Content uploaded by Sherry H Stewart
Author content
All content in this area was uploaded by Sherry H Stewart on Oct 08, 2015
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rigs20
Download by: [Dalhousie University] Date: 08 October 2015, At: 10:07
International Gambling Studies
ISSN: 1445-9795 (Print) 1479-4276 (Online) Journal homepage: http://www.tandfonline.com/loi/rigs20
Predicting gambling behaviour and problems from
implicit and explicit positive gambling outcome
expectancies in regular gamblers
Melissa J. Stewart, Sherry H. Stewart, Sunghwan Yi & Michael Ellery
To cite this article: Melissa J. Stewart, Sherry H. Stewart, Sunghwan Yi & Michael Ellery (2015)
Predicting gambling behaviour and problems from implicit and explicit positive gambling
outcome expectancies in regular gamblers, International Gambling Studies, 15:1, 124-140, DOI:
10.1080/14459795.2014.1000357
To link to this article: http://dx.doi.org/10.1080/14459795.2014.1000357
Published online: 09 Feb 2015.
Submit your article to this journal
Article views: 100
View related articles
View Crossmark data
Predicting gambling behaviour and problems from implicit and
explicit positive gambling outcome expectancies in regular gamblers
Melissa J. Stewart
a
*, Sherry H. Stewart
a,b,c
, Sunghwan Yi
d
and Michael Ellery
e
a
Department of Psychology & Neuroscience, Dalhousie University, Halifax, Canada;
b
Department
of Psychiatry, Dalhousie University, Halifax, Canada;
c
Department of Community Health and
Epidemiology, Dalhousie University, Halifax, Canada;
d
Department of Marketing & Consumer
Studies, University of Guelph, Guelph, Canada;
e
Addictions Foundation of Manitoba, Winnipeg,
Canada
(Received 27 August 2014; accepted 15 December 2014)
Outcome expectancies are the positive or negative effects that individuals anticipate
may occur from engaging in a given behaviour. Although explicit outcome
expectancies have been found to play an important role in gambling, research has
yet to assess the role of implicit outcome expectancies in gambling. In two studies, we
investigated whether implicit and explicit positive gambling outcome expectancies
were independent predictors of gambling behaviour (i.e. amount of time spent and
money risked gambling; Study 1) and problem gambling severity (Study 2). In both
studies, implicit positive gambling outcome expectancies were assessed by having
regular gamblers (N¼58 in Study 1; N¼96 in Study 2) complete a gambling outcome
expectancy reaction time (RT) task. A self-report measure of positive gambling
outcome expectancies was used to assess participants’ explicit positive gambling
outcome expectancies. Both the RT task and self-report measure of positive
gambling outcome expectancies significantly contributed unique as well as shared
variance in the prediction of self-reported gambling behaviour (Study 1) and problem
gambling severity (Study 2). Findings from the current research point to the importance
of using both direct and indirect assessment modes when examining the role of
outcome expectancies in gambling.
Keywords: gambling; addictive behaviour; problem gambling; cognition; addiction
Outcome expectancies involve the perceived positive or negative effects that individuals
anticipate may occur from engaging in a given behaviour. Over the past 25 years, an
abundance of research has highlighted the role of outcome expectancies in addictive
behaviours, such as alcohol use (Goldman, Darkes, & Del Boca, 1999; Sayette, 1999).
Indeed, alcohol outcome expectancies have been found to be associated with increased
alcohol consumption and alcohol-related problems (Goldman et al., 1999). In light of the
theoretical significance of outcome expectancies in the alcohol field, along with the
similarities between alcohol use and gambling as addictive behaviours (American
Psychiatric Association, 2013; Potenza, 2006), it is natural to postulate that outcome
expectancies may also be relevant in the gambling domain. However, researchers have
only recently turned their attention towards investigating the role of outcome expectancies
in gambling. Similar to research on alcohol outcome expectancies, the emerging gambling
literature has found that gambling outcome expectancies are associated with increased
levels of gambling behaviour and problems (e.g. Gillespie, Derevensky, & Gupta, 2007b;
q2015 Taylor & Francis
*Corresponding author. Email: stewart.melissa@dal.ca
International Gambling Studies, 2015
Vol. 15, No. 1, 124–140, http://dx.doi.org/10.1080/14459795.2014.1000357
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Shead & Hodgins, 2009; St-Pierre, Temcheff, Gupta, Derevensky, & Paskus, 2014;
Wickwire, Whelan, & Meyers, 2010).
Although previous research on gambling outcome expectancies has been helpful in
elucidating the role of outcome expectancies in gambling, such research has primarily
relied on direct modes of assessment (e.g. self-report questionnaires). ‘Direct’ or ‘explicit’
measures refer to a class of measurement procedures that tap into cognitions and
behaviours thought to be deliberate and controlled, and those that involve conscious
engagement (De Houwer, 2006). Despite significant research advances made using direct
modes of assessment, as well as the importance of assessing outcome expectancies via
direct measures, its limitations have been increasingly recognized in the alcohol outcome
expectancy literature (e.g. Palfai & Ostafin, 2003). For example, it is unlikely that a
specific episode of drinking is the result of a deliberate and conscious consideration of the
expected outcomes of drinking – the construct reflected in self-report measures of alcohol
outcome expectancies.
Influenced by cognitive psychology, addiction researchers have increasingly adopted
the view that outcome expectancies are represented in the associative memory network
(Goldman et al., 1999; Stacy, 1997). In accordance with this view, gambling outcome
expectancies are operationalized as the speed with which the concept of gambling (or
exposure to gambling cues) activates a given outcome expectancy in an individual’s
memory. For example, individuals who have a very strong positive outcome expectancy of
gambling should experience an increased activation of the positive outcome expectancy
when exposed to gambling to gambling activities or gambling cues compared to with weak
positive gambling outcome expectancies. In order to assess individual differences in the
strength of outcome expectancies, addiction researchers have employed ‘indirect’ or
‘implicit’ measures, such as reaction time (RT) tasks, in conjunction with direct measures
(Kramer & Goldman, 2003; Wiers et al., 2002). In contrast to direct measures, indirect
measures are thought to assess attitudes and cognitions in an automatic manner, are said to
be unconscious and involuntary in nature and appear to influence individuals’ memory
without explicit recall or introspection (see De Houwer, 2006).
Consistent with the reflective-impulsive model of social behaviour (Strack & Deutsch,
2004), both direct (e.g. self-report) and indirect (e.g. RT tasks) measures can be considered
complementary in assessing outcome expectancies in that self-report measures assess
deliberative determinants of behaviour, while RT measures are said to assess automatic
determinants of behaviour (see Wiers & Stacy, 2006). As direct and indirect measures
have been found to tap into different facets of outcome expectancies in the alcohol field
(e.g. de Jong, Wiers, van de Braak, & Huijding, 2007; Kramer & Goldman, 2003), and
provide unique contributions to the prediction of alcohol use behaviours in the alcohol
outcome expectancy area (Wiers et al., 2002), it may be similarly important to utilize both
assessment modes when examining the role of outcome expectancies in gambling.
However, to date, research has yet to assess the utility of direct and indirect measures of
gambling outcome expectancies in independently predicting gambling behaviour and
associated problems.
In order to facilitate research in this area, Stewart and colleagues recently conducted a
set of studies (Stewart, Yi, Ellery, & Stewart, under review; Stewart, Yi, & Stewart, 2014)
that examined the impact of gambling cue exposure on the activation of implicit and
explicit gambling outcome expectancies. Consistent with previous research in the alcohol
field (e.g. Palfai & Ostafin, 2003; Wall, Hinson, McKee, & Goldstein, 2001), results of
these studies revealed an activation of positive but not negative gambling outcome
expectancies following exposure to gambling cues. Similar to research in the alcohol field
International Gambling Studies 125
Downloaded by [Dalhousie University] at 10:07 08 October 2015
(e.g. Jones, Corbin, & Fromme, 2001), these results suggest that negative outcome
expectancies reflect outcomes less proximal to gambling than positive outcome
expectancies. In light of such findings, it appears important to assess whether implicit
and explicit positive gambling outcome expectancies are capable of predicting gambling
behaviours and gambling problems, as has been previously shown in the alcohol outcome
expectancy area (e.g. Kramer & Goldman, 2003; Wiers et al., 2002).
Therefore, the purpose of the current set of studies was to investigate whether implicit
and explicit positive gambling outcome expectancies independently predicted two
important indices of gambling behaviour (i.e. amount of time spent and money risked
gambling; Study 1), as well as problem gambling severity (Study 2). Specifically, the
current research assessed the incremental contributions of direct (i.e. explicit) and indirect
(i.e. implicit) measures of positive gambling outcome expectancies in the prediction of
self-reported gambling behaviour and problem gambling severity. Overall, it was
predicted that the direct and indirect measures of positive gambling outcome expectancies
would predict unique as well as shared variance in the amount of time spent and money
risked gambling (Study 1), and problem gambling severity (Study 2).
Study 1
Method
Participants
Participants (N¼58; 38 males and 20 females) for this investigation consisted of
gamblers who were part of a larger study investigating the effect of gambling cue exposure
on implicit and explicit gambling outcome expectancies (Stewart et al., 2014). Participants
ranged in age from 19 to 61 years (M¼29.97, SD ¼12.04). Thirty-six participants were
recruited from the Halifax Regional Municipality in Nova Scotia, while the remaining 22
participants were recruited from the greater Guelph area in Ontario. No significant
differences were found between the two sites on any key demographic or outcome
variables.
To qualify for participation, individuals had to meet the following inclusion criteria:
(a) 19 years of age or older, (b) gambled at a casino or online at least three times over the
past two months, and (c) reported English as their native language (given that RT measures
require extremely rapid responses to English words). Individuals were excluded if they
were currently attempting to quit gambling or engaging in treatment for problem
gambling.
Using the Problem Gambling Severity Index (PGSI) from the Canadian Problem
Gambling Index (CPGI; Ferris & Wynne, 2001), participants consisted of 3 non-problem
gamblers, 8 low-risk gamblers, 32 moderate-risk gamblers and 15 high-risk/problem gamblers.
Total scores on the PGSI ranged from 0 to 25 (M¼6.26; SD ¼5.20). In terms of
gambling behaviour (as measured by the Gambling Timeline Followback [G-TLFB;
Weinstock, Whelan, & Meyers, 2004]), participants reported spending between 1.92 and
204.00 hours (M¼35.84, SD ¼39.07) gambling over the three months prior to
participating in the study. The amount of money participants risked over the three months
prior to completing the study ranged from $42 to $3690 (M¼$991.41, SD ¼$916.09).
Measures
Problem gambling symptoms. The nine-item PGSI scale of the CPGI (Ferris & Wynne,
2001) was used to assess the presence and severity of gambling problems among
126 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
participants for sample description purposes. The PGSI contains five items that assess
problem gambling behaviour (e.g. ‘Have you bet more than you could really afford to
lose?’) and four items addressing the negative consequences of gambling (e.g. ‘Has
gambling caused you any health problems, including stress or anxiety?’). For each item,
respondents indicated the frequency with which they had engaged in the behaviour or
experienced the given consequence in the last 12 months using a 4-point Likert scale
ranging from 0 (never)to3(almost always).
Self-reported gambling behaviour. The Gambling Timeline Followback (G-TLFB;
Weinstock et al., 2004) was used to obtain a self-report measure of the amount of time
participants spent gambling, as well as the amount of money risked gambling (i.e. the
amount of money wagered gambling) over the three months prior to taking part in the
study, which were used as outcome measures in the current study. The G-TLFB is a
behavioural assessment instrument that consists of an individually administered
retrospective calendar (covering the past three months) that collects information on
gambling frequency, duration, type of game played, intent, risk, win –loss, and number of
standard alcoholic drinks consumed while gambling.
Self-reported gambling outcome expectancies. The 3 positive subscales (enjoyment/
arousal, self-enhancement and money) of the 23-item Gambling Expectancy
Questionnaire (GEQ; Gillespie, Derevensky, & Gupta, 2007a) were used to assess self-
reported positive gambling outcome expectancies. The GEQ served as the direct measure
of positive gambling outcome expectancies in the present study. Similar to previous
research in this area (Stewart et al., 2014), the 3 positive subscales of the GEQ were
combined in the present study in order to obtain a 15-item global measure of participants’
positive gambling outcome expectancies. Participants were asked to what extent they
expect each item/positive outcome (e.g. ‘I win money’; ‘I feel excited’) will occur when
gambling on a 7-point Likert scale ranging from 1 (no chance)to7(certain to happen).
In the current sample, this 15-item measure demonstrated good reliability (
a
¼.83).
Gambling outcome expectancy RT task. Adapted from the classic affective priming task
(Fazio, Sanbonmatsu, Powell, & Kardes, 1986), this RT-based task was used to assess
implicit gambling outcome expectancies. The task was designed to measure how quickly
individuals respond to positive and negative gambling outcome expectancy words (i.e.
targets) immediately after being primed by gambling versus control (i.e. track and field)
pictures. The task was executed via Empirisoft Inc.’s DirectRT experimental psychology
software (Jarvis, 2010). The target word exemplars were selected based on a review of
established self-report measures of gambling outcome expectancies, as well as synonyms
of words from these measures. In total, there were 10 positive outcome expectancy words
(e.g. ‘excitement’, ‘relaxation’) and 10 negative outcome expectancy words (e.g. ‘anxiety’
‘guilt’) used as targets. In addition, 10 gambling-related and 10 non-gambling-related
pictures were used as primes. The task began with 1 block of 4 practice trials, and 2 blocks
with 20 test trials each (i.e. total number of trials was 44). The stimuli for practice trials
were different from those presented during the test trials. During the task, each outcome
expectancy target word was presented twice: once preceded by a gambling prime picture,
and once preceded by a non-gambling prime picture. The order of primes and targets
within each block was counterbalanced across participants.
International Gambling Studies 127
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Each trial started with the 200 ms presentation of either a gambling-related or non-
gambling-related (i.e. track and field) picture in the centre of the screen. This was
immediately followed by a blank screen (100 ms), then by the presentation of a target word
(in the centre of the screen as well) that had either a positive (e.g. ‘enjoyment’) or negative
(e.g. ‘tension’) connotation. Participants were asked to respond to words that had a
negative connotation by clicking the ‘Z’ key on the keyboard, and to respond to words that
had a positive connotation by clicking the ‘/’ key (which was reversed for half of
participants). The length of the inter-trial interval was 1000 ms. Participants were told that
they needed to pay attention to the pictures presented on the screen as their memory for the
pictures may be tested later. Participants were also informed that the first four trials were
practice trials.
Procedure
As the current investigation was part of a larger study (Stewart et al., 2014), only those
methodological details germane to the present study are described, as the protocol for the
larger study is available elsewhere. Upon arrival at the laboratory, participants provided
informed consent. The G-TLFB (Weinstock et al., 2004) was then administered in order to
obtain a measure of participants’ self-reported gambling behaviour over the past three
months. Participants then engaged in some tasks germane to the larger study (Stewart
et al., 2014) and completed a demographic questionnaire. Following this, more tasks
pertaining to the larger study were completed. Participants then engaged in the RT task
followed by completion of the GEQ (Gillespie et al., 2007a). Participants were then
debriefed and compensated $30.
1
Results
Prior to testing our hypotheses, a log transformation was performed on the RT data to
reduce the characteristic positive skewness of RT latencies and normalize the distribution
(see Fazio, 1990; Greenwald, McGhee, & Schwartz, 1998). Also, following the
recommended procedures for correcting extremely slow and fast responses (Greenwald
et al., 1998), values below 300 ms were recoded to 300 ms and those above 3000ms were
recoded to 3000 ms. The same procedures were employed in Study 2. Following this,
correlation analyses (see Table 1) were conducted on the variables of interest, which
included the amount of time spent and money risked gambling, participants’ self-reported
Table 1. Correlations between explicit and implicit positive gambling outcome expectancies and
measures of gambling behaviour (Study 1).
1234
1. Time spent gambling – – – –
2. Money risked gambling .73** – – –
3. Positive outcome expectancy RT .34** .32* – –
4. Positive GEQ .42** .34** .23 –
Note: Time spent gambling and money risked gambling was assessed by the Gambling Timeline Followback
(G-TLFB; Weinstock et al., 2004). The positive outcome expectancy RT was computed by subtracting the log-
transformed RTs to trials involving gambling primes from the log-transformed reaction times involving non-
gambling primes. Positive GEQ ¼scores on the positive subscale of the Gambling Expectancy Questionnaire
(Gillespie et al., 2007a).
*p,.05 ** p,.01.
128 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
positive gambling outcome expectancies, and the RT difference to positive expectancy
targets following gambling primes versus non-gambling primes. This RT difference score
was computed by subtracting the mean log-transformed RT to trials involving gambling
primes from the mean log-transformed RT to trials involving non-gambling primes.
As such, a higher positive value represented a faster RT to positive targets when they
followed gambling primes (i.e. greater implicit positive gambling outcome expectancies).
As displayed in Table 1, both the self-report and RT task measure of positive gambling
outcome expectancies were moderately positively correlated with the amount of time
spent and money risked gambling. In contrast, the self-report and RT measure of positive
gambling outcome expectancies were not significantly correlated.
To test our prediction regarding the incremental contributions of the RT measure and
self-report measure of positive gambling outcome expectancies in the prediction of
gambling behaviour, a set of hierarchical regressions were conducted. Beginning with the
amount of time spent gambling, two hierarchical multiple regression analyses were
conducted to assess the unique and shared contributions of the self-report and RT measure
of positive gambling outcome expectancies to the prediction of this measure of gambling
behaviour (see Table 2). In both cases, participants’ self-reported amount of time spent
gambling over the three months prior to taking part in the study (as assessed by the G-
TLFB) served as the criterion variable.
In Model 1, the self-report measure of positive gambling outcome expectancies was
entered into the initial step of the regression and the RT measure of positive gambling
outcome expectancies was entered as an additional predictor in the second step. In Model
2, the RT measure of positive gambling outcome expectancies was entered into the first
step of the regression and the self-report measure of positive gambling outcome
expectancies was entered as an additional predictor in the second step of the regression.
In Model 1, after controlling for the self-report measure of positive gambling outcome
expectancies in Step 2 of the regression, the RT measure of positive gambling outcome
expectancies remained a significant predictor of the amount of time spent gambling,
DR
2
¼.062, p¼.04. In Model 2, after controlling for the RT measure in Step 2 of the
regression, the self-report measure of positive gambling outcome expectancies remained a
significant predictor of the amount of time spent gambling, DR
2
¼.124, p¼.004.
Table 2. Self-report and RT measures of positive gambling outcome expectancies as predictors of
the amount of time spent gambling (Study 1).
bDFpDR
2
Cumulative R
2
Model 1
Step 1:
Self-report GOE
a
.420 12.005 .001 .177** .177
Step 2:
RT measure of GOE .257 4.508 .038 .062* .239
Model 2
Step 1:
RT measure of GOE .339 7.293 .009 .115** .115
Step 2:
Self-report GOE .257 8.938 .004 .124** .239
Shared Variance
b
.053
*p,.05 ** p,.01.
a
GOE ¼gambling outcome expectancies.
b
This shared variance was calculated by taking the total proportion of explained variance and subtracting the
unique variance explained by the RT measure and the unique variance explained by the self-report measure.
International Gambling Studies 129
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Together, these models accounted for 23.9% of the total variance in the amount of time
spent gambling, R
2
¼.239, p¼.001. Of this explained variance in the amount of time
spent gambling, 26% was contributed uniquely by the RT measure of positive gambling
outcome expectancies, 52% was contributed uniquely by the self-report measure of
positive gambling outcome expectancies, and the remaining 22% was contributed by what
the RT and self-report measure of positive gambling outcome expectancies held in
common (see Figure 1).
Two hierarchical regression analyses were then conducted to assess the distinct and shared
contributions of the RT and self-report measures of positive gambling outcome expectancies
to the prediction of the amount of money risked gambling (see Table 3). In both cases,
participants’ self-reported amount of money risked gambling over the three months prior to
taking part in the study (as assessed by the G-TLFB) served as the criterion variable. Aside
from the difference in criterion variable, the analyses were performed as above.
In Model 1, after controlling for the self-report measure of positive gambling outcome
expectancies in Step 2 of the regression, the RT measure of positive gambling outcome
remained a significant predictor of the amount of money risked gambling, DR
2
¼.062,
p¼.04. In Model 2, after controlling for the RT measure in Step 2 of the regression, the
self-report measure of positive gambling outcome expectancies remained a significant
predictor of the amount of money risked gambling, DR
2
¼.077, p¼.03. Together, these
models accounted for 18.0% of the total variance in the amount of money risked gambling,
R
2
¼.18, p¼.004. Of this explained variance in the amount of money risked gambling,
34% was contributed uniquely by the RT measure of positive gambling outcome
expectancies, 43% was contributed uniquely by the self-report measure of positive
gambling outcome expectancies, and the remaining 23% was contributed by what the RT
and self-report measure of positive gambling outcome expectancies held in common (see
Figure 1). Thus, as hypothesized, results revealed that the self-report and RT measure of
positive gambling outcome expectancies each predicted unique as well as shared variance
in the prediction of the amount of time spent and money risked gambling.
Time Spent
Gambling
(Study 1)
Proportion (%) of Variance Contributed
Money
Risked
Gambling
(Study 1)
Stud
y
1 and 2 Outcome Measures
Problem
Gambling
Severity
(Study 2)
0
5
10
15
20
25 Unique Variance: RT task
Unique Variance: Self-report
Shared Variance: RT & Self-report
Figure 1. Proportion (%) of unique and shared variance contributed by the direct (self-report) and
indirect (RT task) measures of positive gambling outcome expectancies in the prediction of the
amount of time spent and money risked gambling (Study 1) and problem gambling severity (Study 2).
130 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Study 2
Results of Study 1 point to the utility of the direct and indirect measures of positive
gambling outcome expectancies in significantly predicting unique as well as shared
variance in two important indices of gambling behaviour – the amount of time spent and
money risked gambling. While such findings are important, it was also desirable to
examine the relations of positive gambling outcome expectancies with problem gambling
severity. However, this proved difficult in Study 1 due to the limited range of gambling
problems in the sample (i.e. only 19% of the sample scored in the non-problem or low-risk
range on the PGSI; Ferris & Wynne, 2001). To address this limitation and allow for an
examination of the utility of the direct and indirect measures of positive gambling outcome
expectancies in predicting problem gambling severity, we purposely recruited a new
sample of participants that exhibited a broader range of gambling problems (i.e. 40% of
the sample scored in the non-problem or low-risk range on the PGSI; Ferris & Wynne,
2001) in Study 2. Thus, Study 1 and Study 2 were independent studies.
Method
Participants
Participants (N¼96; 66 males and 30 females) for the present investigation consisted of
gamblers who were part of a larger study that investigated the impact gambling
advertisement exposure on implicit and explicit gambling outcome expectancies (Stewart
et al., under review). Participants ranged in age from 19 to 71 years (M¼29.63;
SD ¼12.08). Participants were recruited through posters placed in local universities, and
advertisements in local newspapers and classified websites. Fifty participants were
recruited from the Halifax Regional Municipality in Nova Scotia, 22 participants from the
greater Guelph area in Ontario, and 24 participants from the Winnipeg area in Manitoba.
The three sites did not significantly differ on any key demographic or outcome variables.
In order to be eligible to participate, individuals had to be 19 years of age or older and had
to have gambled at a casino or played any casino games outside a casino, gambled on a slot
machine or video lottery terminal, bet on horses at a racetrack or played dice games for
Table 3. Self-report and RT measures of positive gambling outcome expectancies as predictors of
the amount of money risked gambling (Study 1).
bDFpDR
2
Cumulative R
2
Model 1
Step 1:
Self-report GOE
a
.344 7.499 .008 .118** .118
Step 2:
RT measure of GOE .256 4.153 .040 .062* .180
Model 2
Step 1:
RT measure of GOE .321 6.433 .014 .103* .103
Step 2:
Self-report GOE .285 5.163 .027 .077** .180
Shared Variance
b
.041
*p,.05 ** p,.01.
a
GOE ¼gambling outcome expectancies
b
This shared variance was calculated by taking the total proportion of explained variance and subtracting the
unique variance explained by the RT measure and the unique variance explained by the self-report measure.
International Gambling Studies 131
Downloaded by [Dalhousie University] at 10:07 08 October 2015
money at least once over the past three months. As RT measures require extremely rapid
responses to English words, only individuals whose native language was English were
eligible to participate. Individuals were excluded if they were currently attempting to quit
gambling or receiving treatment for problem gambling. Using the PGSI from the CPGI
(Ferris & Wynne, 2001), participants consisted of 11 non-problem gamblers, 27 low-risk
gamblers, 39 moderate-risk gamblers and 19 high-risk/problem gamblers. Total scores on
the PGSI ranged from 0 to 21 (M¼4.73; SD ¼4.57).
2
Measures
The measures used to assess implicit (gambling outcome expectancy RT task) and explicit
(GEQ; Gillespie et al., 2007a) positive gambling outcome expectancies and problem
gambling severity (PGSI; Ferris & Wynne, 2001) were identical to those used in Study 1
and are not reiterated here. Both the GEQ (
a
¼.86) and PGSI (
a
¼.87) demonstrated
good internal consistency in the present sample.
Procedure
As mentioned earlier, this current investigation was part of a larger study (Stewart et al.,
under review). Only those methodological details germane to the present study are
described here as the protocol for the larger study is available elsewhere. Upon arrival at
the laboratory, participants provided informed consent. Participants then engaged in tasks
relevant to the larger study and completed a demographic questionnaire. Following their
involvement in some other tasks relevant to the larger study (Stewart et al., under review),
participants engaged in the RT task and then completed the GEQ (Gillespie et al., 2007a),
after which they were debriefed and compensated $30.
3
Results
Correlation analyses were conducted to assess the relation between problem gambling
severity, participants’ self-reported positive gambling outcome expectancies on the GEQ
(Gillespie et al., 2007a) and the RT difference to positive expectancy targets following
gambling primes vs. non-gambling primes. As in Study 1, this RT difference score was
computed by subtracting the mean log-transformed RT to trials involving gambling primes
from the mean log-transformed RT to trials involving non-gambling primes. Consistent
with predictions, both the self-report (r(94) ¼.34, p,.001) and RT measure of positive
gambling outcome expectancies (r(94) ¼.26, p¼.006) were positively associated with
level of problem gambling severity. In addition, as in Study 1, the self-report and RT
measures of positive gambling outcome expectancies were not significantly correlated
(r(94) ¼.04, p¼.37).
Given the significant relationship between problem gambling severity and both the
implicit and explicit measures of positive gambling outcome expectancies, it was of
interest to examine the incremental contributions of the RT measure and self-report
measure of positive gambling outcome expectancies in predicting problem gambling
severity. To do so, the same analytic strategy employed in Study 1 was performed.
Specifically, two hierarchical multiple regression analyses were conducted to assess the
unique and shared contributions of the self-report and RT measure of positive gambling
outcome expectancies to the prediction of problem gambling severity (see Table 4).
132 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
In both cases, participants’ level of problem gambling severity (as assessed by the PGSI of
the CPGI [Ferris & Wynne, 2001]) served as the criterion variable.
In Model 1, after controlling for the self-report measure of positive gambling outcome
expectancies in Step 2 of the regression, the RT measure of positive gambling outcome
remained a significant predictor of problem gambling severity, DR
2
¼.059, p¼.01.
In Model 2, after controlling for the RT measure in Step 2 of the regression, the self-report
measure of positive gambling outcome expectancies remained a significant predictor of
problem gambling severity, DR
2
¼.110, p¼.001. Together, these models accounted for
17.5% of the total variance in problem gambling severity, R
2
¼.175, p,.001. Of this
explained variance in problem gambling severity, 34% was contributed uniquely by the
RT measure of positive gambling outcome expectancies, 63% was contributed uniquely
by the self-report measure of positive gambling outcome expectancies, and the remaining
3% was contributed by what the RT and self-report measure of positive gambling outcome
expectancies held in common (see Figure 1). Thus, as hypothesized, results revealed that
the self-report and RT measure of positive gambling outcome expectancies each predicted
unique as well as shared variance in the prediction of problem gambling severity.
Discussion
The current research investigated whether positive gambling outcome expectancies,
assessed using both direct and indirect measures, independently predicted the amount
of time spent and money risked gambling (Study 1), and problem gambling severity
(Study 2). Specifically, the present studies examined the incremental contributions of the
direct and indirect measures of positive gambling outcome expectancies in the prediction
of self-reported gambling behaviour and problems.
As predicted, results from Study 1 revealed that the self-report and RT measure of
positive gambling outcome expectancies each predicted unique as well as shared variance
in the amount of time spent and money risked gambling. Building upon previous research
highlighting the relation between explicit gambling outcome expectancies and
adolescents’ self-reported gambling behaviour (e.g. Wickwire et al., 2010), the present
findings suggest that implicit gambling outcome expectancies are also associated with
Table 4. Self-report and RT measures of positive gambling outcome expectancies as predictors of
problem gambling severity (Study 2).
bDFpDR
2
Cumulative R
2
Model 1
Step 1:
Self-report GOE
a
.340 12.292 .001 .116** .116
Step 2:
RT measure of GOE .243 6.654 .011 .059* .175
Model 2
Step 1:
RT measure of GOE .255 6.524 .012 .065* .065
Step 2:
Self-report GOE .332 12.373 .001 .110** .175
Shared Variance
b
.006
*p,.05 ** p,.01.
a
GOE ¼gambling outcome expectancies.
b
This shared variance was calculated by taking the total proportion of explained variance and subtracting the
unique variance explained by the RT measure and the unique variance explained by the self-report measure.
International Gambling Studies 133
Downloaded by [Dalhousie University] at 10:07 08 October 2015
self-reported gambling behaviour. In sum, our findings indicate that there is merit in
including not only self-report but also RT measures when investigating the role of
outcome expectancies on gambling behaviour. However, our findings should be
interpreted with caution since participants’ gambling behaviour was based solely on their
retrospective self-report (see below for further discussion of this potential limitation).
To increase our understanding of the mechanisms involved in the relationship between
outcome expectancies and gambling behaviour, it is important that future research
examine the impact of such expectancies on prospective gambling behaviour. Indeed,
given the cross-sectional design of the current research, as well as the fact that no studies
to date have established a causal role of gambling outcome expectancies on subsequent
gambling behaviour, an important first step in establishing this causal role would be to
examine the prospective prediction of gambling behaviour in a longitudinal design, similar
to research on alcohol outcome expectancies (e.g. Wiers et al., 2002). Moreover,
considering the current findings, it is important that future research make use of both
indirect and direct assessment modes when examining the role of outcome expectancies on
prospective gambling behaviour.
Similar to Study 1 findings, the results of Study 2 revealed that both the direct and
indirect measures of positive gambling outcome expectancies were significant predictors
of problem gambling severity, with both assessment modes predicting unique as well as
shared variance in problem gambling severity. Such findings build upon previous research
examining the relationship between self-reported gambling outcome expectancies and
gambling problems (e.g. Gillespie et al., 2007b; Shead & Hodgins, 2009) by revealing that
explicit positive gambling outcome expectancies are significant predictors of gambling
problems even after accounting for the effects of another predictor (i.e. implicit positive
gambling outcome expectancies). Along with Study 1 findings, these results highlight the
potential role of implicit cognitions in gambling behaviour and problems. Although
implicit positive outcome expectancies have been found to play an important role in other
addictive behaviours, (e.g. alcohol use; see Wiers & Stacy, 2006), to our knowledge the
current research is the first to highlight the potential utility of implicit positive gambling
outcome expectancies in the prediction of gambling behaviour and associated problems.
In both studies, results revealed that relatively more variation in the prediction of the
amount of time spent and money risked gambling (Study 1) and problem gambling
severity (Study 2) was due to unique rather than shared aspects of the direct and indirect
measures of positive gambling outcome expectancies. Thus, as previously shown in the
alcohol research area (e.g. Wiers et al., 2002), our results suggest that implicit and explicit
positive gambling outcome expectancies are independent and unique predictors of
gambling behaviour (Study 1) and problem gambling severity (Study 2). In line with the
reflective-impulsive model of behaviour (Strack & Deutsch, 2004), such findings suggest
that the RT task and self-report measure of gambling outcome expectancies are tapping
into different underlying constructs. Specifically, the self-report measure may assess
gamblers’ outcome expectancies after conscious deliberation, whereas the RT measure
may assess the automatic activation of participants’ gambling outcome expectancies in
memory by exposure to the gambling prime pictures. An examination of the relationship
between the direct and indirect measures of positive gambling outcome expectancies in
Study 1 and Study 2 provides further support for this claim. Consistent with findings from
the alcohol outcome expectancy literature (e.g. de Jong et al., 2007; Jajodia & Earleywine,
2003; Kramer & Goldman, 2003; Wiers, Van Woerden, Smulders, & De Jong, 2002), the
direct and indirect measures of positive gambling outcome expectancies were not
significantly correlated, suggesting that the underlying constructs assessed using these
134 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
measures are independent. Alternatively, this non-significant correlation may be related to
motivational biases (e.g. social desirability) in self-reported gambling outcome
expectancies or a lack of introspective access to implicitly assessed constructs (Hofmann,
Gawronski, Gschwendner, Le, & Schmitt, 2005). To provide further insight into this
relationship, future research should investigate whether indirect measures of gambling
outcome expectancies assess a unique facet of the gambling cognition domain that is not
accessible via direct assessment modes.
Although the RT and self-report measures of positive gambling outcome expectancies
each provided a unique contribution to the prediction of the amount of time spent and
money risked gambling (Study 1), as well as problem gambling severity (Study 2), the
self-report measure contributed somewhat more variation in the prediction of gambling
outcomes. It is important to note, however, that the greater contribution of the self-report
measure may be due to common method variance bias, which involves the artificial
inflation of relationships between variables that are assessed using the same method (Reio,
2010). Specifically, both the questionnaires measuring gambling behaviour and problems
and those measuring positive gambling outcome expectancies were assessed using self-
report methods.
Limitations
Some limitations of the current research should be noted. First, the present studies are
correlational in nature, making it impossible to infer causality. Second, participants’
gambling behaviour in Study 1 was assessed via self-report, which is prone to social
desirability bias, acquiescence and extreme responding, and demand characteristics
(Paulus & Vazire, 2009). Further, measuring gambling behaviour via self-report relies on
the notion that gamblers are consciously aware of their gambling behaviour, and are
willing to accurately and truthfully report such behaviour to researchers (Birch et al.,
2008). In addition, there was significant variability in some of the outcome measures (i.e.
time spent and money risked gambling) and this should be taken into consideration when
interpreting the present findings. However, it should be noted that this measure of
gambling behaviour (i.e. G-TLFB; Weinstock et al., 2004) has been found to produce
reliable and valid indices of gambling behaviour (Hodgins & Makarchuk, 2003;
Weinstock et al., 2004). Nonetheless, it is important that future research examine whether
gambling outcome expectancies are capable of predicting subsequent gambling behaviour,
as has been found in the alcohol domain (e.g. Palfai & Ostafin, 2003; Wiers et al., 2002).
An additional limitation relates to the fact that a self-report measure of gambling
outcome expectancies designed for adolescent gamblers (i.e. GEQ; Gillespie et al., 2007a)
was used in the present research as there currently is no such measure specifically designed
for adults. Consequently, the GEQ (Gillespie et al., 2007a) may not be the most accurate
representation of the different outcome expectancies held by adult gamblers. That said,
however, the factor structures of other measures of adolescent gambling outcome
expectancies (i.e. the Adolescent Gambling Expectancies Survey; Wickwire et al., 2010)
have been successfully replicated in adult gamblers (Ginley et al., 2013), suggesting that
the outcome expectancies captured by these measures are relevant to both adult and
adolescent gamblers. Further, the current findings point to the relevance and validity of the
GEQ items among adults, as it was predictive of both gambling and problems.
Nonetheless, it is important that future research investigate the specific gambling outcome
expectancies held by adult gamblers and that priority be given to the development of a
self-report measure of adult gamblers’ outcome expectancies. A further limitation relates
International Gambling Studies 135
Downloaded by [Dalhousie University] at 10:07 08 October 2015
to the RT task used in both studies. Specifically, the present studies examined associations
between gambling and outcome expectancies relative to associations with a control
activity (i.e. track and field). Although we chose a control activity that we believed was
comparable to gambling, future research should examine whether similar results are found
using a different control activity. Finally, it is important to note that conclusions of the
current findings are based on the assumption that the RT measure assessed implicit
cognition. However, there is no direct evidence from the present findings that this is the
case. Despite this potential caveat, there is a broad literature base demonstrating that
implicit cognition can be reliably assessed via RT measures (see Stacy & Wiers, 2010).
Implications and directions for future research
The current research has a number of important research and clinical implications.
In relation to research implications, the present findings highlight the importance of
utilizing both direct and indirect measures when investigating the role of outcome
expectancies in gambling as both assessment modes were found to contribute unique
variance in the prediction of gambling behaviour and problems. Given the paucity of
research on gambling outcome expectancies in general, and implicit gambling outcome
expectancies more specifically, results of the present research provide an important
framework for future research examining the impact of implicit and explicit outcome
expectancies in gambling. In addition to assessing whether implicit and explicit positive
gambling outcome expectancies are capable of predicting gambling behaviour as
measured by behavioural observation in either a laboratory or natural environment (as
highlighted above), it is important that future research investigate whether direct and
indirect measures of positive gambling outcome expectancies predict cognitions and
behaviours associated with gambling problems (e.g. craving to gamble, gambling
motives). Similarly, it is important that future research examine potential mediators and
moderators of the relationship between implicit and explicit gambling outcome
expectancies and gambling behaviour, such as gambling abstinence self-efficacy
(Hodgins, Peden, & Makarchuk, 2004), problem gambling severity (Gillespie et al.,
2007b) and self-regulatory resources (Hofmann, Friese, & Strack, 2009). Given that the
current research examined global positive gambling outcome expectancies, it is important
that future research investigate the potential impact of specific positive gambling outcome
expectancies (e.g. winning, excitement) on gambling behaviour and problems among adult
gamblers. Such research may provide insight into whether certain types of positive
gambling outcome expectancies are more closely associated with increased gambling
behaviour and problems than others. It is also important that future research examine the
role of negative outcome expectancies on gambling behaviour and problems. Based on
findings from the alcohol field (e.g. Jones et al., 2001) and emerging gambling literature
(Stewart et al., 2014) suggesting that negative outcome expectancies play a less proximal
role in addictive behaviours, such expectancies may not be closely associated with
problematic gambling.
In relation to clinical implications, results of the present studies point to the potential
utility of targeting implicit gambling-related cognitions, in addition to explicit cognitions,
when designing prevention and treatment initiatives for disordered gambling. In the
alcohol field, support has been shown for the effectiveness of recently developed cognitive
retraining methods that alter implicit alcohol associations from positive to negative
(Houben, Havermans, & Wiers, 2010), with such strategies being associated with reduced
alcohol consumption and improved treatment outcomes (e.g. Wiers, Eberl, Rinck, Becker,
136 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
& Lindenmeyer, 2011). Given the promise of cognitive retraining methods in the alcohol
field, these interventions may be effective in altering implicit positive gambling outcome
expectancies as a treatment for disordered gambling.
In relation to explicit gambling outcome expectancies, it may prove useful to focus on
expectancy challenges as a gambling prevention initiative and intervention for disordered
gambling. Indeed, expectancy challenges, which aim to reduce individuals’ expectancies
about the rewarding properties of an addictive behaviour, have been used to successfully
reduce positive explicit outcome expectancies in the alcohol area (Darkes & Goldman,
1993,1998). Future research in this area may wish to educate gamblers on the role of
outcome expectancies in gambling behaviour and problems, and include exercises in
which gamblers examine environmental factors and cues (e.g. gambling advertisements)
that contribute to the expectancies they hold regarding the positive effects of gambling.
Declaration of interests
Funding sources: This work was supported by a research grant from the Ontario Problem Gambling
Research Centre awarded to Sherry H. Stewart & Sunghwan Yi, as well as an operating grant from
the Manitoba Gambling Research Program (MGRP) of Manitoba Lotteries awarded to Sherry
H. Stewart, Michael Ellery & Sunghwan Yi (Grant #410-2009-1043). However, the findings and
conclusions of this paper are those solely of the authors and do not necessarily represent the views of
Manitoba Lotteries. Melissa J. Stewart was supported by doctoral scholarships from Social Sciences
and Humanities Research Council, Killam Trusts, and Gambling Awareness Nova Scotia.
Conflicts of interest: No potential conflict of interest was reported by the authors.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1. As part of the larger study (Stewart et al., 2014), participants were randomly assigned to one of
two cue exposure conditions: a casino cue video condition (n¼29) or a control cue video
condition (n¼29). Given that cue exposure did not moderate the raw effect of each measure on
the outcomes of interest, nor did it moderate the unique effect of each measure on the outcomes
of interest, we collapsed across cue conditions and only examined participants’ post-cue RT
scores and self-reported outcome expectancies when conducting the analyses.
2. A square root transformation was performed (Tabachnick & Fidell, 2007) to reduce the positive
skewness of the distribution of scores on the PGSI (Ferris & Wynne, 2001). As a comparison of
the analyses conducted using the original and transformed data revealed no differences in
results, the original, untransformed scores on this measure were retained.
3. As part of the larger study (Stewart et al., under review), participants were randomly assigned to
one of two cue exposure conditions: a gambling advertisement cue condition (n¼51) or a
control advertisement condition (n¼45). Given that cue exposure did not moderate the raw
effect of each measure on the outcomes of interest, nor did it moderate the unique effect of each
measure on the outcomes of interest, we collapsed across cue conditions and only examined
participants’ post-cue RT scores and self-reported outcome expectancies when conducting the
analyses.
Notes on contributors
Melissa J. Stewart is currently completing her PhD in Clinical Psychology at Dalhousie University in
Halifax, Nova Scotia. Her research interests include the role of implicit and explicit gambling-
related cognitions, with a particular focus on the role of implicit and explicit outcome expectancies
in gambling. They also include responsible gambling strategies aimed at reducing the risk of
International Gambling Studies 137
Downloaded by [Dalhousie University] at 10:07 08 October 2015
problematic slot machine gambling, as well as prevention and treatment initiatives for disordered
gambling.
Sherry H. Stewart is professor in the Departments of Psychology and Neuroscience, Psychiatry, and
Community Health and Epidemiology at Dalhousie University in Halifax, Nova Scotia. Her research
is focused on psychological factors (e.g. motives, personality, implicit cognitions) contributing to
alcohol abuse, disordered gambling, and the co-morbidity of mental health and addictive disorders.
She has published several clinical trials of novel approaches for the treatment and prevention of
addictive disorders and co-occurring mental health problems.
Sunghwan Yi is an associate professor in the Department of Marketing & Consumer Studies at the
University of Guelph. His research interests include the automatic determinants of addictive,
impulsive and compulsive consumer behaviour (e.g. excessive buying, gambling), as well as
affective motives of compulsive/excessive buying and gambling.
Michael Ellery is currently a clinical specialist with the Addictions Foundation of Manitoba. His
research in the area of gambling has included examining electronic gambling behaviours,
investigating psychometric properties of gambling measures and evaluating the effectiveness of
prevention and treatment initiatives for gambling problems.
References
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders
(5th ed.). Washington, DC: Author.
Birch, C. D., Stewart, S. H., Wiers, R. W., Klein, R. M., MacLean, A. D., & Berish, M. J. (2008). The
mood-induced activation of implicit alcohol cognition in enhancement and coping motivated
drinkers. Addictive Behaviors,33, 565–581. doi:10.1016/j.addbeh.2007.11.004
Darkes, J., & Goldman, M. S. (1993). Expectancy challenge and drinking reduction: Experimental
evidence for a meditational process. Journal of Consulting and Clinical Psychology,61,
344–353. doi:10.1037/0022-006X.61.2.344
Darkes, J., & Goldman, M. S. (1998). Expectancy challenge and drinking reduction: Process and
structure in the alcohol expectancy network. Experimental and Clinical Psychopharmacology,6,
64–76. doi:10.1037/1064-1297.6.1.64
De Houwer, J. (2006). What are implicit measures and why are we using them? In R. W. Wiers &
A. W. Stacy (Eds.), Handbook of implicit cognition and addiction (pp. 11 –28). Thousand Oaks,
CA: Sage Publications.
de Jong, P. J., Wiers, R. W., van de Braak, M., & Huijding, J. (2007). Using the Extrinsic Affective
Simon Test as a measure of implicit attitudes towards alcohol: Relationship with drinking
behavior and alcohol problems. Addictive Behaviors,881, 887.
Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as an
integrative framework. In M. Zanna (Ed.), Advances in experimental social psychology
(Vol. Vol. 23, pp. 75– 109). San Diego, CA: Academic Press.
Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic
activation of attitudes. Journal of Personality and Social Psychology,50, 229 – 238. doi:10.
1037/0022-3514.50.2.229
Ferris, J., & Wynne, H. (2001). The Canadian problem gambling index: User’s manual. Ottawa: S
Canadian Centre on Substance Abuse.
Gillespie, M. A. M., Derevensky, J., & Gupta, R. (2007a). Adolescent problem gambling:
Developing a gambling expectancy instrument. Journal of Gambling Issues,19, 51 – 68. doi:10.
4309/jgi.2007.19.3
Gillespie, M. A. M., Derevensky, J., & Gupta, R. (2007b). The utility of outcome expectancies in the
prediction of adolescent gambling behaviour. Journal of Gambling Issues,19, 69 – 85. doi:10.
4309/jgi.2007.19.4
Ginley, M. K., Whelan, J. P., Relyea, G. E., Simmons, J. L., Meyers, A. W., & Pearlson, G. D.
(2013). College student beliefs about wagering: An evaluation of the Adolescent Gambling
Expectancies Survey. Journal of Gambling Studies. Advance online publication. doi:10.1007/
s10899-013-9403-8
138 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Goldman, M. S., Darkes, J., & Del Boca, F. K. (1999). Expectancy mediation of biopsychosocial risk
for alcohol use and alcoholism. In I. Kirsch (Ed.), How expectancies shape experience
(pp. 223–262). Washington, DC: APA Books.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in
implicit cognition: The implicit association test. Journal of Personality and Social Psychology,
74, 1464–1480. doi:10.1037/0022-3514.74.6.1464
Hodgins, D. C., & Makarchuk, K. (2003). Trusting problem gamblers: Reliability and validity of
self-reported gambling behavior. Psychology of Addictive Behaviors,17, 244 –248. doi:10.1037/
0893-164X.17.3.244
Hodgins, D. C., Peden, N., & Makarchuk, K. (2004). Self-efficacy in pathological gambling
treatment outcome: Development of a gambling abstinence self-efficacy scale (GASS).
International Gambling Studies,4, 99– 108. doi:10.1080/14459790412331296947
Hofmann, W., Friese, M., & Strack, F. (2009). Impulse and self-control from a dual-systems
perspective. Perspectives on Psychological Science,4, 162– 176. doi:10.1111/j.1745-6924.
2009.01116.x
Hofmann, W., Gawronski, B., Gschwendner, T., Le, H., & Schmitt, M. (2005). A meta-analysis on
the correlation between the Implicit Association Test and explicit self-report measures.
Personality and Social Psychology Bulletin,31, 1369–1385. doi:10.1177/0146167205275613
Houben, K., Havermans, R. C., & Wiers, R. W. (2010). Learning to dislike alcohol: Conditioning
negative implicit attitudes toward alcohol and its effect on drinking behavior. Psychopharma-
cology,211, 79– 86. doi:10.1007/s00213-010-1872-1
Jajodia, A., & Earleywine, M. (2003). Measuring alcohol expectancies with the Implicit Association
Test. Psychology of Addictive Behaviors,17, 126– 133. doi:10.1037/0893-164X.17.2.126
Jarvis, B. G. (2010). DirectRT. (Version 2010) [Computer Software]. New York, NY: Empirisoft
Corporation.
Jones, B. T., Corbin, W., & Fromme, K. (2001). A review of expectancy theory and alcohol
consumption. Addiction,96, 57–72. doi:10.1046/j.1360-0443.2001.961575.x
Kramer, D. A., & Goldman, M. S. (2003). Using a modified Stroop task to implicitly discern the
cognitive organization of alcohol expectancies. Journal of Abnormal Psychology,112,
171–175. doi:10.1037/0021-843X.112.1.171
Palfai, T. P., & Ostafin, B. D. (2003). The influence of alcohol on the activation of outcome
expectancies: The role of evaluative expectancy activation in drinking behavior. Journal of
Studies on Alcohol,64, 111– 119.
Paulus, D. L., & Vazire, S. (2009). The self-report method. In R. W. Robins, R. C. Fraley, &
R. F. Krueger (Eds.), Handbook of research methods in personality psychology (pp. 224 – 239).
New York, NY: Guilford Press.
Potenza, M. N. (2006). Should addictive disorders include non-substance-related conditions?.
Addiction,101, 142– 151. doi:10.1111/j.1360-0443.2006.01591.x
Reio, T. G. (2010). The threat of common method variance bias to theory building. Human Resource
Development Review,9, 405– 411. doi:10.1177/1534484310380331
Sayette, M. A. (1999). Cognitive theory and research. In K. E. Leonard & H. T. Blane (Eds.),
Psychological theories of drinking and alcoholism (pp. 247– 291). New York, NY: Guilford
Press.
Shead, N. W., & Hodgins, D. C. (2009). Affect-regulation expectancies among gamblers. Journal of
Gambling Studies,25, 357– 375. doi:10.1007/s10899-009-9131-2
St-Pierre, R. A., Temcheff, C. E., Gupta, R., Derevensky, J., & Paskus, T. S. (2014). Predicting
gambling problems from gambling outcome expectancies in college student-athletes. Journal of
Gambling Studies,30, 47– 60. doi:10.1007/s10899-012-9355-4
Stacy, A. W. (1997). Memory activation and expectancy as prospective predictors of alcohol and
marijuana use. Journal of Abnormal Psychology,106, 61– 73. doi:10.1037/0021-843X.106.1.61
Stacy, A. W., & Wiers, R. W. (2010). Implicit cognition and addiction: A tool for explaining
paradoxical behavior. Annual Review of Clinical Psychology,6, 551– 575. doi:10.1146/annurev.
clinpsy.121208.131444
Stewart, M. J., Yi, S., Ellery, M., & Stewart, S. H. (under review). An examination of the effects of
gambling advertisements on the activation of implicit versus explicit gambling outcome
expectancies among regular gamblers.
International Gambling Studies 139
Downloaded by [Dalhousie University] at 10:07 08 October 2015
Stewart, M. J., Yi, S., & Stewart, S. H. (2014). Effects of gambling-related cues on the activation of
implicit and explicit gambling outcome expectancies in regular gamblers. Journal of Gambling
Studies,30, 653– 668. doi:10.1007/s10899-013-9383-8
Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social behavior.
Personality and Social Psychology Review,8, 220–247. doi:10.1207/s15327957pspr0803_1
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston, MA: Allyn
and Bacon.
Wall, A. M., Hinson, R. E., McKee, S. A., & Goldstein, A. (2001). Examining alcohol outcome
expectancies in laboratory and naturalistic bar settings: A within-subject experimental analysis.
Psychology of Addictive Behaviors,15, 219– 226. doi:10.1037/0893-164X.15.3.219
Weinstock, J., Whelan, J. P., & Meyers, A. W. (2004). Behavioral assessment of gambling: An
application of the Timeline Followback Method. Psychological Assessment,16, 72 – 80. doi:10.
1037/1040-3590.16.1.72
Wickwire, E. M., Whelan, J. P., & Meyers, A. W. (2010). Outcome expectancies and gambling
behavior among urban adolescents. Psychology of Addictive Behaviors,24, 75 – 88. doi:10.1037/
a0017505
Wiers, R. W., Eberl, C., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2011). Re-training automatic
action tendencies changes alcoholic patients’ approach bias for alcohol and improves treatment
outcome. Psychological Science,22, 490– 497. doi:10.1177/0956797611400615
Wiers, R. W., & Stacy, A. W. (2006). Handbook of implicit cognition and addiction. Thousand Oaks,
CA: Sage Publications.
Wiers, R. W., Van Woerden, N., Smulders, F. T. Y., & De Jong, P. J. (2002). Implicit and explicit
alcohol-related cognitions in heavy and light drinkers. Journal of Abnormal Psychology,111,
648–658. doi:10.1037/0021-843X.111.4.648
Wiers, R. W., Stacy, A. W., Ames, S. L., Noll, J. A., Sayette, M. A., Zack, M. M., & Krank, M.
(2002). Implicit and explicit alcohol-related cognitions. Alcoholism: Clinical and Experimental
Research,26, 129– 137. doi:10.1111/j.1530-0277.2002.tb02441.x
140 M.J. Stewart et al.
Downloaded by [Dalhousie University] at 10:07 08 October 2015