned with functional impulsivity (c = .31, p < .01)
and negatively connected with dysfunctional impulsivity
(c = .42, p < .01); such divergent pattern appears consistent with
an ideal BAS measure (Jackson, 2009; Leone & Russo, 2009). In turn,
the Impulsivity items of the AMS were positively related to dysfunctional impulsivity (c = .44, p < .01), and unrelated with functional impulsivity (c = .16, ns). In accordance with results from
the previous models, this pattern suggests that only the subset of
items of the AMS loading on the Appetitive Motivation facet represents the core BAS.
5. Discussion
We reasoned that the AMS could be better modeled by specifying two factors, one related to core BAS functioning and the other
to impulsivity. Such a model appears to outperform the onedimensional alternatives. We anticipated that the Appetitive Motivation facet would show a pattern of associations which were
highly consistent with the empirical definitions of core BAS sensitivity (i.e., linkages with the Extraversion, Drive and RewardResponsiveness facets of the CW-BAS, and with functional impulsivity). Instead, the Impulsive facet of the AMS was expected to relate to classical and dysfunctional impulsivity and with BAS
indicators known to share significant variance with impulsivity.
These hypotheses were empirically supported. Furthermore, the
impulsive cluster within the AMS was weakly related with extraversion, compared with its Appetitive Motivation counterpart. In
brief, as is the case for the CW-BAS, the AMS represents a complex
operationalization of BAS sensitivity. Nevertheless, from our perspective, such complexity is not necessarily a pitfall of the measure; instead, it can be interpreted as a useful feature of the
AMS, as it allows researchers to test different hypotheses that take
into account different BAS-related processes which occur in different stages of reward pursuit (Corr, 2008; Heym et al., 2008). We
believe that BAS-related scales often fail to show simple onedimensional structures because the complex construct they intend
to target is difficult to fit into a one-dimensional Procrustean bed,
and not because of banal psychometric misspecifications. We will
discuss these issues in greater depth below.
5.1. The complex structure of BAS measures: Is it really so bad?
Cooper et al.’s (2008) revision of the AMS acknowledged that
empirical results may cast some doubts on the one-dimensionality
of the scale. Thus, an alternative strategy which allows for multidimensionality is feasible, as our results suggest, and – arguably –
provides a better account of the data. The two factors corresponding to ‘‘core’’ appetitive motivation and to impulsivity depart significantly from a one-dimensional perspective. However, we are
not suggesting that there is something wrong with the AMS; instead, we are convinced that researchers can find in the AMS a sensitive and efficient tool with which to disentangle impulsivityrelated variance in BAS criteria from core appetitive motivation
variance. As has been shown for the CW-BAS scales (Smillie,
660
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
Jackson et al., 2006), it is possible to detect within the AMS a reflection of the reward-reactivity versus trait impulsivity distinction
(e.g., Quilty & Oakman, 2004). It would be interesting to ascertain
whether an appetitive motivation score devoid of almost all conflation with impulsivity is able to predict the behavioral criteria
which have been investigated in previous research relying on the
AMS full score (Smillie & Jackson, 2005; Smillie & Jackson, 2006).
Our results suggest that the 20-item score is conflated with variance relating to impulsivity, and lacks specificity. It could thus be
argued that previous results underestimate trait effects on behavioral criteria. The leaner and more specific factor we propose as
indexing appetitive motivation may reveal stronger associations.
The impulsivity-related cluster of AMS items may turn out to be
useful as well. First, AMS impulsivity may act as an efficient statistical control for any expected effect governed by appetitive motivation. Second, as some behavioral criteria may stem from a mixture
of SR and impulsive action orientation, the two factors of the AMS
may together provide a richer account of complex behaviors. Finally, the Impulsive facet may also reveal how, under some circumstances in need of theoretical definition, behaviors, which are
usually governed by the BAS, may turn out to be dysfunctionally
influenced by impulsivity. Such interactions between behaviors,
situational or personal contingencies and traits (appetitive motivation vs. impulsivity) may help to improve models of maladaptive
behaviors such as addictions, risk-taking and delinquency which
have been rather simplistically traced back to the BAS or to
impulsivity.
5.2. Implications for the conceptualization of the BAS and impulsivity
Over the past few years, the r-RST and most empirical research
have converged on the notion that extraversion (and not impulsivity or psychoticism) can be considered as the appropriate BAS at
the surface personality level (e.g., Depue & Collins, 1999; Smillie,
Pickering et al., 2006; Jackson, 2009). Nevertheless, empirical evidence has consistently shown that most purpose-built BAS measures include some impulsivity variance. For instance, the Fun
scale of the CW-BAS is more closely aligned to impulsivity than
to the core BAS (Leone, 2009; Smillie, Jackson et al., 2006). In a
remarkably similar vein, the AMS can be modeled with two components which are aligned to the core BAS and to impulsivity. Different degrees of theoretical and empirical misspecifications of
these purpose-built BAS measures could account for such pervasive conflation. A first source of confusion traces back to earlier
versions of the RST that explicitly connected BAS and impulsivity.
Nevertheless, it appears that theoretical reasons may account for
the BAS–impulsivity conflation even within the revised RST framework. Even though impulsivity and the BAS may correspond to
independent physiological substrata, their functional outcomes
may be related because of some common mediator (Smillie, Jackson et al., 2006). Another possibility is that the BAS and impulsivity
manifestations have become associated because their manifestations stem from the joint effects of several systems regulating activation and inhibition (Corr, 2002). As the behavioral outcomes of
the BAS are affected by the activity in other systems which are related to conflict detection, inhibition and flight or fight, it may be
possible that the outcomes share variance with impulsivity measures and indices of impulsive tendencies via indirect paths of
association (Smillie, Jackson et al., 2006).
A different account of BAS-impulsivity associations has conceptualized the BAS as encompassing different functions during goal
pursuit: a behavior maintenance function and a consumatory function (Carver & Miller, 2006; Corr, 2008). Impulsivity would be part
of the consumatory function, and would kick in once the likelihood
of attaining the reward became favorable and the incentive was
within reach. In such instances, unleashing impulsive behavior in
order to attain the goal in question may be considered to be adaptive and consistent with a definition of BAS as appetitive motivation and functional impulsivity (Smillie & Jackson, 2006). A
similar account has been proposed in order to tentatively describe
the different functions of the three CW-BAS scales during different
phases of complex incentive-pursuit: Drive and Reward-Responsiveness provide the behavioral maintenance motivation needed
during complex approach behavior involving multiple sub-goals,
whilst the Fun-Seeking component kicks in when the incentive
needs to be seized swiftly (Corr, 2008; Leone, 2009). Our pattern
of results is consistent with the interpretation of the Appetitive
Motivation facet that views the AMS as tapping into the behavioral
maintenance function of BAS. Support for this notion relies on the
negative association between appetitive motivation and impulsivity and its positive linkage with conscientiousness; this pattern is
compounded by the significant association between the appetitive
motivation facet and the CW-BAS Drive and Reward-Responsiveness scales, and its lack of an association with the CW-BAS Fun
scale. Further support is provided by the unique effect of appetitive
motivation on functional impulsivity. In turn, the impulsive facet
within the AMS parallels fun-seeking in the CW-BAS, and it is related to both classical and dysfunctional impulsivity. Applying
Corr’s (2008) reasoning to the AMS, the impulsive component
may represent a theoretically significant part of BAS activity when
the incentive needs to be seized. Instead, when the maintenance
function of BAS is weakened, premature unleashing of the impulsive function may eventually disrupt goal attainment. Of course,
such time-lagged and interactive interplay of the different BAS
components during different phases of goal pursuit needs to be
tested via experimental research, by manipulating the attainability
and proximity of the incentive, as well as the complexity of the
sub-goal structure leading to the final reward.
If the BAS-impulsivity conflation does indeed stem from complex interactions at the physiological and behavioral levels, it
should not be surprising that it is difficult to construct BAS measures which are devoid of impulsivity-related content and which
are orthogonal to impulsivity measures. Responses on personality
questionnaires may be considered as one instance of outcomes of
the BAS (Smillie, Jackson et al., 2006), which often implies a degree
of overlap with impulsivity-related outcomes (Smillie & Jackson,
2005). One-to-one correspondence between pure BAS measures
and BAS sensitivity on the one side, and pure impulsivity measures
and the basic process governing impulsivity on the other, appears
to be a difficult standard to attain, and no single scale developed
under the original RST framework satisfactorily respond to these
requirements. Measures explicitly developed under the assumptions of the r-RST may do a better job, as suggested by recent evidence (Jackson, 2009). Nevertheless, it appears to be possible to
devise within ‘‘classical’’ RST measures as the AMS and the Carver
and White scales relatively clean factors satisfactorily matching
the measurement requirements of the r-RST (Heym et al., 2008).
In the case of classical BAS measures, it appears feasible to identify
components more closely related to the behavioral maintenance
function of the BAS and components aligned with the consumatory
component of the BAS (Corr, 2008). Such distinct factors may be
used to test discriminant, dissociation, mediation or interaction
hypotheses that cannot be tested when the BAS and impulsivity
components are collapsed into a single score.
Some limitations should also be outlined. Our results rely on a
single sample, and replications are warranted. However, the sample is not small by the average standards of personality research,
and – unlike the typical samples in personality research – does
not include psychology students. In addition, the AMS items are
all directed ‘‘pro-trait.’’ This method factor may play a pervasive
role in all the analyses; of course, this method bias is common
amongst any previous research studies which have also used the
L. Leone et al. / Journal of Research in Personality 45 (2011) 655–661
AMS. Common method bias artificially inflates the strength of onefactor structures, and makes it more difficult to devise alternative
solutions. In spite of any effects of method bias, our two factors
were clearly preferable over the one-factor solutions and provided
a clearer pattern of associations with a wide array of criteria. Finally, our results may be, to some extent, dependent on the specific
characteristics of our Italian sample of participants; however, this
could be interpreted as a positive feature, because our results replicate and generalize previous findings using different measures
and different cultural contexts (e.g., Leone & Russo, 2009; Smillie,
Jackson et al., 2006; Smillie & Jackson, 2006). It appears unlikely
that such conceptual consistency would result from any specific
bias occurring in our sample because of methodological artifacts
or culture-specific variance. More likely, consistency of results
across samples, cultures and methods points to robust and generalizable effects.
References
Caprara, G. V., Schwartz, S., Capanna, C., Vecchione, M., & Barbaranelli, C. (2006).
Personality and politics: Values, traits and political choice. Political Psychology,
27, 1–28.
Carver, C. S. (2004). Negative affects deriving from the behavioral approach system.
Emotion, 4, 3–22.
Carver, C. S., & Miller, C. J. (2006). Relations of serotonin function to personality:
Current views and a key methodological issue. Psychiatry Research, 144, 1–15.
Carver, C. S., & 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.
Caseras, X., Avila, C., & Torrubia, R. (2003). The measurement of individual
differences in behavioral inhibition and behavioral activation systems: A
comparison of personality scales. Personality and Individual Differences, 34,
999–1013.
Cooper, A. J., Smillie, L. D., & Jackson, C. J. (2008). A trait conceptualization of
Reward-Reactivity. Psychometric properties of the Appetitive Motivation Scale
(AMS). Journal of Individual Differences, 29, 168–180.
Corr, P. J. (2002). J.A. Gray’s Reinforcement Sensitivity Theory: Tests of the joint
subsystems hypothesis of anxiety and impulsivity. Personality and Individual
Differences, 33, 511–532.
Corr, P. J. (2008). Reinforcement sensitivity theory (RST): Introduction. In P. J. Corr
(Ed.), The reinforcement sensitivity theory of personality (pp. 1–43). Cambridge:
Cambridge University Press.
Depue, R. A., & Collins, P. F. (1999). Neurobiology of the structure of personality:
Dopamine, facilitation of incentive motivation, and extraversion. Behavioural
and Brain Sciences, 22, 491–569.
Dickman, S. J. (1990). Functional and dysfunctional impulsivity: Personality and
cognitive correlates. Journal of Personality and Social Psychology, 58, 95–102.
Eysenck, S. B. G. (1993). The I7: Development of a measure of impulsivity and its
relationships to the superfactors of personality. In W. G. McCown, J. L. Johnson,
& M. B. Shure (Eds.), The impulsive client: Theory research and treatment
(pp. 141–149). Washington, DC: American Psychological Association.
Eysenck, S. B. G., Pearson, P. R., Easting, G., & Allsopp, F. J. (1985). Age norms for
impulsiveness, venturesomeness and empathy in adults. Personality and
Individual Differences, 6, 613–619.
Gray, J. A. (1982). The neuropsychology of anxiety: An enquiry into the functions of the
septo-hippocampal system. Oxford: Oxford University Press.
Gray, J. A., & McNaughton, N. (2000). The neuropsychology of anxiety: An inquiry into
the functions of the septo-hippocampal system (2nd ed.). Oxford: Oxford
University Press.
Heym, N., Ferguson, E., & Lawrence, C. (2008). An evaluation of the relationship
between Gray’s revised RST and Eysenck’s PEN: Distinguishing BIS and FFFS in
Carver and White’s BIS/BAS scales. Personality and Individual Differences, 45,
709–715.
Heym, N., & Lawrence, C. (2010). The role of Gray’s revised RST in the Ppsychopathy continuum: The relationships of psychoticism with a lack of fear
and anxiety, and increased impulsivity. Journal of Personality and Individual
Differences, 49, 874–879.
Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis.
Psychometrika, 32, 179–185.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation
Modelling, 6, 1–55.
Jackson, C. J. (2003). Gray’s Reinforcement Sensitivity Theory: a psychometric
critique. Personality and Individual Differences, 34, 533–544.
Jackson, C. J. (2009). Jackson-5 scales of the revised Reinforcement Sensitivity
Theory (r-RST) and their application to dysfunctional real world outcomes.
Journal of Research in Personality, 43, 556–569.
661
Jackson, C. J., & Smillie, L. D. (2004). Appetitive motivation predicts the majority of
personality and an ability measure: A comparison of BAS measures and a
reevaluation of the importance of RST. Personality and Individual Differences, 36,
1627–1636.
Jöreskog, K. (2005). Structural equation modelling with ordinal variables using LISREL.
Chicago: Scientific Software.
Jöreskog, K., & Moustaki, I., (2006) Factor analysis of ordinal variables with full
information maximum likelihood. <http://www.ssicentral.com/lisrel/techdocs/
orfiml.pdf>.
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.
Leone, L. (2009). Testing conceptual distinctions among Carver and White’s (1994)
BAS scales: A comment and extension on Smillie, Jackson, and Dalgleish (2006).
Personality and Individual Differences, 46, 54–59.
Leone, L., Perugini, M., Bagozzi, R. P., Pierro, A., & Mannetti, L. (2001). Construct
validity and generalizability of the carver white behavioural inhibition system/
behavioural activation system scales. European Journal of Personality, 15,
373–390.
Leone, L., Pierro, A., & Mannetti, L. (2002). Validity of the Italian version of the BIS/
BAS Scale of Carver and White (1994): Generalizability of structure and
relationships with related constructs. Giornale Italiano di Psicologia, 29,
413–434.
Leone, L., & Russo, P. M. (2009). Components of the behavioral activation system
and functional impulsivity: A test of discriminant hypotheses. Personality and
Individual Differences, 43, 1101–1104.
Newman, J. P., MacCoon, D. G., Vaughn, L. J., & Sadeh, N. (2005). Validating a
distinction between primary and secondary psychopathy with measures of
Gray’s BIS and BAS constructs. Journal of Abnormal Psychology, 114, 319–323.
O’Connor, R. M., Stewart, S. H., & Watt, M. C. (2009). Distinguishing BAS risk for
university students’ drinking, smoking, and gambling behaviors. Personality and
Individual Differences, 46, 514–519.
Pickering, A. D., & Gray, J. A. (1999). The neuroscience of personality. In L. Pervin &
O. John (Eds.), Handbook of personality (2nd ed., pp. 277–299). New York:
Guilford.
Pickering, A. D., & Smillie, L. D. (2008). The behavioural activation system:
Challenges and opportunities. In P. J. Corr (Ed.), The reinforcement sensitivity
theory of personality (pp. 120–153). Cambridge: Cambridge University Press.
Quilty, L. C., & Oakman, J. M. (2004). The assessment of behavioural activation the
relationship between impulsivity and behavioural activation. Personality and
Individual Differences, 37, 429–442.
Russo, P. M., Leone, L., & De Pascalis, V. (2011). Cross-cultural validity of the I7
impulsiveness – Venturosomeness – Empathy scales: Evidence from the Italian,
I7. Comprehensive Psychiatry, 52, 446–452.
Satorra, A., & Bentler, P. M. (1988). Scaling corrections for chi-square statistics in
covariance structure analysis. Proceedings of the Business and Economic Statistics
Section of the American Statistical Association, 308, 313.
Smillie, L. D., & Jackson, C. J. (2005). The Appetitive Motivation Scale and other BAS
measures in the prediction of approach and active avoidance. Personality and
Individual Differences, 38, 981–994.
Smillie, L., & Jackson, C. J. (2006). Functional impulsivity and reinforcement
sensitivity theory. Journal of Personality, 74, 47–83.
Smillie, L. D., Jackson, C. J., & Dalgleish, L. I. (2006). Conceptual distinctions among
Carver and White’s (1994) BAS scales: A reward reactivity versus trait
impulsivity perspective. Personality and Individual Differences, 40, 1039–1050.
Smillie, L. D., Pickering, A. D., & Jackson, C. J. (2006). The new reinforcement
sensitivity theory: Implications for personality measurement. Personality and
Social Psychology Review, 10, 320–335.
Urosevic, S., Abramson, L. Y., Harmon-Jones, E., & Alloy, L. B. (2008). Dysregulation of
the behavioral approach system (BAS) in bipolar spectrum disorders: Review of
theory and evidence. Clinical Psychology Review, 28, 1188–1205.
Velicer, W. F. (1976). Determining the number of factors from the matrix of partial
correlations. Psychometrika, 41, 321–327.
Velicer, W. F., Eaton, C. A., & Fava, J. L. (2000). Construct explication through factor
or component analysis: A review and evaluation of alternative procedures for
determining the number of factors or components. In R. D. Goffin & E. Helmes
(Eds.), Problems and solutions in human assessment: Honoring Douglas N. Jackson
at seventy (pp. 41–72). Norwell, MA: Kluwer Academic.
Voigt, D. C., Dillard, J. P., Braddock, K. H., Anderson, J. W., Sopory, P., & Stephenson,
M. T. (2009). BIS/BAS scales and their relationship to risky health behaviours.
Personality and Individual Differences, 47, 89–93.
Wilson, G. D., Gray, J. A., & Barrett, P. T. (1990). A factor analysis of the Gray–Wilson
personality questionnaire. Personality and Individual Differences, 11, 1037–1045.
Zisserson, R. N., & Palfai, T. P. (2007). Behavioral activation system (BAS) sensitivity
and reactivity to alcohol cues among hazardous drinkers. Addictive Behaviors,
32, 2178–2186.
Zumbo, B. D., Gadermann, A. M., & Zeisser, C. (2007). Ordinal versions of coefficients
alpha and theta for likert rating scale. Journal of Modern Applied Statistical
Methods, 6, 21–26.
Zwick, W. R., & Velicer, W. F. (1986). Factor influencing five rules for determining
the number of components to retain. Psychological Bulletin, 99, 432–442.