Tempting fate: Chasing and maladaptive personality

0 downloads 0 Views 991KB Size Report
36.74 years; SD = 12.3), was recruited from four Videolottery Terminal (VLT) gambling venues ... created using a computerized random number generator.

Tempting fate: Chasing and maladaptive personality traits in gambling behavior Giovanna Nigro, Maria Ciccarelli, Marina Cosenza Department of Psychology, Università degli Studi della Campania "Luigi Vanvitelli", Caserta, Italy

Address of the Authors: Department of Psychology, Università degli Studi della Campania "Luigi Vanvitelli" – Viale Ellittico, 31, 81100 Caserta, Italy.

Corresponding author: Prof. Giovanna Nigro – Department of Psychology, Università degli Studi della Campania "Luigi Vanvitelli" – Viale Ellittico, 31, 81100 Caserta, Italy. Tel: + 39 823274432; fax: +39 823323000 E-mail: [email protected]

Authors ORCID: Giovanna Nigro: 0000-0003-3518-2468 Maria Ciccarelli: 0000-0002-7285-1707 Marina Cosenza: 0000-0002-5813-017X

1

Abstract Chasing, or continuing to gamble in an attempt to recoup losses, is a salient feature of problematic gambling. This study, which controlled for gambling severity and alcohol consumption, investigated the association between chasing and maladaptive personality trait domains among habitual gamblers. Participants comprised 126 adult habitual gamblers (73% males) aged between 18 and 69 years. They were administered the South Oaks Gambling Screen (SOGS), the Personality Inventory for DSM-5-Brief Form (PID-5-BF), the Alcohol Use Disorders Identification Test (AUDIT), and a computerized task developed to assess chasing behavior. Participants were randomly assigned to two chasing conditions (Control and Loss). Data were submitted to correlational analysis, univariate and mixed-model ANOVAs, logistic and linear regression analyses. Results showed that the decision to chase was strongly associated with the PID-5-BF Disinhibition domain scores, whereas chasing proneness was related to the Disinhibition, Detachment and Psychoticism domains. Interestingly, chasers scored higher than nonchasers on maladaptive personality dimensions, even after controlling for gender, age, chasing condition, alcohol consumption, and gambling severity. Since these findings support the idea that chasers and nonchasers are different subtypes of gamblers, clinical interventions should take into account the additive role of chasing in gambling disorder.

Keywords: Gambling; gambling disorder; chasing; maladaptive personality domains; DSM-5; PID5-BF Highlights The study analyzed the role of PID-5-BF maladaptive personality domains in chasing (i.e., continued gambling to recoup losses). Strong correlations were found in 126 Italian habitual gamblers between gambling severity and within-session chasing. 2

Chasers outscored nonchasers on PID-5 even after controlling for gambling severity. The PID-5 Disinhibition domain was the strongest predictor of the decision to chase. Disinhibition, Detachment, and Psychoticism predicted chasing persistence.

3

1. Introduction Chasing losses is a salient feature of problematic gambling and represents an important step in the development and maintenance of gambling disorder (Breen and Zuckerman, 1999; Corless and Dickerson, 1989; Lesieur, 1979; Lister et al., 2016; McBride et al., 2010; O’Connor and Dickerson, 2003; Toce-Gerstein et al., 2003). Chasing consists in continuing gambling to recoup previous losses (Lesieur, 1979). “The "chase" begins when a gambler bets either to pay everyday bills that are due or to “get even" from a fall” (Lesieur, 1984, p. 1). According to Lesieur (1979), “If we have to draw a line between the compulsive and noncompulsive gambler (there is some overlap), it is in the amount of “chasing” done by the compulsive gambler” (p. 81). Toce-Gerstein et al. (2003) found that 75,9% of problem gamblers, as defined by life-time NODS score, reported chasing losses. According to Lesieur (1984), it is useful to distinguish between chasing within a session and chasing across sessions. Chasing within a session is typical of regular gamblers, whereas returning later to chase losses is a distinguish characteristic of compulsive gambling. In the same vein, and in line with DSM-IV criterion (APA, 1994), Breen and Zuckerman (1999) introduced the distinction between within- and between-session chasing. Even if chasing is one of the few observable signs for disordered gambling (Gainsbury et al., 2014), the only criterion of gambling disorder absent in substance use disorder (Quester and Romanczuk-Seiferth, 2015), and has been recognized playing a central role in the development of gambling disorder, to date very little is understood about this complex behavior (CampbellMeiklejohn et al., 2008; Nigro et al., 2018; Parke et al., 2016; Worhunsky et al., 2017). Therefore, it is not surprising that experimental research on this topic still remains relatively scarce (for a review, see Lister et al., 2016). Generally speaking, research on chasing focused mostly on the role of individual differences on chasing behavior. In the following brief review of the extant literature we describe aims and main results of studies on chasing, indicating for each contribution the effects due to gender, if any.

4

Breen and Zuckerman (1999) investigated the role of individual differences, namely sensation-seeking and impulsivity, in within-session chasing in a sample of male undergraduates. They found that chasers were higher in impulsivity than nonchasers. O’Connor and Dickerson (2003) found that chasing losses and chasing wins were both associated with impaired control over gambling. No gender difference in chasing, nor in impaired control was observed. Linnet et al. (2006) compared pathological and non-pathological gamblers in episodic chasing. They observed that pathological gamblers showed significantly more chasing and poorer decision-making strategies than non-pathological gamblers, mostly among males. Campbell-Meiklejohn (2008) analyzed the brain mechanisms of chasing losses in minimally-experienced gamblers and found that chasing was associated with increased activity in cortical areas linked to incentive-motivation and an expectation of reward. No gender difference was reported. Kim and Lee (2011) studied the effects of the Behavioral Approach System (BAS) and the Behavioral Inhibition System (BIS) on decision-making. They found that the association between personality traits and winning probabilities influence decisions-making, providing evidence that decision-making and chasing in gambling situations, both after having winning and losing experiences, are affected by personality traits. No difference between male and female participants in chasing behavior was reported. Lister et al. (2016) analyzed the role of gambling goals (i.e., gambling achievementorientation) on chasing behavior in a sample of young adult gamblers. They observed that gamblers with higher winning money motivation were more likely to decide to chase and chased more in response to either losses or wins. No gender difference was reported. Bibby (2016) carried out two experiments analyzing the role of alexithymia and impulsivity in loss-chasing. Results showed that alexithymics are more likely to loss-chase than non-

5

alexithymic individuals. This author did not found main effects or interactions between gender and the other study variables. Parke et al. (2016) investigated the impact of stake size, a risk factor for loss-chasing, on inhibition and reflection impulsivity. Results indicated that decision-making was more impaired at higher stakes in comparison to lower stakes. Gender differences were not tested, since only two females participated in the study. Worhunsky et al. (2017) compared the neurocognitive mechanisms of chasing in individuals with gambling disorder and cocaine-use disorder with healthy controls. Relative to healthy controls, disordered gamblers’ choices to quit chasing were associated with greater engagement of a medial frontal executive-processing network. They found no main effect of gender or group-by-gender interaction between- or within-subjects factors across components by examining activity in functional brain networks. Finally, Nigro et al. (2018) investigated the relation between chasing and decision-making. They found that chasing affected decision-making and that the association between gambling severity and decision-making performance was significantly mediated by chasing. These authors reported that males scored significantly higher than females in terms of both gambling severity and chasing frequency. To our best knowledge, only a handful of studies assessed chasing in gambling using behavioral tasks. With the exception of Linnet et al. (2006), who measured episodic chasing (i.e., sequences of persistent poor choices leading to losses) within the Iowa Gambling Task (IGT), Bibby (2016), Breen and Zuckerman (1999), Campbell-Meiklejohn et al. (2008), Lister et al. (2016), Nigro et al. (2018), and Worhunsky et al. (2017) developed or implemented ad hoc procedures for estimating within-session chasing. More specifically, Breen and Zuckerman measured chasing by means of a computer-generated gambling program, based on the Newman et al. (1987) model, developed for studying antisocial personality. The program was modified to reward a subject at a predetermined rate, which was high initially, but diminished steadily the 6

longer each subject played (p. 1103). The object of the game was to bet on randomly generated cards from a deck of playing cards. Campbell-Meiklejohn et al. (2008) administered a computerized loss-chasing game that required choosing between gambling to recover a loss, at risk of doubling its size, or quitting (p. 294). Lister et al. (2016) asked participants to play slot machines located in an immersive virtual casino environment. After the first 30 spins, participants decided whether they wished to stop or to continue play. The subsequent number of plays following that decision were tallied. Bibby (2016) measured chasing by means of an adapted version of the Cambridge Gambling Task (Rogers et al., 1999), originally developed to assess decision making and risk taking behavior outside a learning context. After 10 practice trials, the participants played the gambling task for 100 trials. Worhunsky et al. (2017) assessed chasing using a modified version of the loss-chase task developed by Campbell-Meiklejohn et al. (2008). Finally, Nigro et al. (2018) developed a computerized chasing task that simulates a card game in which participants played against the house. After the first 30 trials, participants could decide whether to continue or to stop the game. Although in the above-mentioned papers chasing behavior was measured in quite different ways, the tasks share some following overlapping features: the game was apparently chance-determined, the outcomes of the game were manipulated, participants won or lost some money, and both the decision to chase and the number of trials played were considered measures of interest. In addition, participants could stop in any moment during the task. Summing up, the few behavioral tasks devoted to assess chasing focused on within-session chasing, given the difficulties in reproducing in the laboratory between-session chasing, as defined by Lesieur (1979) and the DSM-IV and the DSM-5 criterion for disordered gambling. Although these studies did not report additional information about the psychometric properties of the behavioral tasks, there is no reason to doubt about their ecological validity, mostly because these procedures simulate real life game situations. Although both the DSM-IV and the DSM-5 focus on chasing losses [“After losing money gambling, often returns another day to get even (“chasing” one’s losses)”], in broader sense, chasing refers to gaining more or recouping lost money (e.g., Blaszczynski and Nower, 2002). For 7

instance, O’Connor and Dickerson (2003), who investigated the role of chasing in relation to impaired control over gambling, observed that between-session chasers reported significantly higher impaired control scores than within-session chasers, but found no difference between returning later to chase after large wins or after losing. Ultimately, since the house always wins, in the long run the inability to stop gambling might turn wins in losses. In such a perspective, chasing wins and chasing losses might be rather regarded as two sides of the same coin (Nigro et al., 2018). As noted by Linnet et al. (2006), “one of the inherent problems of chasing to the pathological gambler may be that they do not notice the chasing behavior until it is too late, [...] and then has limited resources for stopping or compensating the behavior” (p. 48). Relative to healthy controls, problem gamblers are less likely to adopt a long term advantageous strategy, even in the face of negative feedback (Linnet, 2013). Although previous research has investigated the role of maladaptive personality traits in the development and maintenance of problematic gambling (for review, see Carlotta et al., 2015), only few studies examined the association between individual characteristics and loss chasing behavior. Prior investigations have shown that loss chasing is linked with impulsivity, sensationseeking, emotional disregulation, hyposensitivity to losses, behavioral disinhibition, and impaired decision-making (Bibby, 2016; Breen Zuckerman, 1999; Linnet et al., 2006; Kim and Lee, 2011; Lister et al., 2016; Nigro et al., 2018; Ochoa et al., 2013; Parke et al., 2016). It is noteworthy to underline that although previous studies clearly indicated that several individual characteristics of gambling addicted individuals are strong predictors of chasing, little effort was made to ascertain to what extent gambling severity and chasing overlap each other, and even less to disentangle gambling severity from chasing. However, to date no study examined the relation between chasing behavior and DSM-5 pathological personality trait domains. The present study aimed to fill this gap by investigating this association in a sample of Italian habitual players. As alcohol and gambling problems show high cooccurrence among both adults and adolescents (e.g., Barnes et al., 2009; Nigro and Cosenza, 2016; 8

Tackett et al., 2017; for reviews, see Rahman et al., 2014 and Rash et al., 2016), we controlled for alcohol consumption. Furthermore, in the attempt to disentangle chasing from gambling involvement we analyzed the association between chasing and maladaptive personality trait domains controlling for gambling severity. In line with previous studies (Lister et al., 2016; Nigro et al., 2018), we expected that gambling severity would be reliably related to within-session chasing. Given that previous research indicated that gambling disordered individuals scored higher on measures of maladaptive personality traits than controls (for a review see Carlotta et al., 2015), and since chasing represents a criterion of gambling disorder, we would expect significant associations between chasing behavior and maladaptive personality traits. More specifically, it was expected that maladaptive personality trait domains would be positively associated with both the decision to chase and chasing proneness, independently of gambling severity.

2. Method 2.1 Participants A convenience sample of 126 Italian adults (73% males), aged between 18 and 69 years (Mage = 36.74 years; SD = 12.3), was recruited from four Videolottery Terminal (VLT) gambling venues offering the same range of gambling activities (games of chances, such as slot machines, casino games, etc.). Of the participants, 37.3% were single, 57.1% married, and 5.6% separated or widowed. With regard to modal occupation status, 24.6% of participants were office workers, 23.8% manual workers, 18.3% unemployed, and 12.7% students. The two inclusion criteria were as follows: 1) participants reported to gamble once a week or more and 2) were 18 years of age or over. Participants were tested individually in a quiet room made available by the management. 220 habitual players were requested to participate voluntarily and no compensation was provided for this study. The percentage of people who declined to participate in the study was 40% (N = 88).

9

2.2 Procedure The ethics committee of the research team's university department approved the present study. Written informed consent was obtained prior to enrolment. Participants subsequently completed a computerized task and three questionnaires. Half of participants were administered the computerized task at the beginning of the session, the other half at the end. As chasing task had two conditions (Control and Loss, respectively), an equal number of participants (N = 63) was randomly assigned to each condition following block randomization procedure. Randomization sequence was created using a computerized random number generator. The three questionnaires were administered in counterbalanced order (counterbalanced measures design: ABC, ACB, BAC, BCA, CAB and CBA). Administration of the instruments required from a minimum of about 25 min to a maximum of about 40 min.

2.3 Measures Participants were administered the South Oaks Gambling Screen (SOGS; Lesieur and Blume, 1987; Italian translation: Cosenza et al., 2014b) to assess the degree and level of problem gambling severity, the Personality Inventory for DSM-5-Brief Form (PID-5-BF)-Adult (Krueger et al., 2012; Italian validation: Fossati et al., 2013), to measure five maladaptive trait dimensions, the Alcohol Use Disorders Identification Test (AUDIT; Saunders et al., 1993) to assess alcohol consumption, drinking behaviors, and alcohol-related problems, and a computerized task developed to measure chasing behavior (ChasIT; Nigro et al., 2018). The SOGS is a self-report instrument made up of 20 scored items and some un-scored items. The total score ranges from 0 to 20. Scores from 0 to 2 indicate no gambling problems, scores of 34 indicate the risk of developing gambling problems, and a score of 5 or above denotes (probable) pathological gambling involvement. The un-scored items request participants to indicate, among others, the frequency of participation in different gambling activities (“not at all”, “less than once a 10

week” or “once a week or more”), the largest amount of money gambled in one day, and parental involvement in gambling. SOGS in the present study had a high internal consistency (Cronbach’s alpha = 0.91). The PID-5-BF is a self-report inventory developed to assess the DSM-5 (APA, 2013) Section III, pathological personality trait domains. This measure consists of 25 items rated on a 4-point Likerttype scale ranging from 0 (very false or often false) to 3 (very true or often true), assessing five dimensions of maladaptive personality traits, including Negative Affect, Detachment, Antagonism, Disinhibition, and Psychoticism. Negative affectivity refers mainly to emotional instability, namely prevalence of negative emotions, such as anxiety, depression, worry, anger, etc., and their behavioral and interpersonal manifestations. Detachment implies avoidance of socio-emotional experience, restricted affective experience and expression, reduced hedonic capacity. Antagonism includes attention seeking, callousness, deceitfulness, grandiosity, and manipulativeness. Disinhibition implies orientation toward immediate gratification, acting impulsively, and inattention to the future consequences of present behavior. Finally, Psychoticism is defined by the exhibition of eccentric or unusual cognitions and behaviors (see Krueger and Markon, 2014). The PID-5-BF is an abbreviate version of the PID-5 (Krueger et al., 2011). Recent studies on both adults and adolescents (e.g., Anderson et al., 2016; Bach et al., 2016; Fossati et al., 2015; for a review, see Al-Dajani et al., 2016) support the use of the PID-5-BF as a screening measure of dimensional maladaptive personality traits. Higher average scores indicate greater dysfunction in a specific personality trait domain. In the present study, the Cronbach α values for the Negative Affect, Detachment, Antagonism, Disinhibition, and Psychoticism scales were 0.70, 0.70, 0.73, 0.89, and 0.71, respectively. The AUDIT is a 10-item measure of alcohol consumption, drinking behavior, and alcohol-related problems. Responses to each item are scored from 0 to 4, giving a maximum possible score of 40. In the present study the Cronbach α value was 0.84.

11

Chasing was measured using a computerized task (ChasIT). The task (ChasIT) simulated a card game in which participants played against the house. The starting amount of money was 10 Euros and participants were encouraged to treat the play budget as real money. Each card reported a number ranging from 1 to 9. Participants won 1 Euro if they had the highest card. If not, they lose the same amount of money. In both cases, participants received a positive (“You won 1 Euro!”) or a negative (“You lost 1 Euro!”) feedback on the computer screen and heard a sound, that varied according to the result. After the first 30 trials, participants were informed that they had completed the first part of the task and informed about the amount of money they had saved or lost with respect to the initial budget. In the first phase of the task wins and losses varied as a function of condition. In the control condition participants saved the entire budget, whereas in the loss condition they lost 12 Euros, namely the entire budget plus 2 Euros. Notwithstanding this deficit, participants were allowed to continue. In the control condition participants were informed that they saved the entire budget. In both conditions, participants could decide whether to continue playing or to stop the game. For the subsequent 30 trials, after each trial participants received a positive o negative feedback and were informed about the amount of residual credit. At this point participants had to decide if they want continuing or stopping the game, by pressing the key “M” to continue or the key “Z” to stop playing. Since participants could continue playing up to the end, the chasing maximum total score was 30. In the control condition the final budget was 10 Euros, in the loss condition minus 14 Euros. The number of wins and losses varied as function of condition (15 and 15 in the first and second part of the control condition and 9 and 21 in the loss condition). The two blocks of wins and losses were randomized, but the sequence was the same for each condition. Participants who chose to stop gaming at the beginning of the second phase of the computerized task were classified “nonchasers” (0), whereas participants who decided to continue gaming were classified “chasers” (1). The decision to continue play or to stop it, as well as the number of trials played were the two dependent measures of interest. 12

3. Results Data analyses were conducted using IBM SPSS version 20.0. The alpha level was set at p < 0.05. All variables were initially screened for missing data, distribution abnormalities, and outliers (Tabachnick and Fidell, 2013). Using p < .001 criterion for Mahalanobis distance, four male and two female participants were eliminated as clear multivariate outliers. This left a final sample size of 126. Pearson correlation coefficients were calculated to examine the relationships among the study variables. Analysis of variance was used to assess mean differences on continuous variables. For categorical data differences in percentages were compared with the Chi-square test. Logistic and linear regression analyses were performed to examine the unique contribution of predictor variables to chasing decision and chasing total score, respectively. In order to control for the presence of multicollinearity, before interpreting the regression coefficients, we calculated the variance inflation factors (VIF), which were below the recommended cutoff of 10 (max. VIF = 1.841; Ryan, 1997). In accordance with SOGS scoring system respondents were classified in the following three categories: non-problem gamblers (score 0-2), problem gamblers (score 3-4), and probable pathological gamblers (score > 5). The percentage of non-problem gamblers was 38.04% among males and 70.59% among females, whereas the percentage of problem gamblers was 15.22% among males and 8,82% among females. Finally, the percentage of probable pathological gamblers was 46.74% among males and 20.59 among females. Results of Chi-square test showed significant association between the two categorical variables (χ 2 (2, N = 126) = 10.65; p < 0.005). However, standardized residuals indicated that only the proportion of females classified as “non-problem gamblers” was significantly larger than the proportion of males (SR = 2.00).

13

Of the total sample (N = 126) 53.3% of males and 41.2% of females were in the control group, whereas 46.7% of males and 58.8% of females in the loss group. Results of Chi-square test did not show significant association between the two categorical variables (χ2 (1, N = 126) = 1.45; p = 0.316). Furthermore, the two groups were homogeneous in terms of age F1, 124 = 0.73; p = 0.393, year of education (F1, 124 = 0.88; p = 0.35), and marital status (χ2 (2, N = 126) = 0.92; p> 0.05). To test for gender differences in the study variables data were preliminarily submitted to univariate ANOVAs. Results showed significant effects due to gender on SOGS (F1, 124 = 11.09; p < 0.001, ηp2 = 0.082), PID-5-BF Detachment (F1, 124 = 4.55; p < 0.05, ηp2 = 0.035) and Antagonism (F1, 124 = 6.68; p < 0.05, ηp2 = 0.051) domains, and the chasing total scores (F1, 124 = 5.63; p < 0.05, ηp2 = 0.043), with males outperforming females. Means and standard deviations as a function of gender are reported in Table 1. As regards chasing, 62.7% of participants decided to chase. The average number of trial played was 7.24 (SD = 10.05). INSERT TABLE 1 ABOUT HERE Furthermore, in order to ascertain whether age, education (in years), and SOGS total score varied as function of gender and experimental condition, three 2 X 2 univariate ANOVAs were performed (see Table 2). Neither significant main effects, nor interaction effects due to experimental condition were observed (all ps > 0.05). As expected, and in line with previous statistical analyses, males scored higher than females on SOGS total score only. INSERT TABLE 2 ABOUT HERE The relationships between all variables were assessed first using Pearson product-moment correlations (see Table 3). INSERT TABLE 3 ABOUT HERE As can be seen, with only very few exceptions zero-order correlations between all variables were moderate to strong in strength. However, since age was in turn positively associated with the Detachment PID-5-BF trait domain (r = 0.32; p < 0.001), and in light of the results of univariate 14

ANOVAs, to determine whether the measures remained correlated after controlling for gender and age, partial correlations among the variables were computed (see Table 4). As can be seen, correlations between the study variables remained significant. INSERT TABLE 4 ABOUT HERE To ascertain whether chasing frequency varied as a function of chasing condition (Control vs. Loss), we run a univariate ANOVA. Results showed a significant effect of condition (F1, 124 = 7.91; p < 0.01, ηp2 = 0.060), indicating that participants played a higher number of trials in the Loss condition than in the Control condition. Chi-square test was used to determine whether there was a relationship between the choice to continue or stop gaming and the score reported on the SOGS item related to chasing behavior (namely, item 4: “When you gamble, how often do you go back another day to win back money you lost?”). Results revealed a significant positive association between the two dummy variables (χ2 (1, N = 126) = 23.36; p < 0.001; Cramer’s V = 0.431), indicating that scores on self-reported betweensession chasing and within-session chasing as measured by the behavioral task were correlated. More specifically, the percentages of between-session chasers, namely participants who scored 1 on item 4, was 48.1%, whereas the percentage of between-session nonchasers was 6.7%. Furthermore, score on the SOGS item 4 was significantly associated with ChasIT total score (r = 0.48; p< 0.001). In order to ascertain whether both the decision to chase and chasing persistence varied as a function of gender and type of game (skill versus chance games) data were submitted to chi-square test and univariate ANOVA. Preliminarily we collapsed the games in two broader categories: skill games (card games, such as poker and blackjack, and sport betting) and chance games (roulette, slot machines games, videolottery, and lotto). Chi-square test did not show a significant association between the decision to chase and the type of game (χ2 (1, N = 126) = 0.536; p = 0.350). Results of 2 X 2univariate ANOVA, with gender and type of game as factors and the number of plays following the decision to chase as dependent variables yielded a significant main effect due to

15

gender (F1, 124 = 5.92; p < 0.05, ηp2 = 0.042), indicating that female participants chased less than males. To determine whether chasers differed from nonchasers on PID5-BF trait domains scores a repeated measures ANOVA with group (nonchasers versus chasers) as a between-subjects factor, the PID-5 domain scores as within-subjects variables, and gender, age, condition, and AUDIT scores as covariates, yielded significant between-subjects effect of PID-5-BF (F1, 120 = 13.41; p < 0.001; ηp2 = 0.101), revealing that, relative to nonchasers, chasers scored higher on the five PID-5BF trait domains. More interestingly, a subsequent analysis showed that these differences remained statistically significant even after inserting SOGS scores among covariates (F1, 119 = 4.11; p < 0.05; ηp2 = 0.043), indicating that chasers outperformed nonchasers on the PID-5-BF trait domains scores over and above gambling severity (see Figure 1). INSERT FIGURE 1 ABOUT HERE To evaluate the relative contribution of gender, age, condition, problematic alcohol use, and the PID-5-BF domains to chasing decision we conducted a hierarchical logistic regression analysis using the decision to stop or continue play as the criterion variable. The results of the final regression model indicated that gender, age, and high scores on the Disinhibition domain predicted significantly the decision to chase (see Table 5). INSERT TABLE 5 ABOUT HERE Finally, results of hierarchical linear regression analysis with ChasIT total score as the dependent variable, and gender, age, chasing condition, PID-5-BF domains, and AUDIT scores as independent variables, indicated that male gender, loss condition, and high scores on Disinhibition, Detachment, and Psychoticism trait domains significantly predicted within-session chasing behavior (see Table 5), with the overall model explaining more than a third part of the total variance (R2adj = 0.37; F5, 110 = 13.34; p < 0.001). INSERT TABLE 6 ABOUT HERE 4. Discussion 16

The present study is the first research that examined the association between the DSM-5 maladaptive personality domains and chasing behavior in a sample of habitual players controlling for alcohol use and gambling severity. Consistent with our initial hypotheses, both the decision to chase and chasing proneness were significantly related to gambling severity, and chasing behavior was significantly associated with high scores on maladaptive personality domains, even after controlling for gambling severity. The positive association of chasing decision and chasing persistence, respectively, with the SOGS total score is inconsistent with the results reported by Breen and Zuckerman (1999), who observed no reliable association between SOGS scores and within-session chasing, but supports the results of Lister et al. (2016), which showed that problem gambling severity scores were positively related to both the decision to chase and the number of chasing spins played in the face of loss. More interestingly, the significant positive correlation between both chasing decision and chasing persistence and the score reported on the SOGS question dealing with chasing represents a kind of indirect validation of the ChaseIT task and suggests the idea that self-reported between-session chasing and within-session chasing could share very similar features. We agree with Parke et al. (2016), according to whom “escalation of monetary wins and losses in a gambling context appears to affect reward and punishment sensitivity in subsequent gambling activity, which in turn, may begin to account for within-session loss-chasing despite the probability of exacerbating the negative consequences of gambling” (p.731). Results of regression analyses demonstrated that the decision to chase was strongly related with male gender and positively associated with age and the PID-5 Disinhibition domain scores, whereas chasing persistence was related, along with gender and chasing condition, to the Disinhibition, Detachment, and Psychoticism maladaptive personality domains. That males were more prone to chase and chased more than female is not surprisingly considering the strong association between SOGS and chasing scores. Indeed, a large body of research demonstrated that male gender is one of the most powerful predictors of gambling disorder. In addition, our results are 17

consistent with prior studies indicating that males chased more than females (Linnet et al., 2006; Nigro et al., 2018). Interestingly, chasing condition (Control vs. Loss) predicted chasing persistence, but not the decision to continue or stop gaming. Summing up, males were more prone to chase than females and chased more when they had to recoup previous losses. Interestingly, the results of repeated measures ANOVA clearly indicated that chasers scored higher on maladaptive personality trait domains than nonchasers, even after controlling for gambling severity. As logistic regression analysis showed, gender and Disinhibition, i.e., the tendency to be impulsive, risk-taking, irresponsible, and distractible, were the most powerful predictors of the decision to chase. According to Krueger and Markon (2014), individuals who scores high on this domain are, among others, impulsive, oriented toward immediate gratification, and insensitive to the future consequences of current behavior. This finding is consistent with the result reported by Breen and Zuckerman (1999) who identified higher levels of impulsivity as a predictor of withinchasing behavior, and Parke et al. (2016) who observed that, independently of winning or losing, players who can choose to gamble at disparate stake sizes showed impairment in reflection impulsivity, namely the tendency to make impulsive decisions. Furthermore, these results are also in line with previous research on gambling addiction, demonstrating that different independent domains of impulsivity, including reward and punishment sensitivity, delay discounting, and cognitive impulsivity (MacKillop, et al., 2014; see also Goudrian et al., 2014), risk proneness, and insensitivity to future consequences (Cosenza et al., 2017; Nigro and Cosenza, 2016; Nigro et al., 2017) are significant correlates of problematic gambling. As regards chasing propensity, results showed that, along with Disinhibition, Detachment, and Psychoticism trait domains were significant predictors of the number of plays following the decision to chase. According to Hopwood et al. (2013) “detached individuals may be emotionally anhedonic or depressed, and may generally tend to avoid and withdraw from others, of whom they may be suspicious” (p. 168). Even if the Detachment trait domain can be regarded as a reason for 18

chasing, rather than a consequence of heavy gambling, it may be that the excessive participation in gambling activities contributes to further impoverish social relations and affective experiences, mostly when gambling do not involve other people. Ultimately, the repeated “interaction” with video lottery terminals and slot machines can result in promoting loneliness and social isolation. Since we found a strong relationship between chasing behavior and gambling severity and given the nature of our sample (frequent players), it is not surprising the association between detachment and chasing. Indeed, prior studies showed that detachment was associated with gambling disorder (Carlotta et al., 2015), suggesting that disordered gamblers tend to escape socioemotional experiences. Nor is surprising that chasing was related to Psychoticism. Recently, Gainsbury et al. (2014) found that individuals who reported chasing losses were more likely to endorse irrational beliefs about gambling wins than those who reported that they were unaffected by previous losses. This results is consistent with previous research demonstrating that a wide range of culturally incongruent odds, cognitive distortions and erroneous beliefs about gambling is related to problematic gambling and foster the persistence of gambling despite negative outcomes (e.g., Cosenza et al., 2014; Johansson et al., 2009; Oei, et al., 2008; Oei and Raylu, 2015; Raylu et al., 2016; Taylor et al., 2014). On the whole, these findings suggest that between- and within-session chasing share similar features and support the idea that chasers and nonchasers represent two different subtypes of gamblers. If chasing is a characteristic of heavy gamblers, it is probably that who chases along a single session do not differ substantially from who chases day by day, even if “control of session duration in a continuous form of gambling is harder than the control of frequency of the sessions themselves, although the two are interrelated” (Dickerson, 1991, p. 334). In brief, it may be that the repeated exposure to gambling cues, as in the case of frequent players, fosters chasing behavior. Future studies should analyze more deeply the relationship between within- and between session chasing. If within- and between-chasers and nonchasers are different sub-types of gamblers, treatment protocols and clinical interventions should take into account the additive role of chasing 19

in gambling disorder. Of course, further research is needed to substantiate the conclusion that the boundaries between within-session and across session chasing are somewhat fuzzier than it looks, whereas the distance between chasers and nonchasers is greater than it looks. In our opinion, the most intriguing and partially unanswered question is whether gambling addiction and chasing share the same neurobiological and cognitive underlying factors. A promising avenue could be the study of brain structure and functional connectivity in chasers and nonchasers (for a review, see Quester and Romanczuk-Seiferth, 2015). Controlling for gambling severity is mandatory, given that some studied clearly indicated that chasing behavior is not limited to pathological gamblers (Canale et al., 2016; for the prevention paradox see also Delfabbro and King, 2017). Future research should also compared chasing losses with chasing wins in order to better understand the role of winning money motivation, as well as of other reasons underlying the decision to chase. Even though the differences we observed between the control and the loss conditions indicated that losses foster chasing, it may be that also wins promote chasing behavior. In all cases, it could be useful to deeply investigate the motivational factors affecting the decision to chase and chasing perseverance. Since from a clinical perspective identifying and separating distinct subgroups of gamblers implies differing management strategies (Blaszczynski and Nower, 2002), future studies on the role of chasing in gambling disorder could contribute to provide useful insights for tailored clinical interventions and preventions programs and, ultimately, to improve the effectiveness of treatments.

4.1. Limitations Although several strengths characterized this study, including the use of a behavioral task to assess chasing, there are some limitations that should be considered when interpreting the present results. First, the participants were recruited using convenient sampling of Italian habitual players. Since prevalence rates of gambling disorder varied across countries, and cross-cultural studies (e. g. 20

Medeiros et al., 2015) suggest that cultural aspects may play a relevant role in gambling addiction, one must be particularly cautious regarding possible generalizability of our findings. Second, the current data are mainly based on self-report measures, which may limit the generalizability of the present study's results due to recall bias and social desirability. Finally, gambling severity was assessed by means of a measure that has been criticized for excessive false positives (Goodie et al., 2013). However, it is worth to note that the SOGS demonstrated satisfactory reliability and validity (Stinchfield 2002; see also Barbaranelli et al. 2013). Finally, since we did not collect data about motivation underlying the decision continue or stop gaming, we are unable to distinguish chasing behavior from gambling habituation or gambling tolerance. In addition, we did not bear out that the decision to stop gaming could results from a lack of interest in continuing playing. Among others, the small amount of money at stake could be one of the factors influencing the choice. On the other hand, people could decide to chase because the task was enjoyable or to reduce the boredom of the experimental condition. Since previous studies did not deeply investigate the motivational determinants of chasing behavior, further research is needed to clarify the role of motivation in gambling persistence. Despite these limitations, the present study provides new information regarding the relationship between chasing behavior and maladaptive personality trait domains.

21

References Al-Dajani, N., Gralnick, T.M., Bagby, R.M., 2016. A psychometric review of the Personality Inventory for DSM–5 PID–5: Current status and future directions. J. Pers. Assess. 98, 62-81. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th ed.. American Psychiatric Press, Washington, DC. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, 5th ed.. American Psychiatric Press, Washington, DC. Anderson, J.L., Sellbom, M., Salekin, R.T., 2016. Utility of the Personality Inventory for DSM-5– Brief Form PID-5-BF in the measurement of maladaptive personality and psychopathology. Assessment, DOI: 10.1177/1073191116676889 Bach, B., Maples-Keller, J.L., Bo, S., Simonsen, E., 2016. The alternative DSM–5 personality disorder traits criterion: A comparative examination of three self-report forms in a Danish population. Personal. Disord. 7, 124-135. Barbaranelli, C., Vecchione, M., Fida, R., Podio-Guidugli, S., 2013. Estimating the prevalence of adult problem gambling in Italy with SOGS and PGSI. J. Gambling Issues, 1-24. Doi: 10.4309/jgi.2013.28.3 Barnes, G.M., Welte, J.W., Hoffman, J.H., Tidwell, M.C.O., 2009. Gambling, alcohol, and other substance use among youth in the United States. J. Stud. Alcohol Drugs 70, 134-142. Bibby, P.A., 2016. Loss-chasing, alexithymia, and impulsivity in a gambling task: Alexithymia as a precursor to loss-chasing behavior when gambling. Front. Psychol. 7, 1-13. Blaszczynski, A., Nower, L., 2002. A pathways model of problem and pathological gambling. Addiction 97, 487-499. Breen, R.B., Zuckerman, M., 1999. ‘Chasing’ in gambling behavior: Personality and cognitive determinants. Pers. Individ. Diff. 27, 1097-1111. Campbell-Meiklejohn, D.K., Woolrich, M.W., Passingham, R.E., Rogers, R.D., 2008. Knowing when to stop: The brain mechanisms of chasing losses. Biol. Psychiatry 63, 293-300. 22

Canale, N., Vieno, A., Griffiths, M.D., 2016. The extent and distribution of gambling-related harms and the prevention paradox in a British population survey. J. Behav. Addict. 5, 204-212. Carlotta, D., Krueger, R.F., Markon, K.E., Borroni, S., Frera, F., Somma, A., et al., 2015. Adaptive and maladaptive personality traits in high-risk gamblers. J. Pers. Disord. 29, 378-392. Corless, T., Dickerson, M., 1989. Gamblers’ self‐perceptions of the determinants of impaired control. Addiction 84, 1527-1537. Cosenza, M., Baldassarre, I., Matarazzo, O., Nigro, G., 2014a. Youth at stake: Alexithymia, cognitive distortions, and problem gambling in late adolescents. Cognit. Comput. 6, 652–660. Cosenza, M., Griffiths, M.D., Nigro, G., Ciccarelli, M., 2017. Risk-taking, delay discounting, and time perspective in adolescent gamblers: An experimental study. J. Gambl. Stud. 33, 383-395. Cosenza, M., Matarazzo, O., Baldassarre, I., Nigro, G., 2014b. Deciding with or without the future in mind: individual differences in decision-making, in: Bassis, S., Esposito, A., Morabito, S.B. (Eds.), Recent Advances of Neural Network Models and Applications. Springer, New York, pp. 435-443. Delfabbro, P., & King, D. (2017). Prevention paradox logic and problem gambling: Does low-risk gambling impose a greater burden of harm than high-risk gambling?. Journal of behavioral addictions, 6(2), 163-167. Dickerson, M. G., 1991. Internal and external determinants of persistent gambling, in: Heather, N., Miller, W.M., Greeley, J. (Eds.), Self-control and the Addictive Behaviours. Maxwell Macmillan, Sydney, pp. 317–338. Fossati, A., Krueger, R.F., Markon, K.E., Borroni, S., Maffei, C., 2013. Reliability and validity of the Personality Inventory for DSM-5 PID-5 predicting DSM-IV personality disorders and psychopathy in community-dwelling Italian adults. Assessment 20, 689-708. Fossati, A., Somma, A., Borroni, S., Markon, K.E., Krueger, R.F., 2015. The Personality Inventory for DSM-5 Brief Form: evidence for Reliability and construct validity in a sample of community-dwelling Italian adolescents. Assessment 24, 615-631. 23

Gainsbury, S.M., Suhonen, N., Saastamoinen, J., 2014. Chasing losses in online poker and casino games: Characteristics and game play of Internet gamblers at risk of disordered gambling. Psychiatry Res. 217, 220-225. Goodie, A.S., MacKillop, J., Miller, J.D., Fortune, E.E., Maples, J., Lance, C.E., et al., 2013. Evaluating the South Oaks Gambling Screen with DSM-IV and DSM-5 criteria: Results from a diverse community sample of gamblers. Assessment 20, 523-531. Goudriaan, A.E., Yücel, M., van Holst, R.J., 2014. Getting a grip on problem gambling: What can neuroscience tell us?. Front. Behav. Neurosci. 8, 1-12. Hopwood, C.J., Schade, N., Krueger, R.F., Wright, A.G.C., Markon, K.E., 2013. Connecting DSM5 Personality Traits and Pathological Beliefs: toward a Unifying Model. J. of Psychopathol. Behav Assess. 35, 162-172. Johansson, A., Grant, J.E., Kim, S.W., Odlaug, B.L., Götestam, K.G., 2009. Risk factors for problematic gambling: A critical literature review. J. Gambl. Stud. 25, 67–92. Kim, D.Y., Lee, J.H., 2011. Effects of the BAS and BIS on decision-making in a gambling task. Pers. Individ. Diff 50, 1131-1135. Krueger, R.F., Derringer, J., Markon, K.E., Watson, D., Skodol, A.E., 2012. Initial construction of a maladaptive personality trait model and inventory for DSM-5. Psychol. Med. 42 1879-1890. Krueger, R.F., Eaton, N.R., Clark, L.A., Watson, D., Markon, K.E., Derringer, J., et al., 2011. Deriving an empirical structure of personality pathology for DSM-5. J. Pers. Disord. 25, 170– 191. Krueger, R.F., Markon, K.E., 2014. The role of the DSM-5 personality trait model in moving toward a quantitative and empirically based approach to classifying personality and psychopathology. Annu. Rev. Clin. Psychol. 10, 477-501. Lesieur, H.R., 1979. The compulsive gambler’s spiral of options and involvement. Psychiatry 42 , 79-87. Lesieur, H.R., 1984. The Chase: Career of the compulsive gambler. Schenkman, Cambridge, MA. 24

Lesieur, H.R., Blume, S., 1987. The South Oaks Gambling Screen SOGS: A new instrument for the identification of pathological gamblers. Am. J. Psychiatry 144, 1184-1188. Linnet, J., 2013. The Iowa Gambling Task and the three fallacies of dopamine in gambling disorder. Front. Psychol. 4, 1-11. Linnet, J., Rojskjaer, S., Nygaard, J., Maher, B.A., 2006. Episodic chasing in pathological gamblers using the Iowa gambling task. Scand. J. Psychol. 4, 43-49. Lister, J.J., Nower, L., Wohl, M.J., 2016. Gambling goals predict chasing behavior during slot machine play. Addict. Behav. 62, 129-134. MacKillop, J., Miller, J.D., Fortune, E., Maples, J., Lance, C.E., Campbell, W.K., et al., 2014. Multidimensional examination of impulsivity in relation to disordered gambling. Exp. Clinical Psychopharmacol. 22, 176-185. McBride, O., Adamson, G., Shevlin, M., 2010. A latent class analysis of DSM-IV pathological gambling criteria in a nationally representative British sample. Psychiatry Res. 178, 401-407. Medeiros, G.C., Leppink, E.W., Yaemi, A., Mariani, M., Tavares, H Grant, J E., 2015. Electronic gaming machines and gambling disorder: A cross-cultural comparison between treatmentseeking subjects from Brazil and the United States. Psych. Res. 230, 430-435. Newman, J.P., Patterson, C.M., Kosson, D.S., 1987. Response perseveration in psychopaths. J. Abnorm. Psychol. 96, 145-148. Nigro, G., Cosenza, M., 2016. Living in the now: Decision-making and delay discounting in adolescent gamblers. J. Gambl. Stud. 32, 1191-1202. Nigro, G., Cosenza, M., Ciccarelli, M., 2017. The blurred future of adolescent gamblers: Impulsivity, time horizon, and emotional distress. Front. Psychol. 8, 1-12. Nigro, G., Ciccarelli, M., Cosenza, M., 2018. The illusion of handy wins: Problem gambling, chasing, and affective decision-making. J. Affect. Disord. 22, 256-259. Ochoa, C., Álvarez-Moya, E.M., Penelo, E., Aymami, M.N., Gómez-Peña, M., Fernández-Aranda, F., et al., 2013. Decision‐making deficits in pathological gambling: The role of executive 25

functions, explicit knowledge and impulsivity in relation to decisions made under ambiguity and risk. The Am. J. Addict. 22, 492-499. O’Connor, J. Dickerson, M., 2003. Definition and measurement of chasing in off-course betting and gaming machine play. J. Gambl. Stud. 19, 359–386. Oei, T.P.S., Lin, C.D., Raylu, N., 2008. The relationship between gambling cognitions, psychological states, and gambling: A cross-cultural study of Chinese and Caucasians in Australia. J. Cross Cult. Psychol. 39, 147-161. Oei, T.P.S., Raylu, N., 2015. Cognitive and psychosocial variables predicting gambling behavior in a clinical sample. Int. J. Ment. Health Addict. 134, 520-535. Parke, A., Harris, A., Parke, J., Goddard, P., 2016. Understanding within-session loss-chasing: an experimental investigation of the impact of stake size on cognitive control. J. Gambl. Stud. 32, 721-735. Quester, S., Romanczuk-Seiferth, N., 2015. Brain imaging in gambling disorder. Curr. Addict. Rep. 2, 220-229. Rahman, A.S., Balodis, I.M., Pilver, C.E., Leeman, R.F., Hoff, R.A., Steinberg, M.A., et al., 2014. Adolescent alcohol-drinking frequency and problem-gambling severity: adolescent perceptions regarding problem-gambling prevention and parental/adult behaviors and attitudes. Subst. Abus. 35, 426-434. Rash, C.J., Weinstock, J., Van Patten, R., 2016. A review of gambling disorder and substance use disorders. Subst. Abuse Rehabil. 7, 3-13. Raylu, N., Oei, T.P.S., Loo, J.M., Tsai, J.S. 2016. Testing the validity of a cognitive behavioral model for gambling behavior. J. Gambl. Stud. 32, 773-788. Rogers, R.D., Owen, A.M., Middleton, H.C., Williams, E.J., Pickard, J.D., Sahakian, B.J., et al., 1999. Choosing between small, likely rewards and large, unlikely rewards activates inferior and orbital prefrontal cortex. J. Neurosci. 19, 9029-9038. Ryan, T.P., 1997. Modern regression methods. Wiley, New York. 26

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. Stinchfield, R.,2002. Reliability, validity, and classification accuracy of the South Oaks Gambling Screen (SOGS). Addict. Behav. 27, 1-19. Tabachnick, B.G. Fidell, L.S., 2013. Using Multivariate Statistics, 6th ed. Pearson, Boston. Tackett, J.L., Krieger, H., Neighbors, C., Rinker, D., Rodriguez, L., Edward, G., 2017. Comorbidity of alcohol and gambling problems in emerging adults: A bifactor model conceptualization. J. Gambl. Stud. 33, 131-147. Taylor, R.N., Parker, J.D., Keefer, K.V., Kloosterman, P.H., Summerfeldt, L.J., 2014. Are gambling related cognitions in adolescence multidimensional?: Factor structure of the gambling related cognitions scale. J. Gambl. Stud. 30, 453–465. Toce-Gerstein, M., Gerstein, D, Volberg, R., 2003. A hierarchy of gambling disorders in the community. Addiction 98, 1661-1672. Worhunsky, P.D., Potenza, M.N., Rogers, R.D., 2017. Alterations in functional brain networks associated with loss-chasing in gambling disorder and cocaine-use disorder. Drug Alcohol Depend. 178, 363-371.

27

Figure 1. PID-5-BF trait domains scores as a function of chasing decision controlling for gender, age, chasing condition, alcohol consumption, and SOGS scores. Error bars indicate standard error of the mean.

1,6 1,4 1,2

MEAN

1 0,8 NONCHASER 0,6

CHASER

0,4 0,2 0 NEGATIVE AFFECT

DETACHMENT ANTAGONISM DISINHIBITION PSYCHOTICISM PID-5-BF DOMAINS

28

Table 1 - Descriptive statistics (means and standard deviations) as a function of gender. Males Mean SOGSa

Females SD Mean

Total sample

SD Mean

SD

5.55

5.13

2.44 2.99

4.71

4.84

Negative Affect

1.29

0.63

1.16 0.49

1.25

0.59

Detachment

0.88

0.56

0.65 0.53

0.82

0.56

Antagonism

0.89

0.58

0.60 0.52

0.81

0.57

Disinhibition

0.95

0.74

0.87 0.75

0.93

0.74

Psychoticism

0.74

0.59

0.77 0.65

0.75

0.60

Chasing total score

8.47 10.56

3.76 7.70

7.20 10.06

AUDITc

5.45

4.71 5.11

5.25

PID-5-BFb

5.05

5.06

a

South Oaks Gambling Screen. Personality Inventory for DSM-5-Brief Form. c Alcohol Use Disorders Identification Test. b

29

Table 2. Sample characteristics as a function of gender and chasing condition and results of 2X2 univariate ANOVAs. Control condition

Loss condition

Univariate ANOVA

Males

Females

Males

Females

(N = 49)

(N = 14)

(N = 43)

(N = 20)

Gender effects Condition effects Mean Age

a

SD

Mean

SD

Mean

SD

Mean

SD

36.86 12.46 33.21 12.31 38.79 12.08 36.05 12.29

F1, 122

p

F1, 122

p

1.64

0.20

0.91

0.34

Education (in years) 10.94

3.86 11.71

3.29 11.28

4.46 12.85

3.54

2.11

0.15

0.84

0.36

SOGSa Total score

4.86

4.84

2.64

2.87

5.39

2.30

3.13

10.98

0.001

0.94

0.33

Chasing Total score

5.28

7.73

2.86

7.22 12.09 12.16

4.40

8.13

6.87

0.01

4.68

0.03

6.35

South Oaks Gambling Screen.

30

Table 3. Pearson’s correlation coefficients among variables.

2 3 4 1. SOGSa 0.497** 0.603** 0.358** PID-5-BFb 2. Negative Affect 0.545** 0.257** 3. Detachment 0.592** 4. Antagonism 5. Disinhibition 6. Psychoticism 7. Chasing total score 8. AUDITc Note. *p < 0.05; **p < 0.01. a South Oaks Gambling Screen. b Personality Inventory for DSM-5-Brief Form. c Alcohol Use Disorders Identification Test.

5 6 7 8 0.697** 0.531** 0.629** 0.341** 0.459** 0.486** 0.289** -

0.550** 0.516** 0.294** 0.522** -

0.378** 0.485** 0.381** 0.469** 0.460** -

0.092 0.177* 0.316** 0.413** 0.183* 0.257** -

31

Table 4. Partial correlations after controlling for gender and age.

2 3 4 1. SOGSa 0.488** 0.576** 0.321** PID-5-BFb 2. Negative Affect 0.539** 0.247** 3. Detachment 0.613** 4. Antagonism 5. Disinhibition 6. Psychoticism 7. Chasing total score 8. AUDITc Note. *p < 0.05; **p < 0.01. a South Oaks Gambling Screen. b Personality Inventory for DSM-5-Brief Form. c Alcohol Use Disorders Identification Test.

5 6 7 8 0.728** 0.561** 0.607** 0.361** 0.466** 0.531** 0.284** -

0.553** 0.544** 0.308** 0.528** -

0.366** 0.473** 0.353** 0.474** 0.474** -

0.101 0.223* 0.309** 0.407** 0.192* 0.259** -

32

Table 5. Results of hierarchical logistic regression analysis. Chasing total score B Gender

SE

Wald

-1.780 0.498 12.783

df

p

Odds ratio (95% CI)

1 0.000

2.235– 15.728

Age

0.046 0.019

5.589

1 0.018

1.008 – 1.087

Conditiona

0.438 0.452

0.941

1 0.332

0.266 – 1.564

Disinhibition

1.284 0.362 12.589

1 0.000

1.776 – 7.335

AUDITb

0.059 0.051

1 0.247

0.960 – 1.172

1.339

Model: χ2 = 41.96. Nagelkerke’s R2 = 0.386. Overall percentage accuracy rate = 77%. a Control vs. Loss. b Alcohol Use Disorders Identification Test.

33

Table 6. Summary of hierarchical linear regression analysis. Chasing total score R2 ΔR2 β

Variable B t p VIF Step 1 Gender -5.299 0.116 0.116 -2.725 -2.725 0.007 1.025 Age 0.019 0.264 0.264 0.792 1.019 a Condition 5.374 3.121 3.121 0.002 1.020 Step 2 Gender -4.679 0.324 0.208 -0.207 -2.735 0.007 1.029 Age 0.043 0.052 0.690 0.492 1.023 Condition 5.132 0.256 3.393 0.001 1.020 Disinhibition 1.244 0.458 6.103 0.000 1.007 Step 3 Gender -3.763 0.381 0.057 -0.167 -2.257 0.026 1.058 Age -0.038 -0.047 -0.594 0.554 1.195 Condition 4.794 0.239 3.290 0.001 1.025 Disinhibition 0.837 0.308 3.623 0.000 1.402 Detachment 1.086 0.301 3.316 0.001 1.600 Step 4 Gender -4.150 0.402 0.021 -0.184 -2.506 0.014 1.072 Age -0.028 -0.035 -0.445 0.657 1.202 Conditiona 4.384 0.219 3.020 0.003 1.045 Disinhibition 0.668 0.246 2.752 0.007 1.586 Detachment 0.829 0.230 2.391 0.018 1.841 Psychoticism 0.622 0.186 2.052 0.042 1.637 Note. B: unstandardized coefficient; ΔR2: R square change; β: standardized regression coefficient; VIF: Variance Inflation Factor. a Control vs. Loss.

34