UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An exploratory study of the stopper device in modern slot machines Chu, Stephanie Wing-Man 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_september_chu_stephanie.pdf [ 1.33MB ]
Metadata
JSON: 24-1.0308678.json
JSON-LD: 24-1.0308678-ld.json
RDF/XML (Pretty): 24-1.0308678-rdf.xml
RDF/JSON: 24-1.0308678-rdf.json
Turtle: 24-1.0308678-turtle.txt
N-Triples: 24-1.0308678-rdf-ntriples.txt
Original Record: 24-1.0308678-source.json
Full Text
24-1.0308678-fulltext.txt
Citation
24-1.0308678.ris

Full Text

    AN EXPLORATORY STUDY OF THE STOPPER DEVICE IN MODERN SLOT MACHINES   by   Stephanie Wing-Man Chu   B.A., The University of Waterloo, 2014      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF ARTS  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Psychology)     THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)       August 2016  © Stephanie Wing-Man Chu, 2016     ii Abstract The stopper device is a common feature on modern slot machines that enables players to brake the spinning reels manually, but with no actual influence on the likelihood of winning. This study explored two potential mechanisms for why players might use the stopper device: (1) by increasing an ‘illusion of control’, and (2) by increasing the speed of game play. Thirty student participants from UBC and 32 community participants from the Greater Vancouver area participated in the study. Unlike previous research on the effects of the stopper device (e.g. Ladouceur & Sévigny, 2005), participants played a real slot machine (instead of pre-programmed slot machine simulator) equipped with a stopper, and were able to decide when and how often they used the stopper. Physiological arousal (heart rate and skin conductance) were monitored during game play. Participants completed the Gamblers’ Beliefs Questionnaire as a measure of trait susceptibility to the illusion of control, and after the game session, participants provided state ratings of the effectiveness of the stopper (i.e. illusory control) and their game play experience. Contrary to predictions, no significant relationships were observed (in either sample) between stopper use and the illusion of control. Stopper use was associated with faster speed of play, as measured by the spin initiation latencies. In addition, wins on trials where the stopper was used tended to increase ongoing stopper use, consistent with reinforcement via operant conditioning. Wins elicited a significant increase in heart rate, but no connection was found between stopper use and physiological arousal. Overall, the pattern of data indicated that stopper use may be better explained in associative learning (reinforcement and speed of play) than by higher-level cognitive appraisals (illusion of control). By increasing speed of play, stopper device may increase gambling losses and harms, and policy implications for the regulation of the stopper feature are considered.    iii Preface The thesis is an original work done by the author, Stephanie Chu. The study was performed at the Centre for Gambling Research at UBC. I am responsible for the study design, study material preparation, data analysis and writing. None of the work from this thesis has been published elsewhere, but preliminary data were presented as posters at the Alberta Gambling Research Institute Conference in Banff, AB in March 2015, and the New Horizons in Responsible Gambling Conference in Vancouver, BC in February 2016.  My supervisor, Dr. Luke Clark, provided continuous guidance and inputs to this project. A fellow graduate student at the Centre, W. Spencer Murch, provided technical support on the set-up of some of the study apparatus. A post-doctoral fellow at the Centre, Dr. Eve Limbrick-Oldfield, provided statistical support on the regression models in the statistical software package SAS. A research assistant, Angela Choi, helped with the data collection from January to June 2015.  Study 1 and 2 were approved by the UBC Behavioural Research Ethics Board under the study title “Casino Lab” (H14-02509).    iv Table of Contents Abstract ........................................................................................................................................... ii Preface............................................................................................................................................ iii Table of Contents ........................................................................................................................... iv List of Tables ................................................................................................................................ vii List of Figures .............................................................................................................................. viii Acknowledgements ........................................................................................................................ ix Chapter One: Introduction .............................................................................................................. 1 Harms of Slot Machines...................................................................................................... 1 Structural Characteristics of Slot Machines ........................................................................ 2 The Stopper Device............................................................................................................. 5 Speed of Play. ......................................................................................................... 8 Stopper and Speed of Play. ......................................................................... 9 Psychophysiology of Gambling. ........................................................................... 11 The Current Study ............................................................................................................. 14 Research Goal 1. ................................................................................................... 15 Hypothesis 1a. ........................................................................................... 16 Hypothesis 1b............................................................................................ 16 Research Goal 2. ................................................................................................... 16 Hypothesis 2.............................................................................................. 16 Research Goal 3. ................................................................................................... 17 Hypothesis 3.............................................................................................. 17 Chapter Two: Method ................................................................................................................... 18 Study 1 .............................................................................................................................. 18 Participants. ........................................................................................................... 18 Apparatus. ............................................................................................................. 19 Slot Machine. ............................................................................................ 19 Psychophysiological Measures. ................................................................ 22 Subjective Measures. ................................................................................ 23 Canadian Problem Gambling Index (CPGI). ................................ 23 Gamblers’ Beliefs Questionnaire (GBQ). ..................................... 23  v Game Experience Questionnaire (GEQ). ...................................... 24 Design. .................................................................................................................. 25 Procedures. ............................................................................................................ 25 Study 2 .............................................................................................................................. 27 Participants. ........................................................................................................... 27 Stopper Use in Real Casinos. ................................................................................ 29 Data Reduction...................................................................................................... 29 Event Marking. ......................................................................................... 29 Event Marking Apparatus. ............................................................ 30 Steps for Time-locking Events...................................................... 30 Spin Initiation Latency Data. .................................................................... 31 Heart Rate Analysis. ................................................................................. 32 Skin Conductance Analysis. ..................................................................... 32 Statistical Analysis. ............................................................................................... 33 Regression Model for Speed of Play Data. ............................................... 34 Regression Model for Heart Rate Data. .................................................... 35 Chapter Three: Results .................................................................................................................. 36 Study 1 .............................................................................................................................. 36 Stopper Presses. .................................................................................................... 36 Subjective Measures. ............................................................................................ 36 Gamblers’ Beliefs Questionnaire (GBQ). ................................................. 36 Game Experience Questionnaire (GEQ). .................................................. 37 Study 2 .............................................................................................................................. 38 Stopper Presses. .................................................................................................... 38 Slot Machine Outcomes. ....................................................................................... 39 Real-world Stopper Usage. ....................................................................... 41 Gamblers’ Beliefs Questionnaire (GBQ). ................................................. 42 Game Experience Questionnaire (GEQ). .................................................. 42 Do Wins Reinforce Stopper Use? ......................................................................... 44 Spin Initiation Latency. ......................................................................................... 45 Heart Rate Data. .................................................................................................... 46 Chapter Four: Discussion .............................................................................................................. 48 Proposed Mechanism 1: Illusion of Control ..................................................................... 49 Proposed Mechanism 2: Increasing Speed of Play ........................................................... 54 Psychophysiological Responses to Slot Machine Outcomes ............................................ 56 Limitations ........................................................................................................................ 57 Conclusions and Implications ........................................................................................... 59  vi References ..................................................................................................................................... 61 Appendices .................................................................................................................................... 69 Appendix A: Abridged Canadian Problem Gambling Index ............................................ 69 Appendix B: Gamblers’ Beliefs Questionnaire ................................................................ 72 Appendix C: Game Experience Questionnaire ................................................................. 74 Appendix D: Stopper Usage in Real Casinos Questionnaire ............................................ 75 Appendix E: Experimenter’s Instructions and Script ....................................................... 76 Appendix F: Conversion Chart for Bonus Remuneration................................................. 79     vii List of Tables Table 1         Participant information for Study 1 ......................................................................... 19 Table 2         Sample items of the illusion of control scale of the GBQ ....................................... 24 Table 3         Four supplementary illusion of control questions added to the GEQ ...................... 25 Table 4         Participant information for Study 2 ......................................................................... 28 Table 5         Two new questionnaire items regarding stopper use in real casinos ....................... 29 Table 6         Correlations between the seven GEQ subscales and stopper presses for Study 1 ... 37 Table 7         The type, number and proportion of outcomes for each participant in Study 2 ...... 39 Table 8         Average GBQ Illusion of Control scores in Study 1 and 2 ..................................... 42 Table 9         Correlations between the seven GEQ subscales and stopper presses for Study 2 ... 42 Table 10       Average GEQ scores in Study 1 and 2 .................................................................... 43 Table 11       Fixed effect analysis for latency time/speed of play. ............................................... 45 Table 12       Fixed effect analysis for heart rate. .......................................................................... 46    viii List of Figures Figure 1         Screenshot of paylines of the modern slot machine used in this study. ................... 5 Figure 2         Dragon’s Fire slot machine used in this study........................................................ 20 Figure 3         The shapes of the 9 paylines. .................................................................................. 21 Figure 4         Spin initiation latency illustration. ......................................................................... 32 Figure 5         Histogram of log-transformed stopper presses for Study 1 .................................... 36 Figure 6         Histogram of stopper presses for Study 1............................................................... 36 Figure 7         Histogram of log-transformed stopper presses for Study 2 .................................... 38 Figure 8         Histogram of stopper presses for Study 2............................................................... 38 Figure 9         The interaction between “stopper/no-stopper wins” and “time”. ........................... 44 Figure 10       Differences in spin initiation latencies following the four slot machine outcome                           types, from the Fixed Effect analysis. .................................................................... 46 Figure 11       Differences in heart rate following the four slot machine outcome types, from the                         Fixed Effect analysis. ............................................................................................. 47     ix Acknowledgements  I would like to thank my supervisor, Dr. Luke Clark, for his insights, continuous support and guidance on this project. I thank W. Spencer Murch for his technical support during the set-up of this study, Angela Choi for her help in data collection, and Dr. Eve Limbrick-Oldfield for her statistical expertise and support. I am grateful for the intellectual inputs and emotional support from Juliette Tobias-Webb, Becci Griggs, Caylee-Britt Goshko, Dr. Tilman Lesch, Dawn Kennedy, Mario Ferrari and Kent MacDonald. I also thank Dr. Catharine Winstanley and Dr. E. David Klonsky for agreeing to be on my MA Committee.    1 Chapter One: Introduction Slot machines are a common type of electronic gaming machine (EGM), a popular form of gambling among consumers, and one of the most profitable forms of gambling for gambling operators (MacLaren, 2015). From 2005 to 2010, Canadian revenues from casino gambling (including casino-based EGMs) as well as slot machines located outside casinos (e.g. in bars) increased steadily. These two sources contributed 34 and 21% respectively to the total government gambling revenue. Of the revenue from casinos, 93% came from slot machines alone. From these figures, slot machines are estimated to contribute over half of the total gambling revenues in Canada (BC Ministry of Health, 2013). Slot machines are also highly accessible: in British Columbia, there are 22,634 licensed slot machines (Problem Gambling Institute of Ontario, 2011), available not only in casinos, but also in other gambling venues like bingo halls.   Harms of Slot Machines Slot machine gambling is often described as an especially problematic and addictive form of gambling. A report from the Ontario Problem Gambling Hotline (2011) revealed that the vast majority of calls to the helpline were from gamblers concerned about their slot machine play; far more concerns were expressed about slots as compared to any other form of gambling activity. An evaluation of the Voluntary Self Exclusion (VSE) Program (a program that enables gamblers to bar themselves from gambling venues throughout the province) found that between 2006 and 2010, 83% of the surveyed VSE enrollees reported a problem with slot machine gambling (BC Ministry of Health, 2013). Similar evidence was presented recently that 80% of the VSE enrollees named slot machines as their chosen game to play (Cohen, 2015).   2 Slot machine gaming is closely associated with the development of problem gambling. Breen and Zimmerman (2002) studied the development of gambling problems in those who gambled primarily on EGMs compared to those who gambled primarily on more “traditional” forms of gambling, like card games. Breen and Zimmerman (2002) measured (retrospectively) the time taken by these two groups to develop their gambling problems, defined as the interval between the age of regular gambling involvement and the age at which the DSM-IV criterion for pathological gambling were first met. EGM gamblers progressed from initiation to problem gambling in an average of 1.08 years, whereas the traditional gamblers took an average of 3.58 years. MacLaren (2015) studied how the different forms of gambling contributed to problem gambling in Canada. His data showed that EGM players spent more money than players of other forms of gambling. EGM players also experienced elevated risk of problem gambling if they endorsed playing on a weekly basis, spending $50 or more in a typical session, or playing for more than an hour at a time. These data further highlighted the danger of prolonged game play on slot machines.   Structural Characteristics of Slot Machines To investigate why slot machines are relatively so addictive, it is important to study the psychological properties of slot machines, termed “structural characteristics”. Griffiths (1999) made a distinction between situational and structural characteristics of gambling activities. Situational characteristics refer to environmental elements external to the game itself, which get people to gamble in the first place. Some examples include the location, number and accessibility of gambling facilities, and casino advertising. In contrast, structural characteristics are features that are inherent to the gambling game itself. They play a key role in reinforcement, and are  3 likely to contribute to the development and maintenance of excessive gambling (Griffiths, 1993; as cited in Dowling, Smith, & Thomas, 2005, p.39). Structural characteristics may contribute to problem gambling by modifying behaviour through reinforcement learning principles, and leading to continuous behaviours like loss chasing (Griffiths, 1993; as cited in Dowling, Smith, & Thomas, 2005, p.39).   Early generations of slot machines, popular throughout much of the twentieth century, contained three mechanical reels painted with the classic array of symbols, e.g. cherries, “BAR” symbols and “Blazing red sevens”. These machines had a lever on the side that the player pulled to spin the reels. The mechanical structure of these early machines was relatively simple, involving only a few structural characteristics. Since then, technological advancements have greatly enhanced the game features on modern slot machines, making the games more exciting and attractive to gamblers. Modern slot machines now consist of multiple animated reels with symbols and bonus games of a certain theme, accompanied by intense (computerized) sound and visual effects (Dixon et al., 2010). In fact, modern slot machines contain the greatest number of structural characteristics as compared to other forms of gambling (Dowling, Smith, & Thomas, 2005).   Structural characteristics may fuel gambling-related cognitive distortions. These distortions describe a variety of biases in the perceived likelihood of winning, which subsequently encouraged gamblers to continue their gambling habits (Clark, 2010). Research in problem gambling treatments has shown efficacy in cognitive therapies that aimed at correcting these erroneous beliefs (Ladouceur et al., 2001; Ladouceur et al., 2003). The illusion of control is  4 a classic cognitive distortion, defined as “an expectancy of a personal success probability inappropriately higher than the objective probability would warrant” (Langer, 1975, p.313). Typically, people develop an illusion of control when elements from games of skill (e.g. sports) are introduced to a game of chance. This was first demonstrated in a series of experiments in Langer’s (1975) seminal paper on the illusion of control. For instance, participants who were allowed to choose their own lottery ticket would sell the ticket at a higher price, compared to those who were simply given the ticket, despite of the random nature of lottery outcomes. In addition to personal choice, other skill-related components like competition, practice and personal involvement have also been shown to increase one’s confidence of success in games of chance. The illusion of control has also been found to influence gamblers’ betting behaviours in naturalistic settings: Davis, Sundahl and Lesbo (2000) studied the behaviours of craps players, gambling with their own money. In craps, players take turns to throw dice, trying to roll certain desirable numbers. Players tended to bet more, place higher bets and place more “difficult” bets (e.g. on a specific number, rather than a spread of numbers) on their own dice throws, compared to their behaviour on other players’ throws. Similar data were seen in roulette, where laboratory participants took higher monetary risks when they were able to personally throw the ball onto the wheel, compared to when the ball was thrown by the experimenter (Ladouceur & Mayrand, 1987).   At first glance, slot machines may appear to be more passive than these other forms of gambling, as players are merely sitting in front of the machine and pressing the same button (or pulling the side lever) repeatedly. Yet, modern slot machines now consist of several illusion of control devices that are likely to increase gamblers’ sense of involvement in the game. The more  5 actively involved a gambler is with a gambling game, the more likely they are to believe that they are exercising skill, and that their actions have an effect on gambling outcomes (Griffiths, 1993). For example, most modern games allowed gamblers to customize their bet sizes by choosing both the number of paylines they want to bet on, and by varying the bet amount on each payline (see Figure 1 for an illustration of paylines on a modern slot machine). In some games, gamblers can bet on up to 20 paylines, as compared to the single horizontal payline in the classic machines. Many machines also have bonus rounds that resemble computer games. These bonus games can require motor dexterity (i.e. skill) but that dexterity only affects outcomes within the bonus game and has no effect on monetary wins (Harrigan & Dixon, 2009).    Figure 1. Screenshot of paylines of the modern slot machine used in this study. The Stopper Device  The stopper button is another common device on modern slot machines, putatively linked to illusion of control. The stopper allows the gambler to brake the spinning reels manually. This behaviour has no impact on the slot machine outcomes (e.g. wins and losses), which are pre-determined and randomized. As such, the use of the stopper merely causes the spinning reels to arrive at the predetermined outcome more quickly (Harrigan & Dixon, 2009). In a game of slots without using the stopper, the gambler presses the play button to make the reels spin, then waits for the reels to stop to reveal the outcomes. By choosing to use the stopper, the gambler can now  6 actively influence when the outcome appears, instead of being a passive agent who waits for the machine to reveal the outcome. Players may further believe that pressing the stopper influences the final positions of the reels and therefore the type of outcomes received. In this way, the use of the stopper can potentially lead to the development of an illusion of control, where the gambler believes s/he can manipulate randomized machine outcomes and become more skillful at the game. This may increase game harms, because exercising “skill” is a factor that contributes to excessive play (Griffiths, 1993).   The relationship between the illusion of control and the use of the stopper device was studied by Ladouceur and Sévigny (2005), in an experiment using a pre-programmed slot machine simulator. In their Study 1, undergraduate participants played 30 spins on the game without using the stopper, and then played another 30 games where they were asked to use the stopper on every trial. As in commercial slot machines, using the stopper did not influence the game outcomes. Results from a debrief questionnaire revealed that use of the stopper increased a range of erroneous beliefs; for example, 87% believed that the symbols could differ if the stopper was used (which is incorrect), and 57% believed they could control game outcomes by pressing the stopper (also incorrect). Participants who endorsed the idea that skill played a role in the slots game were more likely to believe that the stopper improved their chance of winning. Thus, these data suggest that using the stopper device fuelled the illusion of control, by enhancing the skill component to a game of chance. Study 2 in Ladouceur and Sévigny’s paper used a between-groups design to measure how the use of stopper influenced gambling persistence. The group who were instructed to use the stopper device played twice as many spins as the group who were  7 not able to use the device. This important study demonstrated the link between stopper usage and both the cognitions and behaviour (i.e. persistence) of gamblers.   One shortcoming of the Ladouceur and Sévigny (2005) study was a lack of ecological validity. They employed a pre-programmed slot machine simulator and the participants were told when to use, and when not to use, the stopper. These conditions were not very realistic because in real casinos, slot machine outcomes are random, and gamblers are not told how they should play the game. It is also unrealistic to think that gamblers who use the stopper will do so without fail, on every spin. Rather, in natural slots play, gamblers may use the stopper some of the time, exploring the impact that it has on the frequency of winning. The two experiments by Ladouceur and Sévigny (2005) study might not be able to capture participants’ natural patterns of using the device. As a second point, the gamblers tested by Ladouceur and Sévigny (2005) were university students. Gambling behaviours in university students tend to be quite different from community-recruited samples of gamblers. Community gamblers tend to engage in more forms of gambling, possibly due to differences in disposable household income (Marmurek, Switzer, & D’Alvise, 2014). Differences also exist in motivations for gambling (Marmurek, Switzer, & D’Alvise, 2014). Therefore, there are reasons to believe that what was found in the student sample may not be generalizable to the community gamblers.   Another study conducted by Nastally, Dixon and Jackson (2009) attempted to look at how the stopper device and the machine win rate would affect gamblers’ preference on slot machines. The study used two slot machine simulators: one with and one without a stopper device. The win rate was manipulated as either low (10%) or high (80%). Participants preferred  8 the simulators with the 80%-win rate, regardless of whether the game was equipped with a stopper device or not. Nastally and colleagues concluded that the win rate had a stronger effect that the stopper device in determining game preferences. However, their conclusions are limited by the use of graduate students as their sample, and also from a likely demand characteristic arising from the extreme difference in win rate (10% vs. 80%). This salient difference in win rate may have obscured any influence of the stopper device.  Speed of Play. Continuous forms of gambling, such as slot machines and roulette, involve a short interval between the gambler placing their bet and the outcome being revealed (Dickerson, 1993). Following the outcome, it is immediately possible for the gambler to place another bet. In contrast, “discontinuous” forms of gambling (such as lotteries) impose a longer interval between bet and outcome, and also it may not be possible for the player to gamble again in succession. These structural characteristics related to the timing of the game imbue continuous forms with a much higher speed of play, and this may be associated with the increased risk of problematic gambling that is often associated with continuous forms (Griffiths, 1993). Modern EGMs are regarded as the form of (continuous) gambling that has the highest speed of play. Each spin on a slot machine takes approximately 3 to 5 seconds, and regular players play at a rate that is close to this maximum capacity (as cited in Dowling, Smith, & Thomas, 2005, p.40). Harrigan and Dixon (2009) studied the speed of play on two popular modern slot machines (“MoneyStorm” and “Lobstermania”) and found that by initiating a spin every 3 seconds, many players will achieve close to 1,200 spins per hour.    9 Several studies have looked at the impact that speed of play has on the relationship between slot machines and problem gambling. Problem gamblers play slot machines at a significantly higher speed of play, relative to non-problem gamblers (as cited in Dowling, Smith, & Thomas, 2005, p.40; Linnet et al., 2013), and speed of play correlates with problem gambling severity on the South Oaks Gambling Screen (SOGS) (Linnet et al., 2013). Ladouceur and Sévigny (2006) compared groups of players assigned to high-speed and low-speed conditions on gambling persistence. They found that gamblers in the high speed condition played 2.5 times more spins than those in the low speed condition, and were more likely to underestimate the number of spins they had played. Loba et al. (2001) manipulated two sensory features (i.e. slow/fast speed of play and sound on/off) on a standard slot machine. A group of participants with problem gambling tended to persist more in the “fast play speed and sound on” condition. The combined effects of slowing the speed and turning off the sound lowered ratings of enjoyment, excitement and tension-reduction for problem gamblers. A separate study looking at speed of play alone found that a slower game negatively affected the enjoyment and satisfaction of recreational gamblers (as cited in Dowling, Smith, & Thomas, 2005, p.40).   Stopper and Speed of Play. Besides any putative effect on the illusion of control, the use of the stopper device in the middle of a spin will inherently act to shorten the spin latency, and speed up the reveal of the spin outcomes. In this way, players can influence the timing of the game, and thereby regulate their speed of play. Harrigan and Dixon (2009) tested whether stopper use increased the speed of play on slot machines, showing that use of the stopper sped up the game by up to 50%; the player could spin every 1.5 seconds, instead of every 3 seconds when the stopper was not used. Participants from the Loba et al. (2001) study reported that they would  10 be more likely to play slots if they can stop the reels themselves. Most studies looking at speed of play in slot machines have used a simple measure, dividing the total duration of play time by the number of spins. However, this conflates two temporal components. First, the spin duration is directly and inherently reduced by using the stopper device; based purely on mathematics (not psychology), the player can make more spins in a fixed period of play. However, the player can also control the interval between the spin outcome being revealed, and the initiation of the next spin. Some studies have investigated how this latency changes as a function of the outcome on the previous spin, known as the post-reinforcement pause (PRP). Post-reinforcement pauses (PRPs) were noted in early animal studies in operant chambers (“Skinner boxes”), where animals responding on unpredictable schedules would pause after reward delivery, before engaging in the next operant response (Felton & Lyon, 1966). In the context of slot machine gambling, this phenomenon is observed in the latency between a win and the initiation of the next spin.   Both animal (Peters, Hunt, & Harper, 2010; Scarf et al., 2011) and human (Dixon et al., 2013) studies have shown that subjects demonstrated PRPs following appetitive, winning outcomes compared to losing outcomes. The length of the PRP was also positively correlated with the size of the win (Dickerson et al., 1992; Delfabbro & Winefield, 1999). It is possible that when a player receives a win, ongoing behaviour is inhibited for a consummatory period as the player internally “celebrates” the win (Dixon et al., 2013). This pause causes the delay before the initiation of the next spin.   In addition to demonstrating the presence of PRPs for wins compared to losses, Dixon et al. (2013) looked at the spin initiation times following some other outcomes. (The term “PRP” is  11 not technically appropriate here because there is no reinforcement involved). Their study revealed some interesting latencies results regarding near-misses, a type of outcome that closely resembles wins but fall just short. Latencies were shorter following near-misses, not only compared to the latency after a win, but also compared to the latency after regular losses. Dixon et al. (2013) interpreted this very rapid initiation latency as an indication that near-misses were unpleasant and frustrating; even more so than regular losses, mostly due to the fact that they frustrate the gamblers’ goal of winning big money (especially when they are “so close” to achieving the goal). Thus, spin initiation latencies can be used to measure gamblers’ urgency to continue playing, in a way that is further sensitive to the different game outcomes.   Psychophysiology of Gambling. The measurement of physiological responses has provided important information about the psychological basis of gambling. Gambling activities trigger robust changes in physiological arousal: games like blackjack, horse racing betting and slot machine gambling elicit elevated levels in heart rate, skin conductance and cortisol (e.g. Anderson & Brown, 1984; Coventry & Norman, 1997; Meyer et al., 2000; Coventry & Hudson, 2001). Similar results have been obtained with the use of gambling tasks in controlled laboratory settings (e.g. Sharpe et al., 1995; Ladouceur et al., 2003), where real monetary reinforcement is available. To date, most psychophysiological studies of gambling have examined tonic changes; i.e. averaged levels of physiological responses across several minutes of gambling. For instance, Moodie and Finnigan (2005) monitored the participants’ heart rate while they played slots in an actual gambling venue. They reported higher tonic heart rate in test session where a jackpot win was experienced, and this change in heart rate was greater for wins of larger sizes. Despite the interesting findings, this focus on tonic measures is problematic, and especially so for research  12 on slot machine gambling. Real slot machine play involves spins being made every 3 to 6 seconds (Dixon et al., 2011), and the outcomes they get from each spin occur randomly. Therefore, a variety of events may happen within a 60 second window of gambling: a gambler could have had a long streak of losses, increased their bet size, or experienced a major jackpot win or bonus game. The distinctiveness of these events would confound tonic physiological responses (Lole et al., 2012) and if the physiological responses are averaged across a minute or longer, it would be impossible to isolate individual gambling outcomes.   For these reasons, measurement of “event-related” or phasic responses may be much more informative, but only a small number of studies have sought to do this. Goudriaan et al. (2006) investigated the physiological responses during the Iowa Gambling Task; a neuropsychological test that involves risky choices, and gain and loss outcomes, but with only limited resemblance to true gambling. Both problem gamblers and healthy controls experienced a decrease in heart rate after loss outcomes. After wins, healthy controls showed an increase in heart rate while problem gamblers experienced a slight decrease in heart rate. Goudriaan et al. (2006) attributed the blunted heart rate increase to lowered reward sensitivity in problem gamblers. In slot machine gambling, other structural characteristics, like bonus features, were found to be highly arousing. Dixon et al. (2010) studied phasic heart rate changes using heart rate deceleration, a measure derived by comparing the inter-beat intervals (IBIs) immediately before and after the gambling outcome is revealed. Heart rate deceleration, widely considered to be an orienting response to surprising stimuli, was more pronounced following wins than losses. This is consistent with an orienting response as wins are infrequent outcomes.   13 Other studies have looked at phasic changes with regard to skin conductance. Skin conductance is regarded as the “gold standard” for measuring arousal levels due to its reliability and high sensitivity (e.g. Lole et al., 2012, p.155). Generally speaking, wins elicit higher skin conductance levels as compared to losses (Dixon et al., 2010; Lole et al., 2012; Wilkes, Gonsalvez, & Blaszczynski, 2010). The increase in skin conductance was also proportionate to the size of the win outcomes, with large wins producing larger skin conductance than small wins (Lole et al., 2012). Near-miss outcomes (e.g. where two cherries are present on the payline, but with the third cherry narrowly missing) were also shown to increase skin conductance responses to a level that was higher than full misses (Dixon et al., 2011; Clark et al., 2012) and in one study even higher than for actual wins (Dixon et al., 2011), demonstrating the further significance of these events to gamblers.    Psychophysiological arousal is thought to play an important role in maintaining slot machine gambling. Classic accounts of gambling describe arousal as a key reinforcer of gambling behaviour, one which may even be more important than the monetary wins (Brown, 1986). However, the mechanism by which changes in arousal increase gambling is unclear. Interpreting arousal as a straightforward correlate of subjective excitement, Boyd (1982) proposed that arousal itself was the “gambler’s drug”. Individuals who are frequent/problem gamblers show greater increases in physiological arousal (Moodie & Finnigan, 2005) and subjective excitement (Linnet et al., 2010) during gambling sessions, compared to non-problem gamblers. Yet, changes in physiological arousal could also reflect negative emotional responses to the game, such as frustration. Following a frustrating task, participants’ tonic heart rate increased by over 20 beats per minute (as cited in Dixon et al., 2011); indeed, the increase in  14 arousal resulting from a frustration test was larger than that observed following positive events (as cited in Dixon et al., 2011). As such, frustration-induced arousal is a possible mechanism behind the effects of near-misses (Dixon et al., 2011).   An experiment by Clark et al. (2012) forged a link between a manipulation of personal control and the psychophysiological response to gambling outcomes. Their results showed that increases in arousal following near-misses were greater on trials on which the participant personally selected their icon on a simulated slot machine game, compared to trials without such personal control. Participants also gave higher ratings of their motivation to continue gambling on personal control trials. Clark et al. (2012) argued that physiological arousal was linked to the cognitive appraisal of illusory control in games of chance. An increase in personal control can potentially make a gambling game more exciting. In Dow Schüll’s (2012) book Addiction by Design, a slot machine player was quoted as saying that the classic kind of slot machines “didn’t draw [him] in” because he was just waiting for the outcomes to appear. However, the elements of choice and skill available in modern machines can “heightened players’ absorption” and turn slots into a “compelling and interactive involvement” (p.114). Therefore, one may hypothesize that the heightened game involvement and excitement arising from skill and choice (provided by the stopper device) elicits increases in psychophysiological arousal during the game.   The Current Study  This study had two main goals. First, it aimed at exploring two possible mechanisms underlying gamblers’ use of the stopper device, outlined above: (1) through cognitive distortions, specifically the illusion of control; and (2) by increasing the speed of play. I further set out to  15 examine whether these effects were mediated by increased physiological excitement during slot machine play.   Research Goal 1. I aimed to replicate and extend the experiments by Ladouceur and Sévigny (2005) regarding the relationship between stopper use and the illusion of control. The limited ecological validity in the Ladouceur and Sévigny (2005) study was improved by the use of a quasi-experimental design. Participants played an actual, commercially available slot machine with outcomes that were not preprogrammed and were fully randomized. Participants were not instructed on when to use the stopper; rather, they were able to explore the device and use it as often as they wanted. Through these measure, I attempted to approximate slot machine play in actual casinos.   I studied the relationship between stopper use and the illusion of control from three perspectives: trait characteristics, state-related ratings of the game, and behavioural changes. Questionnaires were used to measure both the trait characteristics and state-related responses. The trait measure, the Gamblers’ Beliefs Questionnaire (GBQ), investigated whether the individual was generally prone to having an illusion of control. The state-related measure was administered immediately after the period of game play, and investigated whether the individual experienced illusory control in response to the game itself. A behavioural analysis investigated whether the use of the stopper device changed following win outcomes on which the stopper had been used. That is to say, is the stopper response reinforced by successful outcomes? If the participants had a (distorted) cognitive inference about the stopper affecting their chances of  16 winning, this should be strengthened by a win, and visible as increased use of the stopper device after (relative to before) the win.  Hypothesis 1a. I hypothesized that higher illusion of control scores (based on either the trait questionnaire or state-related items) would predict increased use of stopper device during game play.  Hypothesis 1b. I hypothesized that participants would use the stopper device more in the 5 trials following a successful (stopper) win compared to the 5 trials preceding the win, and also compared to the trials either side of a winning outcome that did not involve the stopper device.   Research Goal 2. Use of the stopper device should also increase the average speed of play. I aimed to explore whether using the stopper would specifically influence the spin initiation latency. Through studying the spin initiation latencies, it is possible to investigate whether using the stopper would influence the players’ urgency to start new spins (and potentially to play more spins).    Hypothesis 2. I hypothesized that the spin initiation latencies found in this study would replicate previous PRP findings, with longer latencies following win outcomes as compared to loss outcomes, and that spin initiation latencies would be shorter on trials when the stopper was used.    17 Research Goal 3. I explored physiological excitement as a potential mediator of the two proposed mechanisms (illusion of control, speed of play) of stopper usage.   Hypothesis 3. I hypothesized that a higher number of stopper presses during game play would be positively related to physiological arousal, as measured by heart rate and skin conductance responses.   These hypotheses were explored in two studies. The first study was conducted in university students, who were mostly inexperienced slot machine users. Given the high degree of complexity of a modern multi-line slot machine, it is likely that my participants spent much of their session figuring out the basics of the game, rather than really playing the game. This would likely affect their tendency to explore novel aspects of the game, such as the stopper device. This could also undermine game enjoyment and immersion. In addition, most of the participants in Study 1 were non-problem gamblers (which fit well with the profile of university students) and this limited the scope of the study, as we could not draw any conclusions beyond novices and inexperienced players. In light of these issues, a similar Study 2 was done on community participants who were regular slot machine players.     18 Chapter Two: Method Study 1 Participants. Thirty student participants (15 males; 15 females) with a mean age of 22.6 took part in Study 1. All participants were recruited at the University of British Columbia Vancouver campus through poster advertisements. To be eligible for the study, the participants needed to: (1) be a UBC student, (2) be at least 19 years old (the legal age to gamble in the British Columbia), (3) not have experienced gambling problems currently or in the past, (4) not be taking any anti-depressant, anti-anxiety or anti-psychotic drugs and (5) not have any allergies to gels, adhesives or sanitizing agents (e.g. rubbing alcohol) that were used in the psychophysiology monitoring. Information concerning past experience playing slot machines was collected using the Canadian Problem Gambling Index (CPGI) [Section 1 Question g]: “In the past 12 months, how often did you bet or spend money on slot machines in a casino?” (Ferris & Wynne, 2001). Of our 30 participants, 7 had played slots at least once in the past year, and only 1 of these played regularly (6 to 11 times in past year). The other 27 participants had no experience playing slot machines over the past year (see Table 1).    Gambling status was assessed using the Problem Gambling Severity Index (PGSI), a section of the CPGI. Participants scoring 1 - 2 were considered “low-risk gamblers” i.e. with few negative consequences associated with their gambling. Scores of 3 to 7 were considered “moderate-risk gamblers”, and scores of 8 or above were considered “problem gamblers” (Ferris & Wynne, 2001). The PGSI was administered immediately following Consent, and by our ethical protocol, any participants scoring in the problem gambling range were discontinued from the study (with remuneration) and provided with information on resources for problem gambling.  19 In our sample of 30 participants, there were 21 non-problem gamblers, 6 low-risk gamblers and 3 moderate-risk gamblers. No problem gamblers were recruited in this sample.   Table 1.  Participant information for Study 1 Note. Age information was not available for 5 non-problem gamblers.  Apparatus. Slot Machine. The study was conducted using an actual slot machine, housed in our Casino Lab on campus. The slot machine was a modern multi-line slots game called “Dragon’s Fire” (Williams Interactive, WMS), which employed a dragon and knights theme (e.g. 3 purple dragon symbols are awarded bonus spins). The screen showed 5 animated reels and 3 horizontal lines (see Figure 2). There were 9 pay-lines (horizontal, diagonal, or a mix of both, spanning across reels 1 to 5 – see Figure 3). Past research has shown that gamblers playing multi-line slot machines typically use the “mini-max” strategy, placing the minimum bet on the maximum number of pay-lines (Harrigan et al., 2011). In light of this finding, the bet per spin in the current study was fixed based on the “mini-max” strategy: 1 credit on each of the 9 paylines. Each spin thus cost 9 credits. Three counters on the screen displayed the current bet size, credits paid on that spin, and a running balance of the credits remaining. The machine had a “Repeat Bet” button Group No. Age Males Females Slots playing frequency over the past 12 months 6-11 times 1-5 times Have not played slots Non-problem gamblers 21 22.7 9 12 0 4 17 Low-risk gamblers 6 20.0 4 2 1 1 4 Moderate-risk gamblers 3 27.7 2 1 0 1 2 Problem gamblers 0 - - - - - -  20 on the right hand side of the facia. This button was used both to initiate a new spin and as the stopper button. As participants were not permitted to modify their bet size in this experiment, this means that all behaviours during the slot machine session involved the player interacting with a single button.   Following an initial press on the “Repeat Bet” button, the reels immediately started to spin. In the absence of a stopper press, the reels would naturally stop in a left-to-right sequence, to reveal any winning lines to the player. If the participant pressed the button again during the spin, this acted as the stopper, and all 5 animated reels would stop immediately. If at least three identical winning symbols were shown on any paylines, there would be a credit payout. Payouts increased with more identical symbols, and also if there were identical symbols across multiple lines.   Figure 2. Dragon’s Fire slot machine used in this study.  21  Figure 3. The shapes of the 9 paylines. On any spin, the possible outcomes could be divided into 5 basic types (Nb. These types varied widely in frequency):  (1) losses occurred if the machine returned zero credits, and thus the player lost all bet credits. On losses, there were no visual or auditory effects as the reels stopped spinning. (2) regular wins occur when the credits won exceeds the number of credits bet. Wins were accompanied by a flashing square around the winning payline(s) as well as a winning jingle. The jingle played while the payout counter would count upwards, to reveal how many credits had been won; therefore, the jingle duration was proportionate to the size of the win. (3) loss-disguised-as-wins (LDWs) occur when the machine returns a (non-zero) amount of credits that is nevertheless less than the amount bet. Like regular wins, LDWs are accompanied by visual and auditory feedback. (4) bonus wins occurred in reaction to 3 (or more) specific “purple dragon” symbols, which could appear anywhere on the screen. On these rare outcomes, players were awarded 10 to 40 free spins depending on how the number of dragon symbols. Typically, players would receive some regular wins or LDWs from the bonus spins, but note that if a player  22 did not get any wins during this period, the outcome was re-classified as a loss. The bonus spins were initiated automatically by the machine, i.e. no initiation press was required from the player, but the use of the stopper button during the spin continued to be effective on bonus spins. (5) near-misses; the purple dragon ‘bonus spins’ enable a further type of near-miss event, if only two purple dragons were presented. Note that the delivery of each purple dragon was accompanied by a dramatic auditory flourish, with this flourish increasing in pitch to successive dragons. Thus, if only 2 dragons were revealed, the player was aware of having been close to getting a bonus win.  All outcomes were randomized by the machine, and thus the machine offers no experimental control over the frequency of each outcome type. In particular, some participants did not receive any bonus wins, and some participants experienced only a small number of regular wins. The payback percentage for the game was set at 90.09%; this is the percentage of bets returned to the player over the lifetime of the machine (Harrigan & Dixon, 2009), which thus showed considerable volatility between participants in a 20-minute session.   Psychophysiological Measures. Heart rate and skin conductance were collected using a Biopac systems MP150, and were viewed and analyzed using the Biopac AcqKnowledge 4.4 software. The Biopac amplified and digitized the physiological responses collected from the participants. To obtain heart rate, three disposable electrocardiogram (ECG) sensors were placed on the participants’ body: one at each shoulder (directly below the collarbones) and one on the lower left abdomen. For skin conductance, two disposable sensors were placed on the index and  23 ring fingers of the non-dominant hand. A pulse sensor (not analyzed) was placed on the middle finger of the non-dominant hand. The responses collected were sent to a laptop running AcqKnowledge 4.4, for real-time viewing and recording through two wireless transmitters, which the participants wore on their waist and wrist for the duration of the slot machine session.  Subjective Measures. Three self-report measures were used to collect information about gambling habits, illusion of control and game enjoyment (see Appendix A, B and C).   Canadian Problem Gambling Index (CPGI). The Problem Gambling Severity Index (PGSI) section of the CPGI (Ferris & Wynne, 2001) was used to assess gambling habits and any symptoms of problem gambling. The PGSI was developed using factor analysis, and item analysis in which redundant items were iteratively eliminated. This resulted in a final 9-item questionnaire. Factor analysis on the 9 items showed that each item has good factor loading (at least 0.5). Most items had moderate to strong correlations with items from the South Oaks Gambling Screen (SOGS), an early measure of problem gambling based on the DSM-III criteria. Ferris & Wynne (2001) showed that Cronbach’s alpha for the CPGI was 0.84 and test-retest Pearson r was 0.78, demonstrating high internal consistency within items and test-retest reliability respectively.  Gamblers’ Beliefs Questionnaire (GBQ). The GBQ (Steenbergh et al., 2002) was developed to assess gambling-related cognitive distortions, with two subscales representing “Luck” and “Illusion of Control”. My analyses focused on the GBQ Illusion of Control subscale as a trait-related measure of the tendency to over-estimate control in games of chance.  24 Participants rated 21 statements for the extent that they agreed/disagreed (see Table 2 for sample items). Steenbergh et al. (2002) performed a factor analysis to identify the two subscale domains and to examine the validity of the questionnaire. Items that scored below .30 in factor loading were dropped; the final 21-items all scored at least .35 on factor loading. Convergent validity of the GBQ was demonstrated through its relationship with gambling behaviours and problem gambling. Moderate Pearson r correlations were found between the average amount of time spent on gambling and the scores from the two GBQ subscales. Problem gamblers, identified by the SOGS, scored significantly higher than non-problem gamblers on the GBQ. Steenbergh et al. (2002) checked the relationship between social desirability response biases and GBQ scores, finding no significant relationship between the two variables. Cronbach’s alpha for the Luck and Illusion of Control subscales were .90 and .84 respectively, establishing internal consistency. Table 2.  Sample items of the illusion of control scale of the GBQ Sample item 1: “My knowledge and skill in gambling contribute to the likelihood that I will make money.” Sample item 2: “I think of gambling as a challenge.” Sample item 3: “My gambling wins are evidence that I have skill and knowledge related to gambling.”  Game Experience Questionnaire (GEQ). State-related emotional and cognitive responses to the slot machine were measured by the GEQ (IJsselsteijn et al., 2008). The GEQ was originally designed to captures 7 dimensions of video game play experience: Sensory and Imaginative Immersion, Tension, Competence, Flow, Negative Affect, Positive Affect, and Challenge. However, it has since been applied into slot machine gambling research (e.g. Templeton et al., 2015). IJsselsteijn et al. (2008) had taken steps to examine the test-retest reliability and construct validity. For example, they found that players’ covert behaviours (e.g. force exerted on a computer mouse) while gaming was positively correlated with the Flow  25 dimension. Limited data exists on formal psychometrics for the GEQ. The original GEQ was supplemented with four supplementary questions (adapted from Ladouceur and Sévigny, 2005), to investigate participants’ specific beliefs about the stopper device during the game play session (See Table 3 for the supplementary questions).  Table 3. Four supplementary illusion of control questions added to the GEQ 1. Do you believe that a player can influence the symbols on the screen after having activated the play button? If so, how? 2. Is there a method for controlling the outcome of the game after the play button has been activated? If so, what is that method? 3. If you were to obtain a winning combination, would it be due to chance, skill or a combination of the two? Why is that? 4. Are there any strategies that could enable you to increase your chance of winning after the play button has been activated? If so, what are these strategies?   Design. The current study used a quasi-experimental repeated-measures design. Participants were asked to play a real slot machine with randomized outcomes during the experiment. Hence, the design is quasi-experimental, in so far as it was not possible to predict nor manipulate the types or frequencies of outcomes that would be delivered on the machine. Each participant experienced a unique mix of outcomes, e.g. some of them had more wins than others. Instead of assigning participants to “stopper” and “no-stopper” conditions (as in Ladouceur & Sévigny, 2005), participants were free to use the stopper whenever and however many times they wanted. Thus, participants attended a single slot machine session, during which the stopper could be used for some (or even none) of the spins, depending on the participants’ preferences, and thus constituting a repeated-measures design.   Procedures. Respondents to the poster advertisement were contacted for a telephone screening interview to ensure eligibility, and then scheduled for a 1-hour session. On arrival at  26 the lab, participants gave informed consent, which included details on remuneration (i.e. $10 per hour plus bonus from the slots game, if any) and video-taping during study. Immediately after the consent process, each participant filled out the CPGI in the waiting room, and the PGSI section was scored visually by the experimenter to confirm the absence of problem gambling. Participants then entered the “Casino Lab” where our slot machines are located. The Casino Lab mimics the atmosphere of an actual casino, using comfortable casino stools and dim lighting. Four slot machines are arranged in a row, and the participants for this study were seated at a machine called “Dragon’s Fire”, as one of two machines in the lab equipped with a stopper device.    After attaching the psychophysiological equipment, the experimenter explained the basic rules of the slot machine (See Appendix E for the script), which included a description of the stopper device; participants were informed that they could use this feature whenever and however many times they wanted to. After making sure that the participants understood the game, the participant was given $40 (2 x $20 bills) to play the slot machine, and instructed to feed the money into the machine. The participant then played the machine at their own pace, for 20 minutes. The $40 endowment amount was chosen in order to sustain participants for 20 minutes based on the machine’s programmed rate of return. However, as the profits are randomized, a small number of participants did exhaust their credits before the 20-minute period was up. When this happened, the experimenter terminated the play session at the moment the credit balance fell below 9 (i.e. not enough credits to make a spin). For these participants, their full set of game play data was included in the analyses, and the statistical fixed regression model  27 was used to account for differences in the number of outcomes across each participant (see Statistical Analysis).    Following the slot machine session, the experimenter removed the psychophysiology sensors and participants completed the GEQ and GBQ. In addition to the $10 remuneration, a bonus payment between $2 and $12 was awarded based on the credits remaining on the slot machine, using a conversion chart (see Appendix F). Note that due to ethical reasons, we could award participants major jackpot outcomes from the machine, requiring the conversion chart. Before leaving the session, participants were debriefed and given some brochures on the mathematics of slot machines, and problem gambling help services in BC.   Study 2  Study 2 was identical to Study 1, with the following exceptions:   Participants. Thirty-two participants (17 males; 15 females) with a mean age of 44.1 took part in Study 2. Participants were recruited from Craigslist, an online advertisement platform in Vancouver. For eligibility, participants did not need to be UBC students, but needed to have played a slot machine at least once over the past 12 months. From similar experiments, we expected this inclusion criterion would actually capture a broad range of gambling involvement, from occasional players through to regular gamblers with some symptoms of problem gambling.   Gambling involvement and problem gambling symptoms were collected using the CPGI and PGSI (Ferris & Wynne, 2001). Over the past 12 months, 15 of 32 participants had played the  28 slot machine at least weekly, and 8 played on a monthly basis. In Study 2, participants scoring in the “problem gambler” range (8 or above) on the PGSI were not discontinued. Rather, we ensured that the participants were not currently seeking treatment or other means of quitting gambling, and that slot machine gambling was a regular behaviour that these participants were engaged in outside of the lab (for at least the past 12 months). In addition, resources for responsible gambling and information on various problem gambling helplines were provided during debrief. In the resultant sample of community-recruited regular gamblers, there were 2 non-problem gamblers, 7 low-risk gamblers, 14 moderate-risk gamblers, and 9 problem gamblers (see Table 4).   Table 4. Participant information for Study 2      Slots playing frequency over the past 12 months Group No. Age Males Females Daily 2-6 times/week Once a week 2-3 times/month Once/month 6-11 times 1-5 times Have not played slots Non-problem gamblers 2 39.0 0 2 - - - - - 1 1 - Low-risk gamblers 7 44.6 2 5 - 1 3 - 1 - 2 - Moderate-risk gamblers 14 40.3 11 3 - 3 3 2 1 3 2 - Problem gamblers 9 50.9 4 5 - 4 1 4 - - - -    As participants were recruited from the Greater Vancouver area to attend a session at UBC, I increased the remuneration from $10 to $15. The bonus payment was unchanged from Study 1.   29  Stopper Use in Real Casinos. I added 2 newly-devised items to ask participants about their real-world tendency to use the stopper device in their day-to-day slot machine play (see Table 5). The answers for stopper use frequency were coded as follows: “Rarely” was coded as “0”, “Sometimes” was coded as “1”, “Most of the time” was coded as “2” and “All the time” was coded as “3”. This questionnaire was completed after the GEQ and GBQ.  Table 5. Two new questionnaire items regarding stopper use in real casinos 1. When you visit gambling venues, do you play slot machines with the stop button function? 2. If yes, how often do you make use of the stop button function?  Data Reduction. I used Study 1 to explore the behavioural data, to develop and optimize the procedures for event marking, and to refine my hypotheses. As the UBC students’ limited experience with slot machines became evident during the running of Study 1, the analysis focusses only on the data from the subjective measures. Study 2 interrogated the behavioural and psychophysiological data in more detail.    For Study 2, data from one participant was not analyzed because he failed to follow the instructions. Heart rate data were not available from one other participant because of an allergic reaction to the ECG sensors (this condition was unknown to the participant prior to the study). This resulted in 31 participants on all measures except heart rate, where 30 participants were available.  Event Marking. Several techniques were devised to place event markers on the Biopac psychophysiology trace, in order to obtain measures of player behaviour (e.g. number of stopper presses) and synchronize that behaviour with the psychophysiological responses:   30  Event Marking Apparatus. The first challenge was to record player behaviour in terms of presses on the spin button (i.e. reel initiation spins and stopper responses). This was achieved by the use of “MaKey MaKey” (Joylabz, New York), an Arduino IDE-based keyboard proxy. This device makes use of the principle of open and closed circuits, in which it sends an impulse to the laptop simulating a keyboard key press whenever the circuit on the MakeyMakey board was closed. A wire in the MakeyMakey circuit was taped to the top of the play button, and positioned so that the circuit only closed (by the wire touching the metal chassis of the slot machine) when the button was pressed.  The device sent impulses to the AcqKnowledge laptop (Intel Core i5) through a USB cable. Using this device, I was able to mark (in real time) the participants’ spin/stop button presses on the physiological trace in Acqknowledge.   The second challenge was to mark the outcomes that were occurring on the slot machine. A Logitech HD C165 webcam recording the slot machine screen was connected to the Acqknowledge laptop and the video was directly synchronized with the real-time physiological trace in Acqknowledge. In this way, outcomes were matched with corresponding physiological changes.  Steps for Time-locking Events. After data collection, all psychophysiological data (heart rate and skin conductance) were processed in the AcqKnowledge 4.4 software. All events (i.e. play/stop button presses and outcomes) that happened during the slot machine play were time-locked on the recording trace, which enabled me to conduct event-related analyses. The time-locking for play/stop button presses was done in real-time, during the slot machine session using  31 the MakeyMakey keyboard proxy. The event-marking of the slot machine outcomes was performed manually, after the experimental session, using the synchronized video recording of the slot machine display. A “time stamp” was placed in AcqKnowledge at the moment the outcome was revealed. For losses and near misses, the event was marked when the final reel stopped spinning; the absence of visual and auditory feedback at that moment signalled the loss. When the participant received a win or LDW, a jingle was played after the fifth reel stopped spinning, followed by the “credit paid” counter counting up to the amount of credit the machine paid to the participants. Therefore, for wins, I dropped a time stamp when the “credit paid” counter stopped counting up. This would be the moment when the participants realized the size of the win or LDW they were getting. These time stamps formed the basis of the phasic analysis.   Spin Initiation Latency Data. The “spin initiation latency” was calculated by the interval (in milliseconds) between the outcome being revealed and the initiation of the next spin (see Figure 4). In other words, this was the length of time that the participant waited to initiate a new spin. Several steps were taken to clean the data for further analysis. First, all bonus spins were removed, as these were rare events and did not involve initiation presses. Second, 3 participants did not use the stopper device at all during the session, and these participants were excluded from the behavioural analysis. Third, for each participant, an outlier removal procedure using the Van Selst and Jolicouer (1994) method was performed for each outcome. The Van Selst and Jolicouer method determined the criterion for outliers based on the number of observations within the relevant cells, which is appropriate here as participants experienced a different mix of outcomes. 2.7% of the data were removed after the Van Selst and Jolicouer procedure.   32  Figure 4. Spin initiation latency illustration.  Heart Rate Analysis. The AcqKnowledge 4.4 software converted the ECG recording into heart rate during acquisition based on each inter-beat interval. This was sampling at 1000Hz. For analysis, heart rate was averaged across a 2-second bin before each outcome event marker (i.e. baseline value), and a 4-second bin after each outcome event marker (i.e. post outcome value). For trials where the post outcome value was greater than the baseline value, that data point was averaged across each of the outcomes in that outcome type. This resulted in 10 cell means for each participant, i.e. 5 types of outcomes (loss, LDW, win, bonus and near-miss) crossed with 2 stopper conditions (with and without stopper).    Skin Conductance Analysis. Skin conductance data were analysed using the same procedure as for heart rate. Preliminary analysis indicated that the skin conductance changes (which typically peak around 5s following a stimulus) were too sluggish to resolve as event-related responses during our slot machine game. Modern slot machines involve a spin every 3 seconds approximately, and this can be sped up by a further 50% (i.e. 1.5 seconds per spin) with the use of the stopper device (see also Dixon and Harrigan 2009). As a result, the skin conductance response for any given spin becomes easily “contaminated” by delayed time-course  33 for the previous spin. For these reasons, skin conductance data were disregarded for further analysis.   Statistical Analysis. Data were analyzed using Statistical Analysis Software (SAS) University Edition (SAS Institute Inc., Cary, NC) and SPSS Statistics (Version 22, IBM Corp, Armonk, NY). Multi-level regression models were used for the analysis of the speed of play and heart rate data, where every trial (i.e. each spin during game play) from each subject was entered into a regression model. In this approach, the correlation between trials for each subject is accounted for at the first level of the model, and the effects of interest were investigated at the second level of the model. The trial-by-trial data was entered into the regression model and were analyzed by the GENMOD procedure in SAS.    I elected to use regression instead of ANOVA because the trial-by-trial data for speed of play and heart rate was unbalanced, specifically due to two reasons: (1) The machine outcomes were randomized by the machine so that many participants did not experience all types of outcomes, and (2) the participants were free to use the stopper whenever and however many times they want. Therefore, missing data were numerous within the data sheet. For example, if a participant chose not to use the stopper, then all data regarding the stopper condition (across all five outcome types) was missing. ANOVA cannot be used on such unbalanced data where there is missing data for certain condition(s).   Some other classes of regression model were considered for the analysis but were also considered unsuitable. Logistic regression was not possible because the machines outcomes were  34 not binary and there were more than two outcomes that the analysis needed to account for. Generalized Estimating Equations (GEE) was not possible because GEE requires a balanced study design.   Regression Model for Speed of Play Data. I used a fixed-effect approach to investigate the predictor of stopper use on speed of play across all slot machine outcomes. The fixed effect model was picked because not only it could run on unbalanced data, it also treated each participant as their own control and therefore it eliminated any bias resulted from the different mix of outcomes that each participant experienced. In order to achieve this, Subject was entered into the model as a categorical predictor. Four levels of Outcomes (Loss, LDW, Win, Near-miss) were entered as a single categorical predictor, with Loss as the reference condition. The interaction term for “Outcome (Loss, LDW, Win, Near-miss) by Stopper use (Yes, No)” was entered as the third predictor variable. The criterion variable is speed of play, i.e. the log-transformed latency time measured in milliseconds.    After the initial regression analysis, I ran a follow-up analysis with the same model, but with an additional fourth predictor variable, the “Participant by Trial Number” interaction. This was done to control the potential bias resulted from a general increase in speed of play as the game play progressed. It is possible that participants might speed up naturally as the game progressed, as they got more familiar and comfortable with the slots game. This could confound the initial findings due to a lack of experimental control over when (and how many times) a participant experienced a certain outcome. For instance, this general increase in speed could  35 coincidently couple with a cluster of wins at the very end of the game, which would create a spurious relationship between wins and shorter latency.   Regression Model for Heart Rate Data. The fixed effects approach was again used to analyzed the heart rate data due to the same reasons mentioned above (i.e. unbalanced data). The model was identical to the one used for the speed of play data, except for the fact that the follow-up analysis was not done (as the “speed increase” confound mentioned above was exclusive to the speed of play data).   Leverage, Cook’s Distance and the Variance Inflation Factor (VIF) was calculated to identify cases that had undue influence to the model and/or poor model fit. Unless otherwise specified, the assumptions of each model was met, meaning that the model was reliable and was not unduly influenced by any cases.    36 Chapter Three: Results Study 1 Stopper Presses. The number of stopper presses for each participant is displayed in Figure 5. The distribution of stopper presses is positively skewed, with a modal bin of 0-33 presses. Only 10 participants in total pressed the stopper more than 100 times within their session, representing “heavy users”. To correct for skew, this variable was log-transformed; the transformation did successfully normalize the data (Figure 6). All further analyses on the “stopper presses” variable used the log-transformed values, although tables and figures display the raw scores.                 Subjective Measures. Gamblers’ Beliefs Questionnaire (GBQ). The Illusion of Control subscale scores from the GBQ were not significantly correlated with the number of stopper presses, r(28)=.22, p=.23.   Figure 5. Histogram of stopper presses for Study 1 (M=84.37, SD=89.69). Figure 6. Histogram of log-transformed stopper presses for Study 1.  37 Game Experience Questionnaire (GEQ). Table 6 shows the correlation matrix for the 7 GEQ subscales against the number of stopper presses.  Table 6.  Correlations between the seven GEQ subscales and stopper presses for Study 1  1 2 3 4 5 6 7 8 1. Stopper presses -        2. GEQ – Competence .20 -       3. GEQ – Immersion .06 .36 -      4. GEQ – Flow .20 .23 .31 -     5. GEQ – Tension/Annoyance .02 -.02 .01 .27 -    6. GEQ – Challenge  .37* .18 .18 .57** .07 -   7. GEQ – Negative Affect -.04 .05 .09 .22 .36 .07 -  8. GEQ – Positive Affect .22 .68** .30 .40* -.17 .48** .06 - *Correlation is significant at the .05 level. **Correlation is significant at the .01 level.    Players who used the stopper device more tended to report higher GEQ Challenge scores, r(28)=.37, p<.05. No significant relationship was found between stopper presses and other scales, including Competence, r(28)=.20, p>.05; Immersion, r(28)=.06, p>.05; Tension/Annoyance, r(28)=.02, p>.05; Negative Affect, r(28)=-.04, p>.05; and Positive Affect, r(28)=.22, p>.05.  Four additional questions assessing illusion of control during game play were added to the GEQ. For each question, participants were asked to give a written response if they had answered “Yes” to the first sub-question. Participants who answered “Yes” to questions 1, 2 or 4, and who mentioned the use of stopper in their written response, were classified as “stopper believers”. Six of 30 students (20%) were included in this group. Interestingly, 5 of these 6  38 participants responded to item 3 indicating that slot machines were due to chance alone, and only 1 participant endorsed a combination of chance and skill.  There was no significant difference in the number of stopper presses between stopper believers and non-believers on an independent sample t-test, t(28)=-1.25, p=.22.   Study 2 Stopper Presses. The distribution for the number of stopper presses is shown in Figure 7 and 8. As in Study 1, the distribution showed high positive skew, with a modal bin of 0-33 presses, and only 15 participants using the stopper device more than 100 times. As in study 1, this variable was log-transformed for further analysis, and the transformation successfully corrected the skew. An independent sample t-test was done to compare the stopper presses in Study 1 and 2, and no significant difference was found.               Figure 7. Histogram of stopper presses for Study 2 (M=74.48, SD=89.80).  Figure 8. Histogram of log-transformed stopper presses for Study 2.   39 Slot Machine Outcomes. The distribution of slot machine outcomes for each participant in Study 2 is summarized in Table 7. On average, the participants played 206 spins with their endowed $40. Losses comprised of 81.4% of the outcomes, while LDWs, wins, bonuses and near-misses comprised of 5.46%, 8.72%, 0.38% and 4.07% of the outcomes respectively.  Table 7.  The type, number and proportion of outcomes for each participant in Study 2      Loss  LDW P# Total spins  No stopper Stopper Loss total Loss proportion (%)  No stopper Stopper LDW total LDW proportion (%) 1 231  58 136 194 83.98  6 5 11 4.76 2 102  86 2 88 86.27  5 0 5 4.90 3 175  129 7 136 77.71  11 1 12 6.86 4 201  144 10 154 76.62  13 1 14 6.97 5 209  140 33 173 82.78  12 3 15 7.18 6 245  81 128 209 85.31  2 5 7 2.86 7 196  149 0 149 76.02  10 0 10 5.10 8 237  90 109 199 83.97  9 6 15 6.33 9 206  157 9 166 80.58  8 3 11 5.34 11 173  120 23 143 82.66  8 0 8 4.62 12 225  193 5 198 88.00  14 1 15 6.67 13 132  69 43 112 84.85  2 1 3 2.27 14 211  104 54 158 74.88  12 0 12 5.69 15 230  74 107 181 78.70  11 2 13 5.65 16 180  138 0 138 76.67  12 0 12 6.67 17 223  181 3 184 82.51  17 0 17 7.62 18 185  138 6 144 77.84  9 0 9 4.86 19 178  102 45 147 82.58  11 0 11 6.18 20 293  49 176 225 76.79  4 13 17 5.80 21 121  96 0 96 79.34  11 0 11 9.09 22 180  59 86 145 80.56  4 6 10 5.56 23 372  73 235 308 82.80  5 13 18 4.84 24 265  91 127 218 82.26  7 8 15 5.66 25 186  107 44 151 81.18  6 0 6 3.23 26 346  7 280 287 82.95  0 15 15 4.34 27 186  121 21 142 76.34  11 3 14 7.53 28 178  58 93 151 84.83  5 4 9 5.06 29 237  174 16 190 80.17  15 1 16 6.75 30 145  112 15 127 87.59  4 0 4 2.76 31 212  176 1 177 83.49  8 0 8 3.77 32 119  97 1 98 82.35  5 0 5 4.20 Average 205.77  108.81 58.55 167.35 81.37  8.29 2.94 11.23 5.46  40 Table 7. (continued) The type, number and proportion of outcomes for each participant in Study 2       Win  Bonus P# Total spins  No stopper Stopper Win total Win proportion (%)  No stopper Stopper Bonus total Bonus proportion (%) 1 231  9 11 20 8.66  0 1 1 0.43 2 102  4 0 4 3.92  0 0 0 0.00 3 175  17 2 19 10.86  0 0 0 0.00 4 201  21 1 22 10.95  1 0 1 0.50 5 209  11 2 13 6.22  1 0 1 0.48 6 245  6 9 15 6.12  1 1 2 0.82 7 196  24 0 24 12.24  1 0 1 0.51 8 237  12 5 17 7.17  0 1 1 0.42 9 206  18 1 19 9.22  1 0 1 0.49 11 173  16 2 18 10.40  0 0 0 0.00 12 225  4 1 5 2.22  0 0 0 0.00 13 132  7 4 11 8.33  0 0 0 0.00 14 211  20 8 28 13.27  0 0 0 0.00 15 230  10 14 24 10.43  0 0 0 0.00 16 180  19 0 19 10.56  3 0 3 1.67 17 223  12 0 12 5.38  1 0 1 0.45 18 185  22 2 24 12.97  0 0 0 0.00 19 178  10 4 14 7.87  0 0 0 0.00 20 293  10 22 32 10.92  1 1 2 0.68 21 121  10 0 10 8.26  0 0 0 0.00 22 180  6 15 21 11.67  0 1 1 0.56 23 372  9 19 28 7.53  0 3 3 0.81 24 265  13 11 24 9.06  0 0 0 0.00 25 186  13 4 17 9.14  0 4 4 2.15 26 346  0 30 30 8.67  0 0 0 0.00 27 186  18 2 20 10.75  1 0 1 0.54 28 178  4 7 11 6.18  0 1 1 0.56 29 237  20 0 20 8.44  0 0 0 0.00 30 145  8 0 8 5.52  0 0 0 0.00 31 212  17 0 17 8.02  0 0 0 0.00 32 119  11 0 11 9.24  0 1 1 0.84 Average 205.77  12.29 5.68 17.97 8.72  0.35 0.45 0.81 0.38  41 Table 7. (continued) The type, number and proportion of outcomes for each participant in Study 2                                  Subjective Measures. Real-world Stopper Usage. Participants were asked to indicate their frequency of stopper use when gambling in the casino. A significant positive correlation was found between their rated stopper use frequency in the casino, and the number of stopper presses during test session,    Near-miss  P# Total spins  No stopper Stopper NM total NM proportion (%) 1 231  3 2 5 2.16 2 102  5 0 5 4.90 3 175  8 0 8 4.57 4 201  10 0 10 4.98 5 209  7 0 7 3.35 6 245  7 5 12 4.90 7 196  12 0 12 6.12 8 237  3 2 5 2.11 9 206  8 1 9 4.37 11 173  3 1 4 2.31 12 225  7 0 7 3.11 13 132  2 4 6 4.55 14 211  11 2 13 6.16 15 230  2 10 12 5.22 16 180  8 0 8 4.44 17 223  8 1 9 4.04 18 185  8 0 8 4.32 19 178  2 4 6 3.37 20 293  5 12 17 5.80 21 121  4 0 4 3.31 22 180  2 1 3 1.67 23 372  3 12 15 4.03 24 265  3 5 8 3.02 25 186  1 7 8 4.30 26 346  0 14 14 4.05 27 186  4 5 9 4.84 28 178  0 6 6 3.37 29 237  10 1 11 4.64 30 145  4 2 6 4.14 31 212  10 0 10 4.72 32 119  1 3 4 3.36 Average 205.77  5.19 3.23 8.42 4.07  42 r(29)=.46, p<.01. No significant relationship was found between casino stopper use and Illusion of Control score from GBQ, r(29)=.32, p=.08. Gamblers’ Beliefs Questionnaire (GBQ). The number of stopper presses was not significantly correlated with the GBQ illusion of control subscale, r(29)=.01, p=.94. An independent sample t-test showed that the average GBQ Illusion of control score was significantly higher in Study 2 (i.e. regular slot machine players) than Study 1 (i.e. UBC students) (see Table 8), t(61)=-4.28, p<.001.  Table 8. Average GBQ Illusion of Control scores in Study 1 and 2  GBQ – Illusion of Control Score Average (SD) Study 1 23 (7.50) Study 2 31.10 (7.97)  Game Experience Questionnaire (GEQ). The correlation matrix for the seven GEQ subscales against stopper presses in shown in Table 9. Table 9.  Correlations between the seven GEQ subscales and stopper presses for Study 2  1 2 3 4 5 6 7 8 1. Stopper presses -        2. GEQ – Competence -.20 -       3. GEQ – Immersion .05 .45* -      4. GEQ – Flow -.06 .39* .64** -     5. GEQ – Tension/Annoyance .33 -.32 -.17 -.04 -    6. GEQ – Challenge  -.16 .53** .62** .55** .11 -   7. GEQ – Negative Affect .15 -.17 -.54** -.49** .59** -.39* -  8. GEQ – Positive Affect -.46** .69** .34 .41* -.52** .32 -.46** - *Correlation is significant at the .05 level. **Correlation is significant at the .01 level.    43 Participants who made more stopper presses during the slot machine session showed lower levels of Positive Affect in response to the game, r(29)=-.46, p<.01. However, no significant relationship was found between stopper presses and other scales, including Competence, r(29)=-.20, p>.05; Immersion, r(29)=.05, p>.05; Flow, r(29)=-.06, p>.05; Tension/Annoyance, r(29)=.33, p>.05; Challenge, r(29)=-.16, p>.05; and Negative Affect, r(29)=.15, p>.05. An independent sample t-test was done to compare the GEQ scores from Study 1 with those from Study 2. The Study 2 scores for Competence, t(59)=-2.25, p=.03, Immersion, t(59)=-2.47, p=.02, Flow, t(59)=-2.73, p=.008, Challenge, t(59)=-2.06, p=.04, and Negative Affect, t(59)=-2.47, p=.02, were significant higher than those in Study 1. However, the scores for Tension/Annoyance, t(59)=-.81, p=.42, and Positive Affect, t(59)=1.20, p=.23, were not significantly differed.  Table 10.  Average GEQ scores in Study 1 and 2  GEQ Average (SD)  Competence* Immersion* Flow** Tension/ Annoyance Challenge* Positive Affect Negative Affect* Study 1 0.73 (.83) 1.10 (.79)  0.98 (.78) 0.80 (.79) 0.50 (.81)  1.45 (.93) 1.17 (.96) Study 2 1.31 (1.13) 1.69 (1.06) 1.65 (1.08) 1.00 (1.10) 1.00 (1.06) 1.13 (1.14) 1.82 (1.11) *T-test comparing the Study 1 vs Study 2 scores is significant at the .05 level. ** T-test comparing the Study 1 vs Study 2 scores is significant at the .01 level.    On the additional GEQ question about their experiences of the stopper device, 16 of 31 participants (52%) were coded as stopper believers. Of those stopper believers, 9 of them believed that a winning combination was due to both chance and skills, while the rest of them believed that it was due to chance alone. There was no significant difference in the number of stopper presses between stopper believers and non-believers in an independent sample t-test, t(29)=1.09, p=.28.   44 Do Wins Reinforce Stopper Use? For each participant, I separated each instance of a regular win depending on whether the stopper device was or was not used on that spin. Note that we had no control over the types of machine outcomes and also when the stopper was pressed. Specifically, 9 participants did not experience any wins in either the stopper or no-stopper conditions, and were excluded from this analysis. I calculated the proportion of stopper presses during the five prior spins before the win, and the five successive spins after the win. These proportion values were entered in a 2x2 repeated measures ANOVA with Stopper (did use, did not use) and Time (before/after) as factors. A significant interaction of Stopper x Time was observed, F(1,21)=6.34, MSE=.01, p<.05 (see Figure 7). This was qualified by a significant main effect for the Stopper factor, F(1,21)=19.46, MSE=.03, p<.001, such that stopper use was higher either side of stopper wins, relative to no-stopper wins, and no significant main effect for Time factor, F(1, 21)=.12, MSE=.01, p=.73. A simple effects analysis was run on the interaction effect. Stopper presses increased significantly during the 5 spins after a stopper win compared to the 5 spin baseline, t(21)=-2.32, p<.05. On a no-stopper win, the frequency of stopper presses did not change significantly from before to after the win outcome, t(21)=1.28, p=.21.   Figure 9. The interaction between “stopper/no-stopper wins” and “time”. 00.10.20.30.40.55 spins before the win 5 spins after the winStopper presses(measured in proportion)No-stopper win Stopper win 45 Spin Initiation Latency. There was a main effect of outcome for LDWs, wins and near-misses, such that the latencies for each of these outcomes was significantly longer than the latency following losses. Note that when generating main effects from a Fixed Effect model, one level of the variable (Outcome) must be assigned as the reference level for other levels to compare against. Losses were selected as the reference category because they are the most common outcomes in slot machine gambling, and are the natural baseline in the ‘post-reinforcement pause’ effect. In addition, there was a significant Outcome by Stopper interaction for losses, LDWs, and wins. The latencies for these three outcomes were significantly shorter when the stopper was used than when it was not used (See Table 11 for relevant statistics and Figure 8 for an illustration of the regression output). These results remained significant in the follow-up analysis adding the participant x trial number interaction term, and therefore there is no indication that participants naturally varied their speed as the game progressed. Table 11.  Fixed effect analysis for latency time/speed of play  Types of outcomes  Loss LDW Win Near-miss Variables β (SE) p β (SE) p β (SE) p β (SE) p Outcomes - - .45 (.03) <.001 .31 (.03) <.001 .19 (.04) <.001 Outcomes x Stopper use interaction -.29 (.02) <.001 -.19 (.05) <.001 -.22 (.06) 0.0003 -.10 (.06) .10 Note. All levels of Outcomes were compared to Loss (baseline). SE = standard error. Subject was also entered as a categorical predictor but the individual beta values are not reported due to their arbitrary nature (derived from comparison to the first subject).    46  Figure 10. Differences in spin initiation latencies following the four slot machine outcome types, from the Fixed Effect analysis.  Heart Rate Data. There was a significant main effect of outcome for wins, meaning that the heart rate after a win was significantly higher than those of losses (See Table 12 for relevant statistics and Figure 9 for an illustration of the regression output). Table 12.  Fixed effect analysis for heart rate  Types of outcomes  Loss LDW Win Near-miss Variables β (SE) p β (SE) p β (SE) p β (SE) p Outcomes - - 3.06 (1.60) 0.06 4.00 (1.39) .004 1.22 (1.89) 0.52 Outcomes x Stopper use interaction 0.32 (0.90) .72 -4.44 (2.86) .12 .25 (2.27) .91 2.70 (3.00) .36 Note. All levels of Outcomes were compared to Loss (baseline). SE = standard error. Subject was also entered as a categorical predictor but the individual beta values are not reported due to their arbitrary nature (derived from comparison to the first subject).   47  Figure 11. Differences in heart rate following the four slot machine outcome types, from the Fixed Effect analysis.  48 Chapter Four: Discussion The current study investigated two possible mechanisms for the use of the stopper device on modern slot machines: that slots players use the stopper device because it increases their perception of skill and control over the outcomes, or that slots player use the stopper device to make the game faster. Several precautions were taken to improve the ecological validity of the study: i) the study was conducted in the Casino Lab, using an ambiance similar to an actual casino in terms of seating and lighting; ii) a real modern slot machine was used in the study, iii) participants were able to decide when to use and not use the stopper. In my community sample (Study 2), the number of stopper presses during the laboratory session was significantly and positively correlated with their self-reported preferences for EGMs equipped with stoppers in their real casino gambling. This supported the presence of external validity of behaviour recorded in the Casino Lab.  The results indicated minimal relationship between stopper use and the illusion of control. Specifically, there was no significant correlation between stopper use (in either the lab setting or as reported for real casinos) and the illusion of control subscale on the trait-related measure (GBQ). Results from the state-related measure (GEQ) showed a positive relationship between stopper use and the Challenge scale in the student sample, although this relationship was not substantiated in Study 2. Rather, in the community sample, a negative correlation was seen between stopper use and GEQ Positive Affect (in response to the game). Supplementary questions on the GEQ indicated that 20% of student sample and 52% of community sample were “stopper believers” (i.e. they reported that using the stopper would have some sort of impact on the outcomes and/or on the likelihood of winning), but interestingly, across both samples, no  49 significant difference was seen for the frequency of stopper use between the stopper believers and non-believers. Despite the lack of support for the cognitive account to explain stopper behaviours, the behavioural analysis in Study 2 revealed that stopper presses were reinforced by successful outcomes during the game, i.e. when a win was generated with the use of stopper, participants increased their stopper presses following such outcomes.   In addition, a connection was found between stopper use and an increase in speed of play. Spin latency data were analyzed for the community participants, showing shorter spin latencies following spins that involved the use of the stopper. This is interpreted as an indication of the urgency to start a new game. The analysis of outcome-related psychophysiological responses indicated a significant heart rate increase following win outcomes, but no significant relationship was found between stopper use and heart rate changes.   Proposed Mechanism 1: Illusion of Control The first hypothesis focused on the relationship between players’ use of the stopper device and one of the cardinal cognitive distortions, the illusion of control. Illusory control was assessed using a trait questionnaire, as well as a state measure of illusory control in response to the slot machine session. No significant relationship was found between stopper use (in either the laboratory session or based on the real casino tendency) and the illusion of control subscale on the Gamblers’ Beliefs Questionnaire, i.e. the trait measure. This finding was unexpected, as a previous stopper study (Ladouceur & Sévigny, 2005) found that a group of players who were instructed to use the stopper showed erroneous cognitions about the game, that were consistent with illusory control. The discrepancy results call into question to whether GBQ has successfully  50 captured the participants’ illusory control (if any) in the current study. First, it is important to note that community participants scored significantly higher than the student participants on the GBQ. This indicates internal validity of the GBQ; for example, in earlier work, regular and problem gamblers typically show higher scores on gambling-related cognitions than healthy non-gamblers (Clark, 2010). Specifically using the GBQ, Myrseth, Brunborg, & Eidern (2010) found that pathological gamblers reported more cognitive distortions than non-pathological gamblers. Similarly, Floyd, Whelan, & Meyers (2006) studied whether warning messages modified gambling beliefs and behaviours in a laboratory setting, using the GBQ as a measure of cognitive distortions. They found that the effectiveness of warning messages was predicted by GBQ scores, such that lower GBQ scores were observed in participants who had received messages about gambling-related cognitive distortions. The GBQ illusion of control subscale also predicted gamblers’ preference for skill-based games (e.g. card games and sports betting) (Floyd, Whelan, & Meyers, 2006).   Despite this strong prior evidence of validity for the GBQ, it is possible that the illusion of control subscale may not measure the same aspects of illusory control that are elicited by the slot machine stopper device. An alternative questionnaire for measuring illusion of control, the Gambling-Related Cognitive Scale (GRCS), included three subscales measuring distinct elements of illusory control, namely predictive control (e.g. believing that series of losses are bound to be followed by wins), illusion of control (e.g. performing specific rituals to increase my chance of winning) and interpretative bias (e.g. believing that gambling success is related to one’s skills and knowledge) (Raylu & Oei, 2004). In comparison, the GBQ was an “overarching” measure that assess participants’ cognitive distortions without addressing the specific elements  51 of illusion of control beliefs. The stopper feature may be specifically related to what the GRCS terms interpretative bias, and this may be underweighted on the GBQ subscale.   Results from the state-related measure (GEQ) showed a positive relationship between stopper use and the Challenge scale in the student sample, such that students who used the stopper more tended to describe the game as more challenging. As a game of pure chance, the subjective perception of the slot machine as “challenging” could indicate that these participants experienced some illusory control, which is further linked to their exploration of the stopper device. As this correlation was not confirmed in the community sample of regular slot machine players, it should be regarded with caution. However, challenge was found to be a common motive for gambling in college students (Neighbors et al., 2002), so students may tend to see gambling games as games of skills but not games of chance. Their inexperience in gambling could also fuel this incorrect perception of gambling. Thus, it may make the students more likely than the regular gamblers to view stopper as a “challenging”, skill-related feature.   The community gamblers scored higher overall on 5 of the 7 GEQ scales compared to the students, which may be indicative of a more intense game experience in the regular players in the lab environment. In the community sample, a negative correlation was seen between stopper use and GEQ scale reflecting Positive Affect in response to the game, and this will be considered further below in the section on Speed of Play.   A number of supplementary questions asked participants specifically about their beliefs concerning the effectiveness of the stopper device. On these questions, 20% of the student  52 participants and 52% of the community participants were coded as “stopper believers”. Critically, across both samples, no significant difference was seen for the frequency of stopper use between the stopper believers and non-believers. Thus, although many participants (especially in the community sample) developed some conscious, erroneous beliefs about the stopper device after the play session, these beliefs did not align with how frequently they used the stopper during the session. This apparent decoupling of cognition and behaviour is not unheard of in the gambling field. Sévigny and Ladouceur (2003) studied the “double switching” concept, in which gamblers demonstrated rational perceptions about the game before and after gambling, but made a shift to more irrational cognitions during gambling. They were said to have “switched” their rationality on and off. For example, video lottery gamblers repeatedly tapped the screen at different places, as if this behaviour could affect the outcomes, but before and after the game they explicitly described the outcomes as being governed by pure chance. It is also plausible that explicit, verbalizable cognitions about the game may represent post-hoc rationalizations. In his book Thinking, Fast and Slow, Kahneman (2001) describes human decision-making as arising from a competition between dual processes: System 1 is fast, automatic, and associative in nature; whereas System 2 is slow and effortful, but more capable of logic and problem solving than System 1. As gambling activities provide no predictability between successive events, System 1 may respond by establishing (false) associations between events (in this case, stopper presses and slot machine outcomes). System 2 may then make sense of these associations through a rationalization process, that pressing the stopper must have an impact on the outcomes, consistent with a conscious illusion of control (Wiers & Stacy, 2006). In short, the supplementary questions may reflect System 2’s attempts to retrospectively explain the individual’s use of the stopper device, a behaviour that may have actually been under the control  53 of a separate lower-level system. As a result, these rationalized responses are very loosely related to the actual behaviour under scrutiny (stopper presses). Taking these results together, hypothesis 1a was only minimally supported.   Study 2 also employed a behavioural analysis, based on the cognitive appraisal of ‘skill acquisition’ in the illusion of control hypothesis. For participants who encountered a win on at least one occasion on which they had used the stopper device, I measured the frequency of stopper presses on the subsequent five spins, compared to the five spins prior to the win. I calculated similar values for stopper presses either side of winning outcomes on which the stopper device had not been used. This analysis showed a significant interaction effect (visible in Figure 8) such that the use of the stopper device was reinforced by successful wins; thus supporting hypothesis 1b.   The (false) connection between win outcomes and stopper presses could be a product of the automatic System 1 processing. An increase in the use of the stopper button following perhaps only a single pairing of the stopper behaviour with a win is reminiscent of examples of “one-shot” superstitions from research in social psychology. When an especially good or bad outcome follows a single instance of a notable behaviour, people may develop a belief that the behaviour is causally relevant to the outcome (Risen, 2015). For example, a sports fan may start calling his shirt a “lucky shirt” after watching his team win a big game wearing that shirt. Risen, Gilovich, & Dunning (2007) examined these “one-shot” illusory correlations in relation to the formation of stereotypes. When members of a minority group perform an unusual behaviour, these events may be processed differently, triggering more elaborate mental processing, greater  54 attributional thinking and enhanced subsequent memory. This illusory correlation between the behaviour and the group can then be generalized to other members of the group, resulting in the formation of stereotypes. In the context of slot machine gambling, the infrequent nature of wins makes them a good candidate for a salient positive outcome. The stopper device may be regarded as a novel and unusual feature and perhaps as an unconventional way to play the slot machine. Therefore, when the unusual behaviour (i.e. using the stopper) is paired with the winning outcome, these conditions are optimal for eliciting the “one-shot” superstition between the two events.   Proposed Mechanism 2: Increasing Speed of Play The second mechanism of interest focused on the speed of play, measured by the spin initiation latencies. These spin latencies were shortest following loss outcomes, and longest following the unpredictable wins. This replicates the classic post-reinforcement pause (PRP) phenomenon, shown by Delfabbro & Winefield (1999) and Dixon et al. (2013). The part that was not replicated was the latencies for near-misses. Dixon et al. (2013) found that near-misses actually generated the shortest latencies, even when comparing with losses. They argued near-misses were an especially frustrating type of outcomes (even more so than losses), hence the shorter latencies. However, in the current study, the near-miss latencies were found to be longer than those for losses, and closer to those for wins than for losses. This is more reminiscent of brain imaging data in which near-misses recruit overlapping brain circuitry to winning outcomes (Clark et al., 2009). It is important to note that the near-misses on the Dragon’s Fire slot machine were quite different from the near-misses delivered in the Dixon et al. (2013) study, which could explain the discrepant latency results. On Dragon’s Fire, the near-misses were accompanied by  55 sound effects (e.g. a “clunking” sound when the bonus purple dragon symbol appeared), which were absent in Dixon et al. (2013)’s near-misses. These special effects could make the near-misses of the current study appeared to be more pleasant and more like a win, instead of the frustrating outcome seen by Dixon et al. (2013).   Most importantly, stopper usage was found to increase the gamblers’ speed of play, supporting Hypothesis 2. Significant Outcome by Stopper interactions were seen for losses, LDWs and wins, such that the spin latencies for each of these outcomes were shorter when the stopper was used, compared to when it was not used. The use of the spin latency variable rather than the average number of spins across the player’s session provides a clearer indication that these effects reflect the players’ urge to start the next spin faster. Of course, the stopper additionally speeds up the rate of play simply by reducing the spin duration, thereby increasing the number of spins within a fixed period of time.   Dixon et al. (2013) interpreted shorter spin latencies as a sign that the participants were frustrated by the outcome and that they want to quickly get out of the frustrated state by initiating a new spin as soon as possible. When applying this frustration-based interpretation to the current study, faster speed of play on stopper trials may equate to higher levels of frustration. Although there was no direct measure of frustration in this study, the GEQ findings for Study 2 do align with this interpretation, as there was a negative relationship between stopper usage and the Positive Affect scale from the GEQ. This revealed an interesting dynamic of stopper use – participants increased their speed of play through pressing the stopper, but through this process, participants may enhance frustration and reduce their enjoyment of the game.   56 Psychophysiological Responses to Slot Machine Outcomes  No significant relationship was found between the use of the stopper device and heart rate changes. In the absence of such a univariate relationship, I did not pursue mediator tests to see whether heart rate mediated the relationship between stopper use and either the illusion of control or speed of play. Nevertheless, the psychophysiological analysis did show a significant main effect of winning outcomes, such that heart rate showed a phasic increase following wins compared to losses. This replicates the findings from the Moodie and Finnigan (2005) study, where they asked participants to play slots in an actual gambling setting while monitoring their heart rate. They found that participants experienced an increase in tonic heart rate in sessions where a jackpot was experienced, which aligned with the results from this study. The demonstration of this effect using a phasic (outcome-related) heart rate response strengthens this finding, as tonic recording naturally contains other events (e.g. talking, breaks in play) that add noise to physiological data.   With regard to the null results on the stopper device on psychophysiological responses, there are at least two points of caution here. First, although heart rate is superior to skin conductance in measuring fast changes in physiological arousal, it remains possible that heart rate could fail to capture some phasic changes in response rapid successive stimuli. This is key problem for slot machine behaviour, where spins (and therefore outcomes) naturally occur every 3 seconds approximately, and the problem is intensified by the stopper device. In this regard, programmed simulations may have advantages over quasi-experimental studies using real slot machines, given that the experimenter can manipulate the spin duration on a simulation, to allow time for physiological responses to recover between trials, and precisely characterize the reaction  57 to different outcomes. Second, the event timing markers for win and LDW outcomes were modelled to the moment when the “credit paid” window stopped counting up; this decision was taken in order to calculate the spin initiation latencies (more will be discussed in Limitations). On the slot machine used in the present study, the auditory and visual effects that accompanied wins and LDWs took place while the machine counted up the credits for a win. In an animal study by Barrus and Winstanley (2016), salient win-related audiovisual cues were found to enhance choice of riskier and more disadvantage options in a rat gambling task. By this reasoning, if sensory feedback drives the physiological response to the wins and LDWs, then marking the outcome at the very end of the sensory period may miss some outcome divergent signals.  Limitations  Several limitations should be noted. First, the process of event marking the slot machine events against the psychophysiological trace was complicated and time-consuming. Although we devised a novel method for precisely synchronizing the spin presses to the psychophysiological trace, the slot machine outcomes were marked manually based on time-locked webcam footage of the slot machine display. A decision was made to model the outcomes following wins and LDWs to the termination (i.e. offset, as opposed to onset) of the visual and auditory feedback, as described above. This decision was taken for the calculation of the spin initiation latencies: if wins and LDWs were marked at the onset of the sensory feedback, i.e. as the final reel reached a standstill), then the variable length of the feedback (proportional to the number of credits won) would confound the interval to the subsequent spin. It is likely that participants wait until the jingles completes before starting another spin – this would be a key source of noise in the spin  58 latency calculations. However, in relation to the psychophysiological analysis, this strategy for outcome marking may fail to detect rapid heart rate change to the sensory feedback itself. Future experiments would be advised to use two sets of event markers, one for the spin latency calculations and the other for the psychophysiological response analysis.   More broadly, studies on real slot machines in our Casino Lab trade off greater ecological validity against weaker experimental control. Due to the lack of experimental control, some participants in our sample yield insufficient data to be included in the analysis. This occurs because the outcomes are randomized on a real slot machines, and also because we allowed participants to use the stopper device as often (or as rarely) as they wished – some participants did not use the stopper button at all. Similarly, bonus spins could not be analyzed because there were very few of them (less than 0.5% of the total number of spins, see Table 7), and a proportion of participants did not encounter any these events. This then prevented us from studying the bonus outcomes, which are an exciting event in slot machine gambling and some gamblers seemed to be very motivated to get a bonus outcome. We chose to analyze the spin latencies and heart rate data with fixed effects regression model because of unbalanced outcomes, in which each participant can experience a different mix of the outcomes (e.g. some had multiple big wins and some had long streaks of losses). Although the fixed effect model could handle unbalanced data and take into account the fact that every participant was contributing a different number of outcomes, this statistical correction would not be able to fully clear the noises resulting from the unbalanced design. For instance, a participant who experience multiple big wins throughout the game may have a more positive game experience than someone who only had long streaks of losses. If future studies are intended to study more specific  59 hypotheses with regard to stopper usage (i.e. studies that are not exploratory in nature), they should consider using slot machine simulator with programmable outcomes to achieve a greater level of control.    Last but not least, it is important to note that in the GEQ analysis, no correction was applied for multiple comparisons. In Study 1, a significant correlation was observed between the GEQ Competence scale and the number of stopper presses, but this relationship was not replicated in Study 2. Therefore, the correlation found in Study 1 may actually be spurious and a result of a Type 1 error.   Conclusions and Implications The current study explored two psychological accounts for why slot machine players make use of the stopper device during game play. It was evident that stopper behaviours could be reinforced by successful gambling outcome, but there was no support a cognitive mechanism (illusion of control) was behind this reinforcement link. Spins with the use of stopper were also found to have shorter spin initiation latencies, thus increasing the speed of play. The overall pattern of results indicate that stopper use may be better explained by a lower-level behavioural/associative account based upon reinforcement learning, than a higher-level cognitive account based on conscious appraisals of illusory control.   These findings have implications for responsible gambling. In light of the spin latency results, use of the stopper device clearly increases the speed of play. A faster game speed enables gamblers to play more games per unit of time, and thus risks spending (and losing) more money.  60 Problem gamblers tend to play slot machines at faster play speeds, and game speed is positively related to ratings of enjoyment and satisfaction, in both recreational and problem gamblers (Blaszczynski et al., 2001). Strategies in responsible gambling education and regulation to limit certain game features in slot machine design should therefore be considered to mitigate these potential harms that can arise from the stopper device feature.    61 References Anderson, G., & Brown, R. I. F. (1984). Real and laboratory gambling, sensation‐seeking and  arousal. British journal of Psychology, 75(3), 401-410. Barrus, M. M., & Winstanley, C. A. (2016). Dopamine D3 receptors modulate the ability of win-paired cues to increase risky choice in a rat gambling task. The Journal of Neuroscience, 36(3), 785-794. BC Ministry of Health. (2013). Lower the stakes: A public health approach to gambling in British Columbia. Victoria, BC: Ministry of Health. Retrieved from http://www2.gov.bc.ca/assets/gov/health/about-bc-s-health-care-system/office-of-the-provincial-health-officer/reports-publications/annual-reports/gambling-in-bc.pdf Blaszczynski, A., Sharpe, L. & Walker, M. (2001). The Assessment of the Impact of the Reconfiguration on Electronic Gaming Machines as Harm Minimization Strategies for Problem Gambling. Report for the Gaming Industry Operators Group. Accessed 23 June 16. Available at: http://www.psych.usyd.edu.au/gambling/GIO_report.pdf. Boyd, W. H. (1982). Excitement: The gambler’s drug. In W. R. Eadington (Ed.), The gambling papers. Reno, NV: University of Nevada. Breen, R. B., & Zimmerman, M. (2002). Rapid onset of pathological gambling in machine gamblers. Journal of Gambling Studies, 18(1), 31-43. Brown, R. I. F. (1986). Arousal and sensation-seeking components in the general explanation of gambling and gambling addictions. International Journal of the Addictions, 21(9-10), 1001-1016.  62 Clark, L., Lawrence, A. J., Astley-Jones, F., & Gray, N. (2009). Gambling Near-Misses Enhance Motivation to Gamble and Recruit Win-Related Brain Circuitry. Neuron, 61(3), 481–490. http://doi.org/10.1016/j.neuron.2008.12.031 Clark, L. (2010). Decision-making during gambling: an integration of cognitive and psychobiological approaches. Philosophical Transactions of the Royal Society B: Biological Sciences, 365(1538), 319-330. Clark, L., Crooks, B., Clarke, R., Aitken, M. R., &  Dunn, B. D. (2012). Physiological responses to near-miss outcomes and personal control during simulated gambling. Journal of Gambling Studies, 28(1), 123-137. Cohen, I. (2015). Voluntary Self-Exclusion: BCLC Longitudinal Study [Powerpoint slides]. Retrieved from http://horizonsrg.bclc.com/past-conferences/2015-speakers-and-presenters/irwin-cohen.html Coventry, K. R., & Hudson, J. (2001). Gender differences, physiological arousal and the role of winning in fruit machine gamblers. Addiction, 96(6), 871-879. Coventry, K. R., & Norman, A. C. (1997). Arousal, sensation seeking and frequency of gambling in off‐course horse racing bettors. British Journal of Psychology, 88(4), 671-681. Davis, D., Sundahl, I., & Lesbo, M. (2000). Illusory Personal Control as a Determinant of Bet Size and Type in Casino Craps Games. Journal of Applied Social Psychology, 30(6), 1224-1242. Delfabbro, P. H., & Winefield, A. H. (1999). Poker-machine gambling: An analysis of within session characteristics. British Journal of Psychology, 90, 425.  63 Dickerson, M. (1993). Internal and external determinants of persistent gambling: Problems in generalising from one form of gambling to another. Journal of Gambling Studies, 9(3), 225-245. Dickerson, M., Hinchy, J., England, S. L., Fabre, J., & Cunningham, R. (1992). On the determinants of persistent gambling behaviour. I. High‐frequency poker machine players. British Journal of Psychology, 83(2), 237-248. Dixon, M. J., Harrigan, K. A., Jarick, M., MacLaren, V., Fugelsang, J. A., & Sheepy, E. (2011). Psychophysiological arousal signatures of near-misses in slot machine play. International Gambling Studies, 11(3), 393-407. Dixon, M. J., Harrigan, K. A., Sandhu, R., Collins, K., & Fugelsang, J. A. (2010). Losses disguised as wins in modern multi‐line video slot machines. Addiction, 105(10), 1819-1824. Dixon, M. J., MacLaren, V., Jarick, M., Fugelsang, J. A., & Harrigan, K. A. (2013). The frustrating effects of just missing the jackpot: slot machine near-misses trigger large skin conductance responses, but no post-reinforcement pauses. Journal of Gambling Studies, 29(4), 661-674. Dow Schüll, N. (2012). Addiction by design: Machine gambling in Las Vegas. Princeton University Press. Dowling, N., Smith, D., & Thomas, T. (2005). Electronic gaming machines: are they the ‘crack‐cocaine’ of gambling? Addiction, 100(1), 33-45. Felton, M., & Lyon, D. O. (1966). The post‐reinforcement pause. Journal of the Experimental Analysis of Behavior, 9(2), 131-134.  64 Ferris, J., & Wynne, H. (2001). The Canadian problem gambling index. Ottawa, ON: Canadian Centre on Substance Abuse. Floyd, K., Whelan, J. P., & Meyers, A. W. (2006). Use of warning messages to modify gambling beliefs and behavior in a laboratory investigation. Psychology of Addictive Behaviors, 20(1), 69. Goudriaan, A. E., Oosterlaan, J., de Beurs, E., & van den Brink, W. (2006). Psychophysiological determinants and concomitants of deficient decision making in pathological gamblers. Drug and alcohol dependence, 84(3), 231-239. Griffiths, M. (1993). Fruit machine gambling: The importance of structural characteristics. Journal of gambling studies, 9(2), 101-120. Griffiths, M. (1999). Gambling technologies: Prospects for problem gambling. Journal of gambling studies, 15(3), 265-283. Harrigan, K. A., & Dixon, M. (2009). PAR Sheets, probabilities, and slot machine play: Implications for problem and non-problem gambling. Journal of Gambling Issues, 81-110. Harrigan, K., Dixon, M., MacLaren, V., Collins, K., & Fugelsang, J. (2011). The maximum rewards at the minimum price: Reinforcement rates and payback percentages in multi-line slot machines. Journal of Gambling Issues, 11-29. IJsselsteijn, W., Van Den Hoogen, W., Klimmt, C., De Kort, Y., Lindley, C., Mathiak, K., ... & Vorderer, P. (2008, August). Measuring the experience of digital game enjoyment. In Proceedings of Measuring Behavior (pp. 88-89). Netherlands: Maastricht. Kahneman, D. (2011). Thinking, fast and slow. New York: Farrar, Straus and Giroux.  65 Ladouceur, R., & Mayrand, M. (1987). The level of involvement and the timing of betting in roulette. The Journal of Psychology, 121(2), 169-176. Ladouceur, R., & Sévigny, S. (2005). Structural characteristics of video lotteries: Effects of a stopping device on illusion of control and gambling persistence. Journal of Gambling Studies, 21(2), 117-131. Ladouceur, R., & Sévigny, S. (2006). The impact of video lottery game speed on gamblers. Journal of Gambling Issues, 17. doi: 10.4309/jgi.2006.17.12. Ladouceur, R., Sévigny, S., Blaszczynski, A., O'Connor, K., & Lavoie, M. E. (2003). Video lottery: Winning expectancies and arousal. Addiction, 98(6), 733-738. Ladouceur, R., Sylvain, C., Boutin, C., Lachance, S., Doucet, C., & Leblond, J. (2003). Group therapy for pathological gamblers: A cognitive approach. Behaviour research and therapy, 41(5), 587-596. 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. Langer, E. J. (1975). The illusion of control. Journal of personality and social psychology, 32(2), 311. Linnet, J., Rømer Thomsen, K., Møller, A., & Callesen, M. B. (2010). Event frequency, excitement and desire to gamble, among pathological gamblers. International Gambling Studies, 10(2), 177-188. Linnet, J., Thomsen, K. R., Møller, A., & Callesen, M. B. (2013). Slot Machine Response Frequency Predicts Pathological Gambling. International Journal of Psychological Studies, 5(1), 121.  66 Loba, P., Stewart, S. H., Klein, R. M., & Blackburn, J. R. (2001). Manipulations of the features of standard video lottery terminal (VLT) games: Effects in pathological and non-pathological gamblers. Journal of Gambling Studies, 17(4), 297-320. Lole, L., Gonsalvez, C. J., Blaszczynski, A., & Clarke, A. R. (2012). Electrodermal activity reliably captures physiological differences between wins and losses during gambling on electronic machines. Psychophysiology, 49(2), 154-163. MacLaren, V. V. (2015). Video Lottery is the Most Harmful Form of Gambling in Canada. Journal of Gambling Studies, 32(2), 459-85. doi: 10.1007/s10899-015-9560-z. Marmurek, H. H., Switzer, J., & D'Alvise, J. (2014). A comparison of university student and community gamblers: Motivations, impulsivity, and gambling cognitions. Journal of behavioral addictions, 3(1), 54-64. Meyer, G., Hauffa, B. P., Schedlowski, M., Pawlak, C., Stadler, M. A., & Exton, M. S. (2000). Casino gambling increases heart rate and salivary cortisol in regular gamblers. Biological psychiatry, 48(9), 948-953. Moodie, C., & Finnigan, F. (2005). A comparison of the autonomic arousal of frequent, infrequent and non‐gamblers while playing fruit machines. Addiction, 100(1), 51-59. Myrseth, H., Brunborg, G. S., & Eidem, M. (2010). Differences in cognitive distortions between pathological and non-pathological gamblers with preferences for chance or skill games. Journal of gambling studies, 26(4), 561-569. Nastally, B. L., Dixon, M. R., & Jackson, J. W. (2009). The effect of stopping devices and win rate on preference in slot machine players. Analysis of Gambling Behavior, 3, 27-30. Neighbors, C., Lostutter, T. W., Cronce, J. M., & Larimer, M. E. (2002). Exploring college student gambling motivation. Journal of Gambling Studies,18(4), 361-370.  67 Ontario Problem Gambling Helpline (2011). Gambling activities identified by individuals seeking problem gambling treatment services through OPGH (January 1, 2011 to December 31, 2011). Retrieved from the Ontario Problem Gambling Helpline (OPGH) database. Peters, H., Hunt, M., & Harper, D. (2010). An animal model of slot machine gambling: The effect of structural characteristics on response latency and persistence. Journal of Gambling Studies, 26(4), 521-531. Problem Gambling Institute of Ontario. (2011). Gambling by province – British Columbia. Retrieved from https://www.problemgambling.ca/EN/Documents/GbyP_BC2011.pdf Raylu, N., & Oei, T. P. (2004). The Gambling Related Cognitions Scale (GRCS): Development, confirmatory factor validation and psychometric properties. Addiction, 99(6), 757-769. Risen, J. L. (2015). Believing what we do not believe: Acquiescence to superstitious beliefs and other powerful intuitions. Psychological Review, 123(2), 182-207. Risen, J. L., Gilovich, T., & Dunning, D. (2007). One-shot illusory correlations and stereotype formation. Personality and Social Psychology Bulletin, 33(11), 1492-1502. Scarf, D., Miles, K., Sloan, A., Goulter, N., Hegan, M., Seid-Fatemi, A., ... & Colombo, M. (2011). Brain cells in the avian ‘prefrontal cortex’ code for features of slot-machine-like gambling. PloS one, 6(1), e14589. Sévigny, S., & Ladouceur, R. (2003). Gamblers' irrational thinking about chance events: The ‘double switching’ concept. International Gambling Studies, 3(2), 149-161. Sharpe, L., Tarrier, N., Schotte, D., & Spence, S. H. (1995). The role of autonomic arousal in problem gambling. Addiction, 90(11), 1529-1540. Skinner, B. F. (1953). Science and human behavior. Simon and Schuster.  68 Steenbergh, T. A., Meyers, A. W., May, R. K., & Whelan, J. P. (2002). Development and validation of the Gamblers' Beliefs Questionnaire. Psychology of addictive behaviors, 16(2), 143. Templeton, J. A., Dixon, M. J., Harrigan, K. A., & Fugelsang, J. A. (2015). Upping the reinforcement rate by playing the maximum lines in multi-line slot machine play. Journal of Gambling Studies, 31(3), 949-964. Van Selst, M., & Jolicoeur, P. (1994). A solution to the effect of sample size on outlier elimination. The quarterly journal of experimental psychology, 47(3), 631-650. Wiers, R. W. & Stacy, A. W. (2006). Handbook of implicit cognition and addiction. Thousand Oaks, CA: SAGE Publications Ltd. doi: 10.4135/9781412976237 Wilkes, B. L., Gonsalvez, C. J., & Blaszczynski, A. (2010). Capturing SCL and HR changes to win and loss events during gambling on electronic machines. International Journal of Psychophysiology, 78(3), 265-272.  69 Appendices Appendix A: Abridged Canadian Problem Gambling Index Section 1  First, we’d like to ask some questions about activities you may participate in. People bet money and gamble on many different things including buying lottery tickets, playing bingo, or card games with their friends. I am going to list some activities that you might have bet money on.   Questions Responses  Daily 2-6 times /week About once /week 2-3 times /month About once /month Between 6-11 times /year Between 1-5 times /year Never I prefer not to say a. In the past 12 months, how often did you bet or spend money on Lottery tickets like the 649, Super 7, or POGO? 1 2 3 4 5 6 7 8 9 b. In the past 12 months, how often did you buy daily lottery tickets like Pick 3? 1 2 3 4 5 6 7 8 9 c. In the past 12 months, how often did you buy instant win or scratch tickets like break open, pull tab, or Nevada strips? 1 2 3 4 5 6 7 8 9 d. In the past 12 months, how often did you buy raffle or fundraising tickets? 1 2 3 4 5 6 7 8 9 e. In the past 12 months, how often did you bet on horse races (i.e. live at the track and/or off-track)? 1 2 3 4 5 6 7 8 9 f. In the past 12 months, how often did you play bingo? 1 2 3 4 5 6 7 8 9            Yes No I prefer not to say       Screen for casino gambling: In the past 12 months, have you gambled at any type of casino including illegal or charity casino? 1 2  3                  Daily 2-6 times /week About once /week 2-3 times /month About once /month Between 6-11 times /year Between 1-5 times /year Never I prefer not to say g. In the past 12 months, how often did you bet or spend money on slot machines in a casino? 1 2 3 4 5 6 7 8 9 h. In the past 12 months, how often did you play poker in a casino? 1 2 3 4 5 6 7 8 9 i. In the past 12 months, how often did you play blackjack in a casino? 1 2 3 4 5 6 7 8 9  70 Questions Responses  Daily 2-6 times /week About once /week 2-3 times /month About once /month Between 6-11 times /year Between 1-5 times /year Never I prefer not to say j. In the past 12 months, how often did you play roulette in a casino? 1 2 3 4 5 6 7 8 9 k. In the past 12 months, how often did you play keno in a casino? 1 2 3 4 5 6 7 8 9 l. In the past 12 months, how often did you play craps in a casino? 1 2 3 4 5 6 7 8 9 m. In the past 12 months, how often did you play electronic gaming machines (EGMs), e.g. Game King, other than slot machines? 1 2 3 4 5 6 7 8 9 n. In the past 12 months, how often did you play a sports lottery like Sport Select (e.g. Pro Line, Over/Under, Point Spread)? 1 2 3 4 5 6 7 8 9 o. In the past 12 months, how often did you bet or spend money on sports pools? 1 2 3 4 5 6 7 8 9 p. In the past 12 months, how often did you bet on cards, or board games with family or friends? 1 2 3 4 5 6 7 8 9 q. In the past 12 months, how often did you bet or spend money on games of skill such as pool, bowling, or darts? 1 2 3 4 5 6 7 8 9 r. In the past 12 months, how often did you bet on arcade or video games? 1 2 3 4 5 6 7 8 9 s. In the past 12 months, how often did you gamble on the Internet? 1 2 3 4 5 6 7 8 9 t. In the past 12 months, how often did you bet on sports with a bookie? 1 2 3 4 5 6 7 8 9 u. In the past 12 months, how often did you personally invest in stocks, options, or commodities markets (does NOT include mutual funds or RRSPs)? 1 2 3 4 5 6 7 8 9      71  Section 2 Some of the next questions may not apply to you, but please try to be as accurate as possible. THINKING ABOUT THE LAST 12 MONTHS…  Questions Responses  Never Sometimes Most of the time Almost always a. Have you bet more than you could really afford to lose? 0 1 2 3 b. Still thinking about the last 12 months, have you needed to gamble with larger amounts of money to get the same feeling of excitement? 0 1 2 3 c. When you gambled, did you go back another day to try to win back the money you lost? 0 1 2 3 d. Have you borrowed money or sold anything to get money to gamble? 0 1 2 3 e. Have you felt that you might have a problem with gambling? 0 1 2 3 f. Has gambling caused you any health problems, including stress or anxiety? 0 1 2 3 g. Have people criticized your betting or told you that you had a gambling problem, regardless of whether or not you thought it was true? 0 1 2 3 h. Has your gambling caused any financial problems for you or your household? 0 1 2 3 i. Have you felt guilt about the way you gamble or what happens when you gamble? 0 1 2 3      72 Appendix B: Gamblers’ Beliefs Questionnaire Read each of the statement carefully. Rate to what extent you agree or disagree with each statement by circling a number.    Strongly agree Mostly agree Agree Neutral Disagree Mostly disagree Strongly disagree I think of gambling as a challenge. 1 2 3 4 5 6 7 My knowledge and skill in gambling contribute to the likelihood that I will make money. 1 2 3 4 5 6 7 My choices or actions affect the game on which I am betting. 1 2 3 4 5 6 7 If I am gambling and losing, I should continue because I don’t want to miss a win. 1 2 3 4 5 6 7 I should keep track of previous winning bets so that I can figure out how I should bet in the future. 1 2 3 4 5 6 7 When I am gambling, “near misses” or times when I almost win remind me that if I keep playing I will win. 1 2 3 4 5 6 7 Gambling is more than just luck. 1 2 3 4 5 6 7 My gambling wins are evidence that I have skill and knowledge related to gambling. 1 2 3 4 5 6 7 I have a “lucky” technique that I use when I gamble. 1 2 3 4 5 6 7 In the long run, I will win more money than I will lose gambling.  1 2 3 4 5 6 7 Even though I may be losing with my gambling strategy or plan, I must maintain that strategy or plan because I know it will eventually come through for me.  1 2 3 4 5 6 7 There are certain things I do when I am betting (for example, tapping a certain number of times, holding a lucky coin in my hand, crossing my fingers, etc.) which increase the chances that I will win.  1 2 3 4 5 6 7 If I lose money gambling, I should try to win it back.  1 2 3 4 5 6 7 Those who don’t gamble much don’t understand that gambling success requires dedication and a willingness to invest some money.  1 2 3 4 5 6 7 Where I get money to gamble doesn’t matter because I will win and pay it back. 1 2 3 4 5 6 7 I am pretty accurate at predicting when a “win” will occur.  1 2 3 4 5 6 7 Gambling is the best way for me to experience excitement. 1 2 3 4 5 6 7 If I continue to gamble, it will eventually pay off and I will make money. 1 2 3 4 5 6 7 I have more skills and knowledge related to gambling than most people who gamble.   1 2 3 4 5 6 7    73  Strongly agree Mostly agree Agree Neutral Disagree Mostly disagree Strongly disagree When I lose at gambling, my losses are not as bad if I don’t tell my loved ones.  1  2 3 4 5 6 7 I should keep the same bet even when it hasn’t come up lately because it is bound to win. 1 2 3 4 5 6 7       74 Appendix C: Game Experience Questionnaire Please indicate how you felt while playing the game you just finished for each of the items below.                        Not at all Slightly Moderately Fairly Extremely I was interested in the game’s theme. 0 1 2 3 4 I felt successful. 0 1 2 3 4 I felt bored. 0 1 2 3 4 I found the game impressive. 0 1 2 3 4 I forgot everything around me. 0 1 2 3 4 I felt frustrated. 0 1 2 3 4 I found it tiresome. 0 1 2 3 4 I felt irritable. 0 1 2 3 4 I felt skillful. 0 1 2 3 4 I felt completely absorbed. 0 1 2 3 4 I felt content. 0 1 2 3 4 I felt challenged. 0 1 2 3 4 I had to put a lot of effort into it. 0 1 2 3 4 I felt good. 0 1 2 3 4  75 Appendix D: Stopper Usage in Real Casinos Questionnaire 1. When you visit gambling venues, do you play slot machines with the stop button function?  ☐ Yes     ☐ No    ☐ Don’t know  2. If yes, how often do you make use of the stop button function?  ☐ All the time     ☐ Most of the time ☐ Some of the time     ☐ Rarely     76 Appendix E: Experimenter’s Instructions and Script Welcome/Consent:  Before we start the study, I will first go through some important information about the study with you and after that I will have you signed a consent form. This study requires you to fill out some surveys and play the slot machine for 20 minutes. In order to be eligible for the study, you must be 19 or above; must not have experienced problems with gambling either now or in the past; must not be taking any medication for any mental health problem, such as anti-depressant. Please confirm that you fulfill all of the requirements?  [If the person does not fulfill any one of the requirements, terminate the study now and debrief.]  In this study, we will be measuring physiological responses and will be attaching self-adhesive sensors that have signal gel on your body. Are you allergic to gels or adhesive? [If yes, terminate the study now and debrief.]  During the study, we will be recording a video. However, your face will not be in the frame. We are only interested in recording the screen of the slot machine.   The study will take about an hour of your time to complete. In appreciation for your time, we will pay you $10 ($15 for community participants). You may also earn a cash bonus from the slot machine play session, but the amount will depend on the remaining credit balance in the machine. [Flip to the last page of the consent letter.] At the end of the study, we will refer to this pay chart to determine the amount of your cash bonus.  Now, please take some time to read over the information letter. If you have no problem with the details listed on the letter, please print your name and sign on page 3.  CPGI:  Please fill out this survey and let me know when you are done. [Add up the scores of Section 2 of CPGI. If the total score exceeds 7, inform the participant about their possible problem gambler status and terminate the study and debrief.]  Wash hands:  During the experiment, we will be attaching sensors on your fingers, so we will need you to wash your hands. There is a sink at the end of the hallway. You only need to rinse your hands with water and please refrain from using soap. We have paper towels here for you to dry your hands.  Sensors and wire connection: [Turn on the Biopac wireless data receivers. Affix them onto the participant. The one that says EDA is for the wrist and the one that says ECG is for the waist.]  You will wear one of the receivers on your wrist and the other one on your waist.  Now I am going to have you attached three sensors on your body.   Two sensors will be attached just underneath the collarbone, one for each shoulder. Please feel for the soft spot just underneath your collarbone and attach a sensor there. Repeat on the other side of your body.  The third sensor will be placed just above your left hipbone.   Now I am going to attach some sensors on your fingers. [Stick the 2 self-adhesive EDA (skin conductance) sensors to the index and ring finger of the participant’s left hand, but do not clip on the wire just yet. Affix the finger pulse sensor on his/her middle finger as well.]   Slot machine instruction:  This is the machine that you will be playing. It is a real slot machine, which are the ones that you can find in the casino.  77  I am going to briefly explain the instructions of the slot machine session. You will be given $40 to play and you will insert the money to this slot. You will play the slot machine for 20 minutes and at the end of the session I will note down how many credits you have left in the machine and the amount of your cash bonus will be determined by the pay chart that I have shown you earlier.   $40 will give you 800 credits to start the game with. Each of the spin will cost you 9 credits to play.    This game has 9 pay lines. If you have three identical symbols fall onto the same pay line from left to right, you will earn some credits from the machine. If you have 4 or even 5 symbols aligned on the same line, you will earn even more credits. The amount of credits earned is determined by the symbols you have – usually the rarer the symbols are, the more the credits you’ll earn.   Now I am going to show you the paytable. Here are some of the symbols that are going to show up on the screen and their corresponding credit return sizes. For example, if you get three of this book symbol [bottom left corner] all line up on the same pay line, then you’ll get 25 credits. If you have four of them on the same pay line, you will get 50 credits and so on. The treasure chest and the golden dragon symbols [on top of the paytable] are wild symbols, which means that they can line up with other symbols of the same kind to give you a win.         78  Here are some of the other symbols that may show up on the machine.   This purple dragon symbol is a scattered symbol, which means that they don’t have to fall on any of the pay lines to give you a win. They only need to show up anywhere on the screen. For example, if you get 3 purple dragons anywhere on the screen, the machine will give you 10 free spins, almost like a bonus round.   This is the button that you will press to make spins on the machine. Another important thing to note is that this button also doubles as a stop button, which enables you to stop the spinning wheels at any time. You may use this stop button feature at any time during the session and you can use it as many times as you want.  During the game play, please press only the spin button and refrain from touching the screen or any other buttons on the machine. Also, please keep your left hand as still as possible during the entire session, so that we can get a clean physiological recording from you.   Now, you will have five practice spins.   Do you have any questions? [If not, give the participants $40 and instruct them to feed the bills one by one to the machine.]  You can begin your game session whenever you are ready. [Start timing them. They have 20 minutes to play.] [After they are done, stop the Acqknowledge recordings and then save it immediately. You can then  [When they are ready, ask them to remain on their seat and have them fill out the GBQ and GEQ in front of the slot machine. Meanwhile, calculate their cash bonus (if any).]  Please fill out these surveys. [After they are done, you can debrief them.]  This is the end of the study. You have earned a total of $ _______: $10 (or $15 for community participant) for your time and $ ______ cash bonus from the slot machine session. Please sign this receipt. This will be used as a record that you have received $ _______ from us.   This debrief letter explains what we are trying to look for from this study and if you have any question, please contact Stephanie. There is also a responsible gambling brochure attached at the back for your reference.   This is the end of the study. Thank you so much for participating.     79 Appendix F: Conversion Chart for Bonus Remuneration Slot Machine Pay Chart Credits at the end of the session What you get 0 No bonus 1 – 400 $2 401 – 800   $4 801 – 1200   $6 1201 – 1600 $8 1601 -- 2000 $10 More than 2000  $12  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0308678/manifest

Comment

Related Items