– Masterthesis –
The Role of Approach Bias in Problem Gambling
Ruby SmitsName: Ruby Smits Studentnumber: 10002327 Date: 16-3-2016
MSc Clinical Psychology, Faculty of Social and Behavioural Sciences
Supervisor: Marilisa Boffo, David de Jong, Reinout Wiers;
Addiction Development and Psychopathology, Department of Development Psychology.
Abstract
Background A potential contributor to the course of addiction is the approach bias, an automatic tendency towards drug related stimuli. The importance of such automatic processes has been emphasized by dual process models on substance addiction. The major aim of this study was to apply these models in a behavioural addiction by investigating the role of approach bias in gambling addiction. Here we investigated the existence of an approach bias among gamblers, its relation with and predictive value on gambling problems and habitual behaviour. Since gambling behaviour is often accompanied by alcohol, we investigated the prevalence and moderating factor of alcohol problems. Methods 52 community-recruited gamblers were assigned to either the problematic gambler or the non-problem gambler group, based on their severity of gambling problems. Participants completed two online assessment sessions: one at baseline and one-month apart. Approach bias towards gambling cues was measured with a newly developed gambling Approach Avoidance Task (AAT). Severity of gambling problems, details about habitual gambling behaviour and alcohol use and problems were assessed at both time points. Results Results show a gambling-related approach bias among problem gamblers but not in non-problem gamblers. Gambling approach bias is significantly related to gambling problem severity and predicts gambling frequency over time. No difference was found in alcohol use and problems between problem and non-problem gamblers. Conclusions Consistent with dual process models, gambling cues automatically activate approach tendencies towards them. An increased approach bias relates to problem severity and gambling frequency over time. This study contributes to existent theory by the novel finding that automatic approach tendencies exist in behavioural addiction.
Index
1. Introduction 04
2. Methods 07
2.1 Study Design 07
2.2 Participants and procedure 07
2.3 Gambling AAT 08
2.4 Questionnaire measure 08
2.4.1 Problem Gambling Severity Index (PGSI) 08
2.4.2. Gambling and Time Line Follow Back (TLFB) 09
2.4.3. Alcohol Use Identification Test (AUDIT) 09
2.5 Analysis 09
3. Results 09
3.1 Demographics and assumptions 09
3.2 Group comparison on approach bias 09
3.3. Approach bias and gambling problems and behaviour 11
3.4. Predictive value of approach bias 11
3.5 Group comparison on alcohol use 11
3.6 Moderation alcohol 11
4. Discussion 11
1. Introduction
“Man is condemned to be free; because once thrown into the world, he is responsible for everything he does”, Jean Paul Sartre (1943/1974) firmly stated. One of the best ways to discuss free will is in the context of addiction. There is existing controversy about the extend to which addiction is voluntary. Gene Heyman suggests addiction is like all choices, influenced by conscious motivation (Heyman, 2010). However, a defining paradoxical feature of addiction is the continuation of a disrupting and highly harming behaviour, despite conscious motivations to the contrary (Watson, de Wit, Cousijn, Hommel & Wiers, 2013). This constitutes the destructive nature of an addiction. A considerable amount of research emphasised the importance to differentiate between automatic and more deliberate behaviours in explaining addiction, via dual-process models of cognitive processes underlying behaviour (Eberl, Wiers, Pawelczack, Rinck, Becker & Lindenmeyer, 2013; Evans, 2003; Veenstra & de Jong, 2010; Wiers, Kühn, Javadi, Korucuoglu, Wiers, Walter, Gallinat & Bermpohl, 2013). These suggest that addictive behaviour is the outcome of two qualitatively different types of processes, namely, automatic, impulsive processes and controlled, reflective processes (Beevers, 2005; Wiers, Gladwin, Hofmann, Salemink & Ridderinkhof, 2013). The automatic processes are generally fast, can occur outside of conscious awareness, and we are generally only aware of the behaviour resulting from it (Beevers, 2005; Wiers, Gladwin et al., 2013). From an evolutionary point of view, it is seen as the older process between the two and naturally result in more instinctive behaviour (Evans, 2003). The automatic processes ensure we can naturally move or behave in certain ways, without even have to think about it. This could result in cognitive biases (Molde, Pallesen, saetrevik, Hammerborg, Laberg & Johnson, 2010), which are by nature difficult to voluntarily change and control (Wiers, Gladwin et al., 2013). Reflective processes are generally slower and involve intention and awareness (Beever, 2005). These processes are more analytical (Molde et al., 2010) and accommodate flexible learning and adaptation (Wiers, Gladwin et al., 2013). The reflective process provides higher level of reasoning and can result in ‘well thought’ behavioural outcomes
(Peeters, Wiers, Monshouwe, van de Schoot, Janssen & Vollebergh, 2012), but consequently take more time. These processes underlie thinking and reasoning in everyday life (Evans, 2003) but in addiction the interaction between the processes becomes maladaptive (Wiers, Bartholow, van den Wildenberg, Thush, Engels, Sher, Grenard, Ames & Stacy, 2007). The automatic processes become stronger and if there aren’t enough resources for cognitive control they can take over.
The incentive sensitization theory provides a framework for the working mechanisms of automatic processes involved in dual process models (Robinson & Berridge, 1998; Mogg, Field, & Bradley, 2005), and have so far focused mostly on substance-related addiction (Markou, Kosten & Koob, 1997). According to this view, the brain develops a strong association between a stimuli and a reward. The wanted drug itself produces a positive incentive value, a hedonic state addicted individuals seek (Robinson & Berridge, 2000). Drug related cues are consequently seen as more attractive, grab attention and automatically elicit approach behavioural tendencies (i.e., approach bias) (Mogg et al., 2005). Neurobiological research supported these findings, showing dopamine primarily serves to draw a person’s attention to events that predict or signal reward, such as addiction-related stimuli (Schultz 1998; Franken, 2003). This increased attention activates appetitive processes, including approach behaviour (Franken, Kroon, Wiers, & Jansen, 2000) and can result in guiding behaviours toward the stimuli (Robinson & Berridge, 2000). The persistent nature of addiction disorders becomes more comprehensible in this view. While consciously fighting your needs, the subconscious draws you even more towards your wanted object. The present study focuses on the presence of an automatic approach bias in gambling addiction, which the automatic motivational tendency to approach gambling games and/or sites and related cues in the environment (Boffo, Willemen, Wiers & Dom, 2014). Translated into a real life situation, this could mean a gambler walks into a bar; the only slot-machine present automatically and unwillingly draws attention and the person physically moves more closely to it, without immediately consciously realising it.
To our knowledge, the present study is the first to investigate the role of approach bias in gambling addiction, and in a behavioural addiction in general. It was not until 2013 that the Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorder (DSM 5) (American Psychiatric Association, 2013, www.DSM5.org) included pathological gambling in the substance and other addiction disorders diagnostic category. Pathological gambling has been previously categorized as an impulse control disorder in the DSM IV (American Psychiatric Association 2000). Gambling addiction indeed shares many pathological and phenotypical features identified in substance addictions (Griffiths, 1996; Potenza, 2006), such as craving, tolerance and repeated unsuccessful attempts to cut back or quit. These and other similarities could also imply similar the presence of similar automatic cognitive processes underlying gambling addictive behaviour, as found in substance addictions (references of studies identifying approach bias in different SUDs). However, an essential difference from substance addiction is the absence of the active effect of substance use on brain structure and/or functioning (Lawrence, Luty, Bogdan, Sahakian, Clark, 2009). Therefore, if a gambling approach bias exists it would provide a deeper insight into the pure behavioural features of addiction without the correlated direct effects of the substance on neuropsychological functioning. As previous research on dual-process models and incentive sensitization focused on substance addiction, this study could make an essential contribution to the existing theory. This study could show that incentive value is not only attributed to a stimulus via a direct drug-reward association but also via behavioural associations. Therefore, investigation on this class of addiction is warranted and of great value.
Orford (2011) highlights the importance to shift the limited view that ‘addiction’ entails substance use only, of which research consequently dominates. He emphasizes to put gambling in a more central position of our gaze, because of the long established evidence of its addictiveness. Pathological gambling is a growing health concern, including financial breakdown, impaired relations with family and friends, and psychological problems (Bergh & Kühlhorn, 1994). Since legalization and
privatisation of gambling in industrialized countries is growing, gambling addiction is expected to grow even more (Ladouceur, Sylvain, Boutin, Lachance, Doucet, Leblond & Jacques, 2001). Still, the consequences of abnormal gambling behaviour remain out of sight in daily life (Black & Moyer, 1998).
Studies on the importance of approach bias in substance addiction are emerging. Wiers, Rinck, Dictus & van Den Wildenberg (2009) introduced an approach avoidance task in the addiction field, by using an alcohol applied version of the Approach Avoidance Task (AAT) (Rinck & Becker, 2007). Alcohol dependent patients were requested to push away or pull closer alcohol or control pictures based on a response cue related to the format of the picture (e.g., landscape or portrait), irrelevant to the feature of the picture. Pulling and pushing responses reflected approach and avoidance tendencies, respectively. Initially, the push and pull responses were given by using a joystick device. However, in order to develop a task that would be suitable for online assessment, Co-Investigator Wiers and his colleagues (i.e. Peeters et al., 2012) developed a modification of the task, by using keyboard keys to provide the push and pull responses. Results show that alcohol dependent patients showed a stronger approach bias towards alcohol-related stimuli than non-related stimuli (Peeters et al., 2012; Wiers, Gladwin et al., 2013; Wiers, et al., 2009; Wiers et al., 2011). More studies further explored the substance-related approach bias in other substance addictions (Brevers et al., 2011), including cannabis (Cousijn et al., 2012), nicotine (Mogg et al., 2005; Wiers, Kühn et al., 2013) and heroin stimuli (Zhou et al., 2012). However, studies on automatic processes involved in behavioural addiction disorders such as gambling, are relatively scarce (Hønsi, Mentzoni, Molde, & Pallesen, 2013). Griffiths (1994) introduced the concept of automaticity in gambling behaviour, after finding that the mind of regular gamblers ‘went black’ during periods of gambling and were unable to report on cognitive activity. A study by McCusker & Gettings (1997) confirmed the presence of automatic processes accompanying on going gambling behaviours. By assessing a Stroop task, where participants should name the colour of a word irrelevant to its contents, they found that it took pathological gambler 20 percent
longer to colour-name gambling related words than non-related words. Non-pathological gamblers showed no difference in colour name reaction time between gambling related and non-related words. This indicates an attentional bias (Cox, Fadardi, Intriligator, & Klinger, 2014). Furthermore, an implicit memory bias was found for pathological gamblers, showing automatic unintentional retrieval of gambling related words (Bradley, Mogg, & Millar, 1996). Following the incentive learning theory on substance abuse, an increased attention exists for substance-related stimuli, which activates appetitive processes, including approach behaviour. Orford (2001) showed that these appetitive processes not only result in drug-seeking behaviour, but also result in behavioural affective states as sexual desire, eating and gambling behaviour (Orford, 2001). Neurobiological studies supported this. Attention has focused on drugs affecting the dopamine system, which serves to draw a person’s attention to events that predict or signal reward. Meanwhile, the same system plays an important role in eating and sexual behaviour (Orford, 2001).
Studies that did examine automatic impulsive processes in gambling addiction focused so far only on attentional biases (Boyer & Dickerson, 2003; Brevers et al., 2011; Wölfling, Mörsen, Duven, Albrecht, Grüsser & Flor, 2011). In the context of addiction, this can be defined as an increased selective attentional focus on addiction-related stimuli (Hønsi et al., 2013; Rooke, Hine, & Thorsteinsson, 2008). Problem gamblers show greater attentional bias towards visual win-related gambling stimuli when compared to non-problem gamblers (Molde et al., 2010). The effects of attention bias on gambling addiction may be a base for developing a more comprehensive understanding of underlying cognitive factors in gambling addiction (Hønsi et al., 2013). The current study extends previous research by investigating gambling-related approach bias. The first aim of this study was to investigate whether a gambling approach bias will be stronger among problem gamblers than non-problem gamblers. To measure approach bias we used a newly developed online Gambling Approach Avoidance Task (G-AAT; Boffo, Pronk, Wiers & Mannarini, 2015), based on the alcohol approach avoidance task (Wiers et al., 2009).
Cognitive models of addiction further propose
a relationship between cognitive biases and problem severity (van Hemel-Ruiter, Wiers, Brook, & de Jong, 2015). Previous research focused on substance addiction, showing inconsistencies regarding this relationship (Field & Cox, 2008). Various studies show positive correlations with alcohol and problem severity (Cox, Yeates, & Regan, 1999; Sharma, Albery, & Cook, 2001), as well as negative correlations (Field, Mogg, & Bradley, 2005).The same contradictory results were found for smoking (Mogg & Bradley, 2002; Zack et al. 2001), cannabis (Field, Eastwood, Bradley, & Mogg, 2006; Field, Mogg, & Bradley, 2004) and heroine (Bearre, Sturt, Bruce, & Jones, 2007). In a first study in gambling addiction, no association has been found between attention bias and gambling dependence severity (Brevers et al., 2011).
The relationship between approach bias and gambling problems and behaviour was therefore the second aim of this study. To measure severity of gambling problems we used the Problem Severity Gambling Index (PGSI) (Ferris & Wynne, 2001). The Gambling – Time line Follow back (G-TLFB) (Weinstock, Whelan, & Meyers, 2004) was also used to collect information about habitual gambling behaviour. This assessment method extends diagnostic screening measures by using a more specific behavioural assessment (Weinstock et al., 2004), to identify a functional relationship between the approach bias score and gambling behaviour.
For a proper evaluation of the role of approach bias in gambling addiction, identifying the course of escalation of problems is also an important objective. If approach bias is related to problem severity and behaviour, then it will expected that approach bias would lead to more problems and gambling behaviour, as similarly found in cannabis use (Cousijn et al., 2012). If a predictive value of approach will be found, this will bring a substantial contribution to the understanding of the development of gambling addiction. Therefore, we investigated whether gambling-related approach bias predicts prospective severity of gambling problems and gambling behaviour.
Finally, the role of alcohol use is also taken into account in this study. Gambling behaviour is known to be highly comorbid with other substance use behaviours (Hall, Carriero, Takushi, Montoya, Preston & Gorelick, 2000; Ibáñez,
Blanco, Donahue, Lesieur, Pérez de Castro & Férnandez-Piqueras, 2001). Petry, Stinson & Grant (2005) found that within a sample of pathological gamblers the lifetime prevalence of comorbidity with alcohol use disorder was 73%, compared to 38% for other illicit substance abuse. The risk for severe gambling behaviour increases with alcohol use comorbidity (el-Guebaly, Patten, Currie, Williams, Beck, Maxwell & Wang, 2006). Men who drank alcohol while gambling showed an increase in bet amount and in instances of losing more than they could afford (Giacopass, Stitt & Vandiver, 1998). These findings suggest a significant role of alcohol use in gambling addiction. In the present study we therefore investigated if problem gamblers show more alcohol problems than non-problem gamblers. In line with the dual process model, it would be expected that alcohol would diminish the reflective processes and therefore increase the impact of the automatic processes, in this case the approach bias, on behavioural outcome. Therefore, the present study investigates whether alcohol moderates the effect of approach bias on gambling problems and behaviour.
The rationale for this study was to apply existing models of automatic processes in in substance addiction to a behavioural addiction, namely gambling addiction. Recapitulating, this resulted in the following hypotheses:
1. First, we hypothesized that a gambling-approach bias as measured with the Gambling Approach Avoidance Task is stronger in problem gamblers than in non-problem gamblers at baseline.
2. Second, we hypothesized that gambling approach bias will be significantly correlated with gambling involvement and problems. 3. Third, we hypothesized that
gambling-approach bias at baseline will predict prospective gambling problems and habitual gambling behaviour over time (one-month follow up).
4. Fourth, we hypothesized that problematic drinking behaviour will be greater among problem gamblers than non-problem gamblers.
5. Lastly, we hypothesized that problematic drinking behaviour will moderate the relationship between approach bias and gambling involvement and problems.
2. Methods
2.1 Study design and procedure
This study has been fully conducted online by means of the experimental online platform Lotus (UvA). Participants were pre-screened for eligibility via a registration page. Upon inclusion, participants filled in the questionnaires (Demographics, Problem gambling severity Index, Gambling – Time Line Follow Back, Alcohol Use Disorder Identification Test). The assessment session ended with the G-AAT. The session took about 25 minutes. After one month, a follow-up measurement assessment took place with the same procedure. Participants were reminded via email of the follow-up session. Upon full completion of each session, participants received a compensation of 17€ per session in the form of Bol.com vouchers. The study has been approved by the Ethics Committee of the University of Amsterdam (qualified as ‘standard research’ by Maurits van der Molen, member of the ethics Review Board, UvA).
2.2 Participants and procedure
The participants included 52 gamblers, whereof 26 problem gamblers, with a mean age of M=30,23 (SD=9.5) and 26 male non-problem gamblers, with a mean age of M=31,04 (SD=8,2). All participants were male, though we did not specifically request that. Participants were recruited through online advertisements, with use of banners on local relevant gambling websites and posts on social media like Twitter and Facebook. These advertisements invited people who ‘betted, gambled or played poker’ every now and then. Adults (>18) who gambled at least three times in the past month and were not seeking help for gambling problems were eligible for this study. Participants scoring 3 or more (i.e., at least moderate-risk gamblers) on the Problem Gambling Severity Index (PGSI) of the Canadian Problem Gambling Index (CPGI – Ferris & Wynne, 2001) were assigned to the problem-gambler group; whereas participants scoring 2 or less (i.e., non-problem and low-risk gamblers) to the non-problem-gambler group. Four participants were excluded from analysis of follow-up data, since they only completed the baseline session, resulting in a sample of 48 participants (24 problem gamblers and 24 non problem gamblers) for the follow up measurement.
Assessment
2.3 Gambling Approach Avoidance Task (G-AAT)
Gambling approach bias was assessed online by using an adapted version of the alcohol AAT (Wiers et al., 2009). The AAT is a computerized reaction time task where participants are instructed to respond as fast and accurately as possible to a feature (i.e., tilt direction) of a stimulus unrelated to its content. If a picture is tilted 3° to the left the participant should press the U key and the picture will become smaller on screen, for a push-away response. If a picture is tilted 3° to the right the participant should press the N key for a pull response, and the picture will become larger on screen. Pairing of tilt direction and push/pull responses are counterbalanced between participants. The task is composed of 3 blocks: one practice block of 8 trials and two critical blocks of 64 trials each. Stimuli are presented in push and pull format equally often. Trials are presented in pseudorandom order with maximum three consecutive repetitions of the same stimulus category (i.e., gambling and control stimuli). Half of the presented stimuli contained gambling pictures and the others matched neutral pictures (non-gambling).
Task stimuli were carefully selected and developed (Boffo et al., 2014). Stimuli are 40 pairs of matched pictures (500x500 pixels) of gambling cues and controls for each of four gambling categories: roulettes and dies, slot machines, card games, betting. Target and control pictures contained an equal number of overview and detail pictures for each gambling category and screenshots of websites when gambling category also includes online gambling sites. All gambling pictures were photographed in a naturalistic setting (i.e., casinos, betting rooms, slot machines halls). All pictures were taken with a high-resolution camera, by consistently using the same framing and shooting angle (i.e., from the front) and without using flash. Control pictures include familiar daily objects (or websites for online stimuli) or places completely unrelated to gambling (i.e., no reference to money or cash dispensers, no video games or anything remotely connected to gambling practices) and as similar as possible in complexity and pictorial features (i.e., colour, luminosity, shooting angle) to gambling pictures. At baseline, participants were asked to select the two gambling categories they
experienced to be the most problematic. A random sample of 16 gambling stimuli (8 gambling stimuli per category) and related control pictures was then presented in the task.
Consistent with previous work (Peeters et al., 2012; Wiers et al., 2010) the conventional calculation of AAT scores was based on median reaction time (RT) scores (in ms) rather than mean RT score, to minimize the influence of outliers. The approach bias score was calculated as the difference between the median RT for pushing (avoidance) and the median RT for pulling (approach) for both conditions (gambling/push – gambling/pull, neutral/push – neutral/pull) and as final difference between the two category-specific approach bias scores (Cousijn, Goudriaan, & Wiers, 2011; Rinck & Becker, 2007; Wiers et al., 2009). Test-retest reliability of the gambling approach bias score was poor (rho=-0.027), as well as split half – reliabilities, with rho=-0.12 at baseline and rho=-0.10 at follow-up
2.4 Questionnaire measures
2.4.1 Problem Gambling Severity Index (PGSI)
Severity of gambling problems was evaluated with the PGSI (Ferris & Wynne, 2001). The PGSI is a nine-item self-report scale that assesses the frequency to which participants engaged in problematic gambling behaviours (four items; e.g. ‘how often have you bet more money you could really afford to lose?’) and experienced negative gambling-related consequences (five items; e.g. ‘How often has gambling caused you any health problems, including stress or anxiety?’). Each of the items is scored on a 4-point Likert scale (0=never, 1=sometimes, 2=most of the time, 3=almost always). The total score for severity of gambling problems was calculated by summing the 9 items, with higher scores reflecting more severe gambling problems. The PGSI has demonstrated good internal consistency (α=0.84) and test-retest reliability over three-four weeks (rho=0.78) (Lesieur & Blume, 1987). In this study, acceptable Cronbach’s alphas were found at both time points (α=0.79 at baseline, α=0.76 at follow-up). Test-retest reliability was also satisfactory (rho=0.72).
2.4.2 Gambling and Alcohol Time Line Follow Back (G-TLFB)
To assess gambling behaviour, we used the Gambling – Time Line Follow Back (G-TLFB) (Weinstock et al., 2004). This screening instrument uses a retrospective calendar to measure habitual gambling behaviour over the past 30 days. Seven dimensions of gambling behaviour were assessed: Frequency, duration, intent (amount of money intended to be gambled), risk (actual amount wagered), win–loss (the amount of money won or lost during a gambling episode), and consumption of alcohol while gambling (amount of standard glasses of alcohol). The latter was included in the TLFB because, since alcohol consumption was assumed to lower inhibition (Pihl, Assaad, & Bruce, 1998) and to interact with decisions in wagering (Rosenthal & Lorenz, 1992; Weinstock et al, 2004). A sum scores for each dimension was calculated. Weinstock et al. (2004) found good test-retest reliability for the G-TLFB over a two-week period, with reliabilities for each of the gambling dimensions ranging from r=0.74 to 0.96. In this study we found acceptable test-retest reliabilities for all of the dimensions (ranging from r=0.71 to r=0.82), except for win-loss score (r=0.27). To further evaluate construct validity, correlations between the different dimensions at baseline and at follow-up were computed. All dimensions correlated significantly with each other at both time points (ranging from r=0.30 to r=0.77), except for risk and win-loss at baseline (p=0.36) and all scales with alcohol (ranging from r=-0.07 to r=-0.06).
2.4.3 Alcohol Use Disorder Identification Test (AUDIT)
We administered the Alcohol Use Disorder Identification Test (Saunders, Aasland, Babor, de la Fuente, & Grant, 1993) to assess problematic drinking behaviour. The AUDIT is a 10-item self-report questionnaire assessing typical frequency and quantity of drinking (three items; e.g. ‘How often do you have a drink containing alcohol?’) and symptoms of dependence or adverse consequences of drinking (seven items; e.g. ‘How often during the last year have you found that you were not able to stop drinking once you had started?’) (Bradley, McDonell, Bush, Kivlahan, Diehr & Fihn, 1998). Items were scored on a 5-point Likert
scale (0=never, 1=less than monthly, 2=monthly, 3=weekly, 4=daily or almost daily).
2.5 Data analysis
Before computing AAT approach bias score, error trials were removed from the task and reaction times for correct response were checked for extreme outliers (<200ms). Two participants displayed an error rate greater that 35% of the total amount of trials in the baseline or follow-up assessment session (Wiers et al., 2011) and their task data was then excluded from the analyses. The analyses plan involved the use of parametric tests (t-test, repeated-measure ANOVA and hierarchical regression) to test the study hypotheses. All dependent variables were screened for normality assumptions. In case the variable was not normally distributed, a non parametric version of the same test was used.
3. Results
3.1 Demographics and assumptions
At baseline, the groups did not differ in age or gender (p<0.05). Most variables met the assumption of homogeneity of variance between the two groups, except for the PGSI at both baseline and follow-up (p<0.01), and TLFB Frequency at follow-up (p<0.05). The assumption of normality was violated for all variables (p<0.01), except for AUDIT and gambling approach bias at follow-up (p>0.05).
3.2 Group comparison on approach bias
At baseline, participants showed a gambling approach bias significantly different from zero (p=0.048), with a non-parametric one-sample Wilcoxon Signed Rank Test. A non-parametric Mann-Whitney test was used to examine group differences on approach bias scores at baseline (see table 1). In line with hypothesis 1, problem gamblers (Mdn=35.25) showed a significantly greater gambling approach bias than non-problem gamblers (Mdn=-4), U=198, z=-2,562, p=0.01 (Fig.1).
Table 1 Mean and standard deviation scores of AAT, PGSI and TLFB in problem gamblers and non-problem gamblers, baseline and follow up.
Baseline Follow up
Problem gambler Non-problem gambler Problem gambler Non-problem gambler
Variable Mean (sd) AAT 49.4 (83.1) -2.9 (59.5) -16.2 (52.8) 13.4 (56.9) PGSI 5.7 (3.1) 1.1 (0.8) 0.9 (3.3) 1.9 (1.3) TLFB – Frequency 362.6 (983.7) 242 (456.2) 881.4 (2369.1) 362.3 (901) TLFB – Duration 1162.8 (876.6) 951.6 (1140.9) 1250.1 (1429) 1150.6 (1458.7) TLFB – Intention 1280.9 (3156.6) 771.9 (1924.9) 2447.2 (7815.9) 2230.2 (7956) TLFB – Risk 1544.2 (3479.1) 630.9 (1875.3) 2437.4 (7598.2) 2120.9 (7689.3) TLFB – Win/Lose 155.3 (693.3) 211.8 (424.8) 322 (679.1) 197.6 (562.6) TLFB – Alcohol 26 (38.7) 69 (187.7) 45 (112.1) 118 (433)
AUDIT 8.5 (5.7) 6.8 (4.3) N.A. N.A.
Note: Mean scores, with SD (standard deviation) in parentheses.
Fig. 1 Approach bias scores (in ms) of the gambling AAT task. Positive approach scores are indicative of approach tendencies, negative of avoidance tendencies.
50.00 40.00 30.00 20.00 10.00 0.00 -10.00
problem gambler non-problem gambler
Problem gambler vs. non-problem gambler group
M
ean appr
oach bias scor
3.3 Relation between Gambling approach bias and gambling problems and behaviour
Table 2 presents the Spearman’s rho correlation coefficients between gambling approach bias and PGSI and TLFB. Results show that approach bias score correlated significantly with the PGSI score at baseline (rho=0.32, p=0.01). No correlation was found between approach bias and any of the TLFB dimensions.
3.4 Predictive value of approach bias on gambling problems and behaviour
To test if gambling approach bias at baseline would predict future gambling problems and behaviour at follow-up, we performed a hierarchical linear regression analysis. In the regression model, baseline PGSI or TLFB scores were entered in the first block to control for baseline measurement; in the second block, group was entered first followed by the baseline gambling approach bias. Gambling approach bias significantly predicted Frequency of gambling episodes over and above baseline gambling frequency, F=2.1, R²=.93, df=48, β=0.09, t=2.2, p=0.03. Group did not predict frequency of gambling episodes (p=0.67). Gambling approach bias did not predict any of the other self-report gambling measures (PGSI, p=0.74; TLFB Duration, p=0.85; TLFB Intention, p=0.29; TLFB Risk, p=0.31; TLFB Win-loss, p=0.18, and TLFB alcohol, p=0.22).
3.5 Group comparison on alcohol use
To investigate whether problem gamblers show greater alcohol consumption and alcohol-related problems than non-problem gamblers, we performed an independent-sample t-test. Scores on the AUDIT and on the TLFB alcohol did not differ significantly between groups (p>0.05).
3.6 Moderating effect of alcohol on the relationship between gambling approach bias and gambling problems and behaviour
Congruent with a failure to find any significant correlation between gambling approach bias and alcohol use and/or problems, alcohol use and problems did not moderate the relationship between gambling approach bias and follow up PGSI or TLFB (ps > 0.05).
4. Discussion
The aim of this study was to investigate whether a stronger approach bias towards gambling stimuli exists among problem gamblers than non-problem gamblers and if this approach bias is associated with and can predict gambling problems and habitual behaviour. Furthermore, amount of alcohol consumption and its moderating role on approach bias and gambling problems and behaviour were
Table 2 correlation coefficients approach bias scores and gambling problems and behavior
Variable rs (p-value) PGSI 0.32 (p=0.01) TLFB – Frequency -0.11 (p=0.23) TLFB – Duration 0.10 (p=0.23) TLFB – Intention -0.07 (p=0.32) TLFB – Risk 0.10 (p=0.23) TLFB – Win/Lose -0.03 (p=0.42) TLFB – Alcohol -0.08 (p=0.28)
assessed. Firstly, results illustrated that problem gamblers showed a stronger gambling-related approach bias than non-problem gamblers. This was in line with our first hypothesis. Second, gambling approach bias was positively related to gambling problems. Unexpectedly, no relation was found between gambling approach bias and habitual gambling behaviour at baseline. Still, approach bias towards gambling cues at baseline did predict gambling frequency after one month after controlling for gambling frequency at baseline, but not gambling problems and any other TLFB index of habitual gambling behaviour. Finally, alcohol use was the same between problem and non-problem gamblers. Congruent with this finding, alcohol use did not moderate the effect of approach bias on gambling problems and gambling behaviour. The finding that problem gamblers showed a stronger approach bias than non-problem gamblers is consistent with previous results finding a drug-related approach bias towards alcohol stimuli (Wiers et al., 2011), cannabis stimuli (Cousijn et al., 2011), cigarette stimuli (Mogg et al., 2005) and heroine stimuli (Zhou et al., 2012). As was similarly found in a study on cannabis use (Cousijn et al., 2011), where cannabis approach bias predicted prospective cannabis use but not cannabis problems, gambling related approach bias also predicted prospective gambling frequency after one month but not gambling problems. In line with dual process models, automatic processes seemed to be a contributor in maintaining gambling addiction. Gambling cues elicited an automatically activated approach tendency towards them. This finding extends previous research by showing the existence of automatic, impulsive processes in a non-substance related addiction. To the best of our knowledge, this is the first study to find that approach bias predicts frequency of a behavioural addiction. By showing that automatic processes also played a role in a behavioural addiction, this result demonstrated that not merely the effect of the drug induces an incentive value and maintains the persistent nature of the disorder.
The most important finding was that the approach bias predicted changes in frequency of gambling behaviour after one month. Engaging in the gambling situation can be seen as a first-step in problematic gambling behaviour. The duration of the gambling period, intended bet, actual amount
wagered and the winning and loses resulting from it, all happens áfter the first step into the gambling situation is made. The frequency of engagement into gambling situations could consequently precede following habitual problematic behaviour. An addicted gambler wouldn’t get drawn into the automaticity of his behaviour in the gambling situation if he wouldn’t be involved in the gambling situation in the first place. Therefore, the fact that the approach bias specifically contributed to the frequency of gambling behaviour is a crucial contribution to an insight of the course and continuation of the gambling addiction.
The finding that alcohol use did not differ between problem and non-problem gamblers was unexpected. A possible explanation for this finding might be the fact that the comparison group was not recruited from a general healthy population. In fact, non-problem gamblers did gamble at least three times or more in the past month, although not experiencing problematic gambling behaviour. It might be possible that the two groups are therefore comparable in characteristics such as alcohol use and therefore show no difference on the AUDIT score. However, this explanation remains speculative until a control group of healthy people who never gambled is tested.
Some potential limitations must be taken into account. First, to assess habitual gambling behaviour the TLFB was used. It has to be considered that this questionnaire is a relatively time-consuming questionnaire, of which participant’s concerns confronted other researchers before (Sobell, Agrawal, Sobell, Leo, Young, Cunningham & Simco, 2003). It requests patience and attention. In the development of the G-TLFB Weinstock et al. (2004) found a pattern of underreporting on the G-TLFB. Remarkably, in a study on alcohol addiction, underreporting was significantly greater among diagnosed alcohol abusers than the non-pathological group (Searles, Helzer, & Walter, 2000). Some entered question marks as values. Past research on alcohol use found that participants report more alcohol consumption with daily diaries than with retrospective methods (Lemmens, Tan, & Knibbe, 1992; Searles, Helzer & Walter, 2000; Simpura & Poikolainen, 1983). Furthermore, the assessment was conducted online. For this reason, there was no supervision over the administration of the questionnaires and tasks. Therefore, opportunities
to give erroneous reports were possibly higher than in more controlled lab assessment sessions. However, an issue gamblers specifically value is anonymity and online assessment supports this value. Furthermore, the online assessment makes it more easy to reach a gambling population, which is hard to reach. Only a few studies investigated the problems encountered in online assessment (Pedersen, Grow, Duncan & neighbors, 2012; Rueger, Trela, Palmeri & King, 2012). In a study on marihuana use (Pedersen et al., 2012) strong psychometric properties were demonstrated for a self-administered, online version of the TLFB and found no differences between an online and an in-person version. A similar result was found in a study on smoking and drinking, which concluded that the web-based TLFB is well suited for assessment (Rueger et al., 2012). These results suggest that online assessment of the TLFB should not be a constraint and even has advantages. In addition to that, due to issues of anonymity, online assessment remains the preferable assessment method. To overcome these limitations it may be worthwhile to use a daily, at home, gambling measurement in future research. A possible way to assess this is by Ecological momentary assessment (EMA) (Stone, Hufford, & Shiffman, 2008). EMA involves assessment of current behaviours and experiences of the participants, in their natural environment. This has proven to minimize recall bias and maximize ecological validity (Stone et al., 2008). A possible way to integrate this assessment method in assessing habitual gambling behaviour in future studies is using a daily report measurement method. Participants shall receive daily notifications, with a reminder to fill in their gambling behaviour that day. It would be preferable if participants could report their daily use ad-hoc via an app on their phone. In this way the notification is easy noticeable and reporting requests minimum effort. It might even be useful to use a voice recording system, on which participants can record verbally their daily use via a touch pad on their telephone, to minimize effort even more. However, more investigation is needed to further elaborate these possibilities. Second, the results on the approach bias task have to be interpreted with some caution, since it has shown poor reliabilities in this study. Limitations of reaction time (RT) measurements are a more often encountered problem (Pachella,
1974; Ratcliff, 1993). Wiers et al. (2013) found similar suboptimal reliability scores. They suggest a possible alternative instrument would be a Stimulus Response Compatibility task (SRC) (Spruyt, de Houwer, Tibboel, Verschuere, Crombez, Verbanck, Hanak, Brevers & Noël, 2013), which is a more direct measure of approach bias than the relatively implicit AAT task. The SRC might be more reliable than the AAT, but different studies highlight the benefits of the AAT over the SRC (Eberl et al., 2013; Veenstra & de Jong, 2010; Wiers et al., 2013). Most importantly, the SRC is sensitive to strategic influences since it measures explicit approach tendencies instead of the implicit measurement of the AAT. Furthermore, in contrast to the SRC, the approach or avoidance tendencies are accompanied with a zooming effect of the stimuli in the AAT, resembling more real-life results of the participant’s movements (Wiers et al., 2013) To the best of our knowledge, a more sufficient measurement method for assessing approach bias in gambling behaviour does not yet exist. Since the gambling specific AAT is newly developed in this study, more research could focus on further improving this task.
Furthermore, the finding that approach bias predicted habitual gambling frequency is a unique, novel finding. However, surprisingly we did not find a predictive value of approach bias on gambling problems and the other TLFB indices of habitual gambling behaviour over time. This might be due to the fact there was a one-month follow-up, which might have been too short to detect changes in gambling problems and habitual behaviour. For the PGSI this is especially an important notification, since the PGSI measures gambling problems over the past 12 months. A one-month follow up would largely result in the same score. It would be recommended to extend the assessment time window, to improve the validity of the results and give the possibility for more substantial changes. A final point that should be kept in mind is the fact that participants were compensated with money. This might have affected their motivations. Possibly the participants only performed the tasks because they would be compensated for it and were not intrinsically motivated to contribute to the study, which could have compromised their accuracy in performance on the task and completing the questionnaires. Future studies could be held with participants who are not compensated for it
with an external reward.
Although some limitations could be considered in this study, there were many strengths. The present study is the first study thus far to investigate and show the presence of a gambling related approach bias underlying gambling problems and predicting future chances of engaging in gambling behaviour. A new gambling approach
avoidance task was developed and conducted online to assure anonymity. This study can be seen as a pilot study for future studies and results offer confidence to further explore approach bias in gambling addiction. This study gives rise for further elaboration and improvement in investigating automatic processes in gambling addiction and behavioural addiction in general.
References
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed., text revision (DSM-IV-TR). Washington, DC: American Psychiatric Publishing, Inc., 2000.
American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5). Diagnostic and Statistical Manual of Mental Disorders 4th edition TR.
Bearre, L., Sturt, P., Bruce, G., & Jones, B. T. (2007). Heroin-related attentional bias and monthly frequency of heroin use are positively associated in attenders of a harm reduction service. Addictive Behaviours, 32(4), 784–792.
Beevers, C. (2005). Cognitive vulnerability to depression: A dual process model. Clinical Psychology Review, 25(7), 975–1002. Bergh, C., & Kühlhorn, E. (1994). Social, psychological and physical consequences of pathological gambling in Sweden.
Journal of Gambling Studies, 10(3), 275–285.
Black, D. W., & Moyer, T. (1998). Clinical features and psychiatric comorbidity of subjects with pathological gambling behaviour. Psychiatric Services (Washington, D.C.), 49, 1434–1439.
Boffo, M., Pronk, T., Wiers, R. W., & Mannarini, S. (2015). Combining cognitive bias modification training with motivational support in alcohol dependent outpatients: study protocol for a randomised controlled trial. Trials, 16(1), 63.
Boffo, M., Willemen, R., Wiers, R. W., & Dom, G. (n.d.). TOP-Gambling training: Study protocol Study protocol, 1–15. Boyer, M., & Dickerson, M. (2003). Attentional bias and addictive behaviour: Automaticity in a gambling-specific modified Stroop task. Addiction, 98(1), 61–70.
Bradley, B. P., Mogg, K., & Millar, N. (1996). Pergamon Memory bias in clinical & non-clinical depression. Science, 34(11), 865–879.
Bradley, K. a, McDonell, M. B., Bush, K., Kivlahan, D. R., Diehr, P., & Fihn, S. D. (1998). The AUDIT alcohol consumption questions: reliability, validity, and responsiveness to change in older male primary care patients. Alcoholism, Clinical and Experimental
Research, 22(8), 1842–1849.
Brevers, D., Cleeremans, A., Bechara, A., Laloyaux, C., Kornreich, C., Verbanck, P., & Noël, X. (2011). Time course of attentional bias for gambling information in problem gambling. Psychology of Addictive Behaviours, 25(4), 675–682.
Cousijn, J., Goudriaan, A. E., Ridderinkhof, K. R., van den Brink, W., Veltman, D. J., & Wiers, R. W. (2012). Approach-Bias Predicts Development of Cannabis Problem Severity in Heavy Cannabis Users: Results from a Prospective FMRI Study. PLoS ONE, 7(9), e42394.
Cousijn, J., Goudriaan, A. E., & Wiers, R. W. (2011). Reaching out towards cannabis: approach-bias in heavy cannabis users predicts changes in cannabis use. Addiction, 106(9), 1667–1674.
Cox, W. M., Fadardi, J. S., Intriligator, J. M., & Klinger, E. (2014). Attentional bias modification for addictive behaviours: clinical implications. CNS Spectrums, 19(3), 215–24.
Cox, W. M., Yeates, G. N., & Regan, C. M. (1999). Effects of alcohol cues on cognitive processing in heavy and light drinkers.
Drug and Alcohol Dependence, 55(1-2), 85–89.
Eberl, C., Wiers, R. W., Pawelczack, S., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2013). Approach bias modification in alcohol dependence: Do clinical effects replicate and for whom does it work best? Developmental Cognitive Neuroscience, 4, 38–51.
el-Guebaly, N., Patten, S. B., Currie, S., Williams, J. V. a., Beck, C. a., Maxwell, C. J., & Wang, J. L. (2006). Epidemiological Associations between Gambling Behaviour, Substance Use & Mood and Anxiety Disorders. Journal of Gambling Studies, 22(3), 275–287.
Evans, J. (2003). In two minds: dual-process accounts of reasoning. Trends in Cognitive Sciences, 7(10), 454–459.
Field, M., & Cox, W. (2008). Attentional bias in addictive behaviours: A review of its development, causes, and consequences.
Drug and Alcohol Dependence, 97(1-2), 1–20.
Field, M., Eastwood, B., Bradley, B. P., & Mogg, K. (2006). Selective processing of cannabis cues in regular cannabis users.
Drug and Alcohol Dependence, 85, 75–82.
Field, M., Mogg, K., & Bradley, B. P. (2004). Cognitive bias and drug craving in recreational cannabis users. Drug and Alcohol
Dependence, 74(1), 105–111.
Field, M., Mogg, K., & Bradley, B. P. (2005). Craving and cognitive biases for alcohol cues in social drinkers. Alcohol and
Alcoholism, 40(6), 504–510.
Franken, I. H. A. (2003). Drug craving and addiction: integrating psychological and neuropsychopharmacological approaches.
Progress in Neuro-Psychopharmacology and Biological Psychiatry, 27(4), 563–579.
Franken, I. H., Kroon, L. Y., Wiers, R. W., & Jansen, a. (2000). Selective cognitive processing of drug cues in heroin dependence.
Journal of Psychopharmacology (Oxford, England), 14(4), 395–400.
Griffiths, M. (1996). Behavioural addiction: an issue for everybody? Journal of Workplace Learning, 8(3), 19–25.
Hall, G. W., Carriero, N. J., Takushi, R. Y., Montoya, I. D., Preston, K. L., & Gorelick, D. a. (2000). Pathological gambling among cocaine-dependent outpatients. The American Journal of Psychiatry, 157(7), 1127–1133.
Heyman, G.M. (2009). Addiction: A disorder of choice. Harvard University Press.
Hønsi, A., Mentzoni, R. A., Molde, H., & Pallesen, S. (2013). Attentional Bias in Problem Gambling: A Systematic Review.
Journal of Gambling Studies, 29(3), 359–375.
Ibáñez, a, Blanco, C., Donahue, E., Lesieur, H. R., Pérez de Castro, I., Fernández-Piqueras, J., & Sáiz-Ruiz, J. (2001). Psychiatric comorbidity in pathological gamblers seeking treatment. The American Journal of Psychiatry, 158(10), 1733–5.
Ladouceur, R., Sylvain, C., Boutin, C., Lachance, S., Doucet, C., Leblond, J. & Jacques, C. (2001). Cognitive Treatment of Pathological Gambling. The Journal of Nervous and Mental Disease, 189(11), 774–780.
Lawrence, A. J., Luty, J., Bogdan, N. a., Sahakian, B. J., & Clark, L. (2009). Impulsivity and response inhibition in alcohol dependence and problem gambling. Psychopharmacology, 207(1), 163–172.
Lemmens, P., Tan, E. S., & Knibbe, R. a. (1992). Measuring quantity and frequency of drinking in a general population survey: a comparison of five indices. Journal of Studies on Alcohol, 53(5), 476–486.
Mogg, K., & Bradley, B. P. (2002). Selective processing of smoking-related cues in smokers: manipulation of deprivation level and comparison of three measures of processing bias. Journal of Psychopharmacology (Oxford, England), 16(4), 385–392.
Mogg, K., Field, M., & Bradley, B. P. (2005). Attentional and approach biases for smoking cues in smokers: An investigation of competing theoretical views of addiction. Psychopharmacology, 180(2), 333–341.
Molde, H., Pallesen, S., Sætrevik, B., Hammerborg, D. K., Laberg, J. C., & Johnsen, B.-H. (2010). Attentional biases among pathological gamblers. International Gambling Studies, 10(1), 45–59.
Orford, J. (2001). Addiction as excessive appetite. Addiction, 96(January 2000), 15–31.
Pachella, R. G. (1974). The interpretation of reaction time in information processing research. Human Information Processing:
Tutorials in Performance and Cognition, (November), 41–82.
Pedersen, E. R., Grow, J., Duncan, S., Neighbors, C., & Larimer, M. E. (2012). Concurrent validity of an online version of the Timeline Followback assessment. Psychology of Addictive Behaviours: Journal of the Society of Psychologists in Addictive Behaviours, 26(3), 672–7.
Peeters, M., Wiers, R. W., Monshouwer, K., van de Schoot, R., Janssen, T., & Vollebergh, W. a M. (2012). Automatic processes in at-risk adolescents: The role of alcohol-approach tendencies and response inhibition in drinking behaviour. Addiction, 107(11), 1939–1946.
Petry N. M., Stinson F.S., Grant B.F:, (2005). Comorbidity of DSM-IV pathological gambling and other psychiatric disorders: Results from the National Epidemiologic Survey on Alcohol and Related Conditions. J Clin Psychiatry, 66: 564-574.
Potenza, M. N. (2006). Should addictive disorders include non-substance-related conditions? Addiction, 101(SUPPL. 1), 142–151. Ratcliff, R. (1993). Methods for dealing with reaction time outliers. Psychological Bulletin.
Rinck, M., & Becker, E. S. (2007). Approach and avoidance in fear of spiders. Journal of Behaviour Therapy and Experimental
Psychiatry, 38(2), 105–120.
Addiction, 95(8s2), 91–117.
Rooke, S. E., Hine, D. W., & Thorsteinsson, E. B. (2008). Implicit cognition and substance use: A meta-analysis. Addictive
Behaviours, 33(10), 1314–1328.
Rueger, S. Y., Trela, C. J., Palmeri, M., & King, a C. (2012). Self-administered web-based timeline followback procedure for drinking and smoking behaviours in young adults. J Stud Alcohol Drugs, 73(5), 829–833.
Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption--II.
Addiction, 88, 791–804.
Searles, J. S., Helzer, J. E., & Walter, D. E. (2000). Comparison of drinking patterns measured by daily reports and timeline follow back. Psychology of Addictive Behaviours, 14(0893-164X (Print)), 277–286.
Sharma, D., Albery, I. P., & Cook, C. (2001). Selective attentional bias to alcohol related stimuli in problem drinkers and non-problem drinkers. Addiction, 96(2), 285–295.
Simpura, J., & Poikolainen, K. (1983). Accuracy of Retrospective Measurement of Individual Alcohol-Consumption in Men - a Reinterview after 18 Years. Journal of Studies on Alcohol, 44(5), 911–917.
Sobell, L. C., Agrawal, S., Sobell, M. B., Leo, G. I., Young, L. J., Cunningham, J. a, & Simco, E. R. (2003). Comparison of a quick drinking screen with the timeline followback for individuals with alcohol problems. Journal of Studies on Alcohol, 64(6), 858–861.
Spruyt, A., De Houwer, J., Tibboel, H., Verschuere, B., Crombez, G., Verbanck, P., … No??l, X. (2013). On the predictive validity of automatically activated approach/avoidance tendencies in abstaining alcohol-dependent patients. Drug and Alcohol
Dependence, 127(1-3), 81–86.
Stone, A. A., Hufford, M. R., & Shiffman, S. (2008). Ecological Momentary Assessment. Annual Review of Clinical Psychology, 4, 1–32.
van Hemel-Ruiter, M. E., Wiers, R. W., Brook, F. G., & de Jong, P. J. (2015). Attentional bias and executive control in treatment-seeking substance-dependent adolescents: A cross-sectional and follow-up study. Drug and Alcohol Dependence. Veenstra, E. M., & de Jong, P. J. (2010). Restrained eaters show enhanced automatic approach tendencies towards food.
Appetite, 55(1), 30–36.
Watson, de Wit, Cousijn, Hommel, & Wiers. (2013). Motivational Mechanisms Underlying the Approach Bias to Cigarettes.
Journal of Experimental Psychopathology, 4(3), 250–262.
Weinstock, J., Whelan, J. P., & Meyers, A. W. (2004). Behavioural assessment of gambling: an application of the timeline followback method. Psychological Assessment, 16(1), 72–80.
Wiers, C. E., Kühn, S., Javadi, A. H., Korucuoglu, O., Wiers, R. W., Walter, H. & Bermpohl, F. (2013). Automatic approach bias towards smoking cues is present in smokers but not in ex-smokers. Psychopharmacology, 229(1), 187–197.
Wiers, R. W., Bartholow, B. D., van den Wildenberg, E., Thush, C., Engels, R. C. M. E., Sher, K. J. & Stacy, A. W. (2007). Automatic and controlled processes and the development of addictive behaviours in adolescents: A review and a model. Pharmacology
Biochemistry and Behaviour, 86(2), 263–283.
Wiers, R. W., Eberl, C., Rinck, M., Becker, E. S., & Lindenmeyer, J. (2011). Retraining automatic action tendencies changes alcoholic patients’ approach bias for alcohol and improves treatment outcome. Psychological Science: A Journal of the American
Psychological Society / APS, 22(4), 490–497.
Wiers, R. W., Gladwin, T. E., Hofmann, W., Salemink, E., & Ridderinkhof, K. R. (2013). Cognitive Bias Modification and Cognitive Control Training in Addiction and Related Psychopathology: Mechanisms, Clinical Perspectives, and Ways Forward.
Clinical Psychological Science, 1(2), 192–212.
Wiers, R. W., Rinck, M., Dictus, M., & Van Den Wildenberg, E. (2009). Relatively strong automatic appetitive action-tendencies in male carriers of the OPRM1 G-allele. Genes, Brain and Behaviour, 8(1), 101–106.
Wiers, R. W., Rinck, M., Kordts, R., Houben, K., & Strack, F. (2010). Retraining automatic action-tendencies to approach alcohol in hazardous drinkers. Addiction, 105(2), 279–287.
Wölfling, K., Mörsen, C. P., Duven, E., Albrecht, U., Grüsser, S. M., & Flor, H. (2011). To gamble or not to gamble: At risk for craving and relapse - learned motivated attention in pathological gambling. Biological Psychology, 87(2), 275–281.
Zack, M., Belsito, L., Scher, R., Eissenberg, T., & Corrigall, W. A. (2001). Effects of abstinence and smoking on information processing in adolescent smokers. Psychopharmacology, 153(2), 249–257.
Zhou, Y., Li, X., Zhang, M., Zhang, F., Zhu, C., & Shen, M. (2012). Behavioural approach tendencies to heroin-related stimuli in abstinent heroin abusers. Psychopharmacology, 221(1), 171–176.
