ArticlePDF Available

Abstract and Figures

This research examined similarities and differences between gambling activities, with a particular focus on differences in gambling frequency and rates of problem gambling. The data were from population-based surveys conducted in Canada between 2001 and 2005. Adult respondents completed various versions of the Canadian Problem Gambling Index (CPGI), including the Problem Gambling Severity Index (PGSI). A factor analysis of the frequency with which different gambling activities were played documented the existence of two clear underlying factors. One factor was comprised of Internet gambling and betting on sports and horse races, and the other factor was comprised of lotteries, raffles, slots/Video Lottery Terminals (VLTs), and bingo. Factor one respondents were largely men; factor two respondents were more likely to be women and scored significantly lower on a measure of problem gambling. Additional analyses indicated that (1) frequency of play was significantly and positively related to problem gambling scores for all activities except raffles, (2) the relationship between problem gambling scores and frequency of play was particularly pronounced for slots/VLTs, (3) problem gambling scores were associated with playing a larger number of games, and (4) Internet and sports gambling had the highest conversion rates (proportion who have tried an activity who frequently play that activity).
Content may be subject to copyright.
Gambling, Gambling Activities, and Problem Gambling
Thomas Holtgraves
Ball State University
This research examined similarities and differences between gambling activities, with a particular focus
on differences in gambling frequency and rates of problem gambling. The data were from population-
based surveys conducted in Canada between 2001 and 2005. Adult respondents completed various
versions of the Canadian Problem Gambling Index (CPGI), including the Problem Gambling Severity
Index (PGSI). A factor analysis of the frequency with which different gambling activities were played
documented the existence of two clear underlying factors. One factor was comprised of Internet gambling
and betting on sports and horse races, and the other factor was comprised of lotteries, raffles, slots/Video
Lottery Terminals (VLTs), and bingo. Factor one respondents were largely men; factor two respondents
were more likely to be women and scored significantly lower on a measure of problem gambling.
Additional analyses indicated that (1) frequency of play was significantly and positively related to
problem gambling scores for all activities except raffles, (2) the relationship between problem gambling
scores and frequency of play was particularly pronounced for slots/VLTs, (3) problem gambling scores
were associated with playing a larger number of games, and (4) Internet and sports gambling had the
highest conversion rates (proportion who have tried an activity who frequently play that activity).
Keywords: problem gambling, individual differences, gambling activities
People can and do gamble on virtually anything. Currently, the
most popular gambling activities are poker, sports betting, various
types of lotteries, bingo, parimutual wagering on (horse and dog)
races, casino games such as black jack and craps, slots, and a
variety of electronic gambling machines (e.g., video poker). Day
trading stocks on the Internet is a more recent addition to this list,
although one whose popularity may have already peaked.
The gambling activities in this (partial) list vary across many
dimensions. Some activities, such as poker involve a degree of
skill; others, such as lotteries, are purely random, chance events.
Some activities, such as poker and craps, are relatively social and
involve a degree of interaction that is sometimes intense and
focused; others, such as slots, are more solitary activities and are
generally pursued as such. The speed of play varies as well, from
craps and blackjack where the outcome is immediate, to weekly
lottery drawings or wagering on sporting events where the out-
come is more delayed. Moreover, gambling allows one to present
certain identities, and a large part of that identity is the game or the
games that one chooses to play (e.g., Holtgraves, 1988).
It would be surprising if these differences between gambling
activities were unimportant, yet research on gambling has often
overlooked them (but see Kessler et al., 2008; Wong & So, 2003).
For example, problem gamblers are often treated as a homogeneous
group, and the different pathways (e.g., different gambling activities)
through which one might become a problem gambler are ignored
(Blaszczynski & Nower, 2002). This is unfortunate, because dif-
ferent gambling activities may vary in terms of the type of person
they attract, as well as the role they play in the development of
pathological gambling. Hence, it is possible that different types of
people will engage in different gambling activities with different
subsequent effects.
Gambling and Individual Differences
Are there differences between people who prefer different gam-
bling activities? Research addressing this issue has been relatively
sparse. However, there has been some research examining differ-
ences between problem gamblers and nonproblem gamblers. There
were early mixed results reported for the traits of sensation seeking
(Anderson & Brown, 1984; Kuley & Jacobs, 1988) and locus of
control (Cameron & Myers, 1966; Ladouceur & Mayrand, 1984).
More recently, however, several studies have converged on showing
that problem gamblers tend to score higher on a cluster of traits
associated with the dimensions of impulsiveness and negative emo-
tionality (Bagby et al., 2007; Slutske, Caspi, Moffitt, & Poulton,
2005). It is possible, however, that this overall profile obscures
some important differences based on preferred gambling activities.
For example, it has been argued that problem gamblers can be
classified into subgroups based on their approach to arousal: a
subgroup that uses gambling as a means of augmenting arousal and
a subgroup that uses gambling as a means of reducing arousal
(Blaszczynski & Nower, 2002). Gambling activities clearly vary in
this regard; some are simple and solitary (e.g., slots) and promote
dissociative states that can serve to reduce arousal. Others are more
complex and social (e.g., craps) and can serve to augment arousal.
In one of the few attempts to examine differences in personality
traits for players of different games, Slowo (1998) found that
people who prefer to play the more exciting casino games were
This research was supported by a grant from the Ontario Problem
Gambling Research Centre. The statistical assistance of James Jones is
gratefully acknowledged.
Correspondence concerning this article should be addressed to Thomas
Holtgraves, Department of Psychological Science, Ball State University,
Muncie, IN 47306. E-mail:
Psychology of Addictive Behaviors © 2009 American Psychological Association
2009, Vol. 23, No. 2, 295–302 0893-164X/09/$12.00 DOI: 10.1037/a0014181
relatively higher on extraversion traits such as activity and excite-
ment. In contrast, poker machine players were significantly higher
on anxiety. Hence, the problem gambling trait of impulsiveness
was more evident in one subset of gamblers (those preferring
fast-paced casino games), and the trait of negative emotionality
was more evident in a different subset (those who preferred poker
More recent research has documented the existence of other
differences between people who prefer different gambling activi-
ties. For example, Petry (2003) asked participants who were seek-
ing admission to a state-run gambling treatment center to indicate
their most problematic form of gambling. Five major groups
emerged (sports, horse/dog racing, cards, slots, and lottery), and
these groups differed in several ways. First, there were clear
gender differences, with sports and horse/dog racing being almost
exclusively men and slot players twice as likely to be women.
Second, these groups differed in terms of gambling frequency
(lottery players gambled the most frequently and card players the
least) and amount of money gambled (lottery players the least and
horse/dog race gamblers the most). Finally, there were differences
in terms of substance abuse (substance abuse, especially alcohol,
was more common among sports betters) and psychiatric variables
(sports and card gamblers had fewer problems than the other
Differences Between Gambling Activities
Rather than focusing on differences between people who play
different games, it is possible to focus on differences between the
games themselves. One manifestation of this approach is the
argument that participation in some gambling activities is more
likely to result in problem gambling than participation in other
gambling activities. It has been argued, for example, that Elec-
tronic Gambling Machines (EGMs) are highly addictive (Produc-
tivity Commission, 1999). In this survey, conducted in Australia, it
was estimated that 22.6% of regular EGM gamblers had a signif-
icant gambling problem, a rate comparable to casino table games
(23.8%) but higher than racing (14.7%) and far higher than lotter-
ies (2.5%). It is very difficult to determine unambiguously the
addictive potential of a game, however. For example, high-
problem gambling rates for EGM players could be the result of
their playing other gambling activities. One alternative measure is
to compute the percentage of gamblers indicating that an activity
is their favorite (based on amount of money spent) who are
problem gamblers. With this measure, people who preferred play-
ing EGMs had the highest problem gambling rate (9.7%), followed
by racing (5.2%), casino gambling (3.5%), and lotteries (.3%).
Another measure is the weekly conversion rate, or percentage of
people who have played an activity who report playing that activ-
ity weekly. In the Productivity Commission report (1999), this rate
was 11.06% for EGMs, a rate lower than that for lotteries (48.5%)
but greater than that for casino table games (2.4%).
And another
measure is the percentage of problem and nonproblem gamblers
who engage in any particular activity. Not surprisingly, problem
gamblers are more likely to play EGMs than are nonproblem
gamblers (Smith & Wynne, 2004; Wynne, 2002), although this
finding is true for most gambling activities. Still, relative to other
activities, EGMs have been rated as one of the most popular
weekly activities for problem gamblers but not for nonproblem
gamblers (Volberg, 1997; Wynne, 2002). Taken together, these
measures suggest a relatively high addictive potential for EGMs.
Present Research
Prior research suggests that individuals who prefer, or at least
more frequently play, different gambling activities differ from one
another in some important ways. The purpose of the present
research was to explore these and other differences (and similar-
ities) in more detail. More specifically, in this research I pursued
the following two major issues. First, is there an underlying
structure for different gambling activities based on the frequency
with which they are played? In other words, do gambling activities
cluster together in any sort of meaningful way? For example, are
people more likely to play slot machines if they play the lottery
versus if they bet on sports? This type of analysis will be useful for
identifying similarities and differences between gambling activi-
ties, as well as the role played by these underlying dimensions in
the initiation and development of gambling and problem gambling.
Second, to what extent are different gambling activities associated
with different rates of problem gambling? This is obviously an im-
portant question, but one that is not amenable to a single, straightfor-
ward analysis. There are no completely unambiguous measures in this
regard. Accordingly, in this research I used a variety of different
analyses and searched for common patterns across these analyses.
First, I examined differences between gambling activities in terms of
their conversion rates and levels of problem gambling. Second, I
focused on differences between people in terms of their problem
gambling status, and whether these differences were associated with
preferences for certain gambling activities and with the number of
gambling activities that one played.
To examine these issues, I used a large, integrated data set
comprised of responses to gambling surveys conducted in several
Canadian provinces between 2001 and 2005. The use of this type
of population-based survey data is important because participants
in many studies in this area have been problem gamblers seeking
treatment (e.g., Petry, 2003). Hence, there is a clear need to
explore these differences with population-based data.
The integrated data set for this study was made available by the
Ontario Problem Gambling Research Centre. This data set con-
sisted of responses to telephone surveys regarding gambling that
had been conducted in several different Canadian provinces be-
tween 2001 and 2005. Respondents were adults (18 or 19 years of
Conversion rates are particularly susceptible to the availability of an
activity and hence should be interpreted with caution.
There has been some dispute over these findings, however (e.g.,
Dowling, Smith, & Thomas, 2005). Most notably, Mizerski and colleagues
(Mizerski, Jolley, & Mizerski, 2002) have argued that the high percentage
of people who are heavy users of EGMs is no different from the distribu-
tion of heavy-use consumers for any consumer product (i.e., 80% of use is
accounted for by 20% of the users). This analysis, however, does not
involve different gambling activity comparisons, only EGM distributions
versus expected consumer behavioral distributions.
age or older) randomly selected with various constraints (e.g.,
stratified by region) in order to approximate the demographic
breakdown for that area. Some of the surveys weighted their
sample based on certain demographic considerations. However,
the assignment of weights was not consistent over these different
surveys and hence they were not used in the present analyses. The
sample size of the combined data set was 21,374. Of these respon-
dents, 12,299 had gambled at least once during the past 12 months
and hence were eligible for inclusion in the present analyses.
Information regarding the procedures used for each survey is
presented in Table 1. Included in this table are references and
URLs for each survey (all survey reports are available online).
Although each survey was created and conducted indepen-
dently, the survey protocol always consisted of a version of the
Canadian Problem Gambling Index (CPGI), a comprehensive set
of questions regarding participation in a variety of gambling ac-
tivities, as well as background questions, substance abuse issues
and a variety of additional demographic variables. There were
some differences between the surveys in terms of the content and
wording of the CPGI items. However, all analyses reported here
are based on identically worded questions (the specific wording for
all analyzed items is given in the Results section).
One important component of the CPGI is a scale designed to
assess problem gambling. This scale consists of nine items and is
referred to as the Problem Gambling Severity Index (PGSI). Each
of the surveys, with one exception, contained these nine items
worded in an identical manner. The one exception was the Na-
tional survey in which a dichotomous (rather than four-response)
format was used for two of the PGSI items. This survey was
excluded from all analyses that included the PGSI. The PGSI was
designed to measure a single, problem gambling construct in a
general population rather than in a clinical context (unlike its
Table 1
Surveys Included in Combined Data Set
Survey province: Alberta
Survey year: 2002
Sample size: 1,804
Reference: Smith & Wynne (2002)
Online report:
Conducted by: The University of Alberta’s Population Research Lab
Stratified: By region
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Reported response rate: 63.6%
Margin of error (95% CI at maximum variance): 2.3%
Survey province: Ontario
Survey year: 2001
Sample size: 4,631
Reference: Wiebe, Single, & Falkowski-Ham (2001)
Online report:
Conducted by: Viewpoints Research Inc.
Stratified: By region, gender, age
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Margin of error (95% CI at maximum variance): 1.4%
Reported response rate: 37%
Survey province: Ontario
Survey year: 2005
Sample size: 3,604
Reference: Wiebe, Mun, & Kauffman (2006)
Online report:
.pdf?docid 7670
Conducted by: Hitachi Survey Research Centre in the Department of
Sociology at the University of Toronto at Mississauga
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Reported response rates: (strict) 82.5% (optimal); 46.4%
Margin of error: Not reported
Survey Province: Manitoba
Survey year: 2001
Sample Size: 3,119
Reference: Patton, Dhaliwal, Pankratz, & Broszeit (2002)
Online report:
Conducted by: Addictions Foundation of Manitoba
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Reported response rate: 40.7%
Margin of error (95% CI at maximum variance): 1.0%
Survey province: British Columbia
Survey year: 2002
Sample Size: 2,500
Reference: Ipsos-Reid & Gemini Research (2003)
Online report:
Conducted by: Ipsos-Reid
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Stratified: By region
Margin of error (95% CI at maximum variance): 2%
Reported response rate: 27%
Survey province: Newfoundland
Survey year: 2005
Sample size: 2,596
Reference: Market Quest Research Group (2005)
Online report:
Conducted by: Market Quest Research Group
Stratified: By region, gender
Sampling households: Random-digit dialing
Sampling within households: Adult (19 years and older) with most
recent birthday
Reported response rate: 25.4% (computed following Scenario C of the
Marketing Research and Intelligence Association [MRIA] of
Margin of error (95% CI at maximum variance): 1.92%
Survey province: National
Survey year: 2001
Sample size: 3,120
Reference: Ferris & Wynne (2001)
Online report:–047a-41ec-
Conducted by: Institute for Social Research at York University
Stratified: By region
Sampling households: Random-digit dialing
Sampling within households: Adult with most recent birthday
Reported response rate: Not reported
CI: Not reported
Note. CI confidence interval.
nearest competitor, the South Oaks Gambling Screen). The mea-
sure was theoretically derived and assesses problem gambling
behaviors (e.g., Chasing: “How often have you gone back another
day to try to win back the money you lost?) and the occurrence of
adverse consequences of gambling (e.g., Guilt: “How often have
you felt guilty about the way you gamble or what happens when
you gamble?”). For each of the nine items, respondents answer on
a four-alternative scale: 0 never;1sometimes;2most of the
time;3Almost always. Responses were summed and, following
convention, respondents were classified into gambling subtypes
based on their PGSI scores as follows: 0 nonproblem gambler;
1–2 low-risk gambler; 3–7 moderate-risk gambler; 8 and
over problem gambler.
Although the PGSI was developed independently, there is some
item overlap with the SOGS. Research on the PGSI indicates that
it has adequate internal consistency and test–retest reliability (Fer-
ris & Wynne, 2001), and that it assesses a single, underlying
problem gambling construct that is correlated with various gam-
bling behaviors (Holtgraves, in press).
Unless otherwise noted, all reported analyses were based on
participants who indicated that they had gambled at least once in
the past 12 months. All analyses were performed with SPSS
(version 15). The Ns varied over analyses, because not all ques-
tions were included in all surveys, and because analyses differed in
their inclusion criteria.
Factor Analysis of Gambling Frequency
The first issue concerned the possibility that there is an under-
lying structure to the frequency with which gambling activities are
played. A principal components analysis (PCA) of the frequency
(0 never;1less than once a month;2at least once a month;
3at least once a week;4daily) with which respondents
reported engaging in eight different gambling activities (lottery,
horse race betting, Internet gambling, bingo, raffles, sport select,
slots/VLTs, bookie)
was conducted on the sample of respondents
with nonmissing data on these eight activities (N10,685).
Because responses to the PGSI constituted ordinal data, a poly-
choric correlation matrix was first constructed and used as matrix
input for the factor analysis (Gilley & Uhlig, 1993; Joreskog &
Moustaki, 2000). In this analysis, the PCA yielded two factors with
eigenvalues greater than one (3.28 and 1.35). Together, these two
factors accounted for 57.9% of the variance.
The factor structure (varimax rotation) that emerged was very
clear (see Table 2), with horse racing, sport select, Internet, and
bookie wagering all loading highly on the first factor and lottery,
bingo, slots/VLTs, and raffles all loading highly on the second
factor. All loadings were greater than .61 on the primary factor and
less than .3 on the secondary factor. This structure suggests that
certain people prefer to play certain types of games, one group that
prefers to play lottery, bingo, slots, and raffles, and a second group
that prefers gambling on the Internet, horse races, and sports.
I conducted exploratory analyses of the characteristics of re-
spondents based on within which factor their favorite gambling
activity fell. Favorite gambling activity was defined as the gam-
bling activity played more frequently than any other gambling
activity. The majority of the factor one group were men (78%), and
the majority of the factor two group were women (52%). In
addition, mean scores on the PGSI were significantly higher for the
factor one group (M.867) than for the factor two group (M
.397), F(1, 4084) 15.83, p.001.
Importantly, this difference
occurred for both men (1.12 vs. .46) and women (.61 vs. .34),
demonstrating that the problem gambling difference between fac-
tors is not a function of greater male participation in factor one and
female participation in factor two; the Factor Gender interaction
was not significant, F(1, 4084) 2.82, p.09.
Differences Between Gambling Activities
Conversion rates. Conversion rates provide an estimate of the
likelihood that individuals who try a particular gambling activity
will come to frequently engage in that activity. Hence, conversion
rates are estimated by computing the ratio of frequent players to
players who have ever played an activity. The conversion rates
(weekly or more) for eight activities were computed and are
presented in Table 3. The numbers in this table refer to the
percentage of players who report having ever played an activity
who currently play that activity at least weekly. For example,
11.95 % of people who have played some form of lotto in the past
12 months play lotteries weekly or more frequently. In contrast,
the conversion rate for raffles is 1.87. The activities clearly differ
in their conversion rates. Most notably, there is a group of three
activities that have conversion rates significantly greater than the
other activities: bookie, sport select, and Internet. This means that
people who have tried these three activities are far more likely to
be frequent players than people who have tried other activities.
Problem gambling differences. This analysis was concerned
with whether there were differences between gambling activities in
terms of problem gambling rates. For example, what percentage of
people for whom activity X is their favorite gambling activity are
Casino gambling was not included in this list because items that
assessed it were not included in some surveys, and because surveys that did
include such items assessed different aspects of casino gambling.
Factor analyses were conducted separately for men, and the results
were identical to the overall analysis. It was not possible to conduct a
separate analysis for women because the distribution of gambling frequen-
cies prevented the generation of a polychoric correlation matrix.
The sample size is reduced because of missing responses on PGSI
items and because of the exclusion of respondents with multiple favorite
Table 2
Factor Structure (Loadings) for Frequency of Engaging in
Different Gambling Activities
Activity Factor 1 Factor 2
Sport select .755 .220
Internet .725 .044
Bookie .880 .133
Horse racing .704 .267
Lottery .162 .775
Raffle .112 .606
Bingo .057 .736
Slots/VLT .296 .700
classified as nonproblem gamblers, low-risk gamblers, moderate-
risk gamblers, and problem gamblers? First, the distribution of
gambling subtypes as a function of favorite gambling activity (as
defined earlier) was computed and the results are presented in
Table 4. A
analysis indicated that the distribution of gambling
subtypes varied over gambling activities,
(21, N4088)
334.59, p.0001. As can be seen in Table 4, rates of nonprob-
lematic gambling were relatively high (78%) for lotteries, raf-
fles, and horse race betting; moderate for bingo and slots/VLTs;
and low (64%) for sport select, Internet and bookies, with
nonproblem gambling 50% or less for the latter two activities.
A one-way analysis of variance (ANOVA) was then conducted
with favorite activity as the independent variable and PGSI score
as the dependent measure. Consistent with the
test, PGSI scores
varied significantly as a function of favorite activity, F(7,80.78)
22.25, p.001.
These results can be seen in the far right column
in Table 4. Post hoc tests using the Dunnett’s C procedure indi-
cated that people for whom playing raffles was the favorite activity
had significantly lower PGSI scores than participants whose fa-
vorite activity was playing slots, lotto, bingo, or sport select. PGSI
scores were elevated for those who favored the Internet and
bookies but the differences were not significant at p.05.
Two parallel analyses were then conducted to examine problem
gambling as a function of frequent participation (weekly or more)
in a gambling activity (regardless of whether the activity was a
person’s favorite). As can be seen in Table 5, the results generally
paralleled the findings for favorite activity. A
analysis indicated
that the distribution of gambling subtypes varied over gambling
(21, N933) 101.28, p.0001. Again, problem
gambling rates were lowest for raffles and highest for gambling
with bookies. There were some differences, however, between the
results of this analysis and those that classified participants based
on their favorite activity. Most notable was the relatively high
problem gambling rate for people who play slots/VLTs on a
frequent basis. Fully 66% of those engaging in this activity on at
least a weekly basis are low risk or greater. This suggests that
frequent slot/VLT play is accompanied by the more frequent
playing of another gambling activity (which is then the favorite).
An ANOVA on PGSI scores (see far right column in Table 5) was
consistent with this pattern and significant, F(7, 84.86) 3.62,
PGSI scores were quite elevated (and significantly
different from other participants, with the exception of frequent
Internet and bookie gambling) for people who frequently played
slots/VLTs. PGSI scores were elevated for frequent Internet and
bookie gamblers, but not significantly greater than other participants.
Differences Between People
Several issues regarding gambling differences as a function of
gambling subtype were examined. First, do problem gamblers
focus and play only one activity as some have previously sug-
gested (e.g., Breen & Zimmerman, 2002; Petry, 2003)? Or are they
more likely to play multiple games (Kessler et al., 2008)? A
composite gambling measure was created representing the number
of the eight different gambling activities one had engaged in
during the past 12 months. This measure was analyzed as a
function of gambling subtype. There was a significant effect of
gambling type for this measure, F(3, 2900) 75.94, p.001, and
follow-up post-hoc tests indicated that all means were significantly
different ( p.05) from one another, with the exception of the
difference between moderate risk and problem gamblers. As can
be seen in figure 1, there is an increase in the number of gambling
activities that individuals engage in as problem gambling severity
increases. In short, it does not appear to be the case that problem
gamblers concentrate on a single game (at least relative to less
problematic gamblers).
One potential limitation with the prior analyses is that partici-
pants often play multiple games, thereby making it difficult to
isolate the relationship between individual games and problem
gambling. To help overcome this, a multiple regression analysis
was conducted in which total PGSI score was treated as the
criterion variable and frequency of play for each gambling activity
served as the predictor variables. All variables were entered si-
The variances differed over gambling groups (and the size of each
group differed widely), and Levene’s test for the equality of variances was
significant ( p.01). Because of this violation of the homogeneity of
variance assumption, Welch’s Fwas computed for the omnibus test and
Dunnett’s C for all post hoc comparisons.
The sample size is reduced because of missing responses on PGSI
items and because of the exclusion of respondents with multiple favorite
Table 3
Conversion Rates for Eight Gambling Activities
Activity (n)
Conversion rate
(weekly or greater)
95% Confidence
Lotto (4,845) 11.95 11.08–12.89
Raffles (9,030) 1.87 1.60–2.17
Bingo (1,939) 17.48 15.85–19.23
Slots/VLT (3,372) 2.76 2.26–3.37
Horse Race (913) 5.15 3.90–6.78
Sport select (849) 23.20 20.49–26.16
Bookie (53) 30.19 19.52–43.54
Internet (149) 20.81 15.07–28.02
Values represent the percentage of people who have played an
activity who play that activity weekly or more.
Table 4
Distribution of Gambling Subtypes and Mean Problem
Gambling Severity Index (PGSI Scores) for Each Favored
Gambling Activity
Activity (n)
Gambling subtype, %
PGSINonproblem Low risk
risk Problem
Lottery (1,221) 81 12 6 1 .54
Raffle (1,915) 93 6 1 1 .13
Horses (71) 78 10 11 1 .77
Bingo (214) 67 22 10 1 .87
Slots/VLT (469) 74 16 8 2 .85
Sport select (160) 64 27 6 3 .92
Internet (32) 50 34 12 3 1.56
Bookie (6) 17 50 17 17 3.33
Note. Means without a superscript in common are significantly different
at p.05 using Dunnett’s C.
multaneously, thereby assessing the effects of each game control-
ling for the effects of all other games. The results are reported in
Table 6. The overall model was significant, R.335, F(8,
7404) 117.17, p.001. Each gambling activity, with the
exception of playing raffles, contributed positively and signifi-
cantly to the problem gambling scores. The size of the beta weights
was relatively modest and ranged between .18 and .05. Consistent
with the analysis of frequently played games, frequency of playing
slots/VLTs was the activity most highly associated with problem
gambling. Importantly, in this case the slots–problem gambling
relationship held when the effects of other gambling activities
were controlled.
Limitations of the present research should be noted at the outset.
First, the response rates for some of the surveys were relatively
low, although generally comparable to rates typically reported for
gambling surveys (Gemini Research, 1994; Shaffer, Hall, &
Vander Bilt, 1997). Second, the surveys differed in various ways
(although not the specific items that were analyzed in this report)
and were conducted for different reasons. Moreover, the surveys
were conducted in different locales that varied in terms of gam-
bling availability. Third, consistent with other research (e.g.,
Kessler et al., 2008), the occurrence of problem gambling was
relatively infrequent, and breaking down problem gambling as a
function of favored gambling activity reduced the sample even
more, thus resulting in some data instability. Within the context of
these limitations, the present research generated several important
First, even though gambling activities vary on a range of di-
mensions, there do appear to be some underlying similarities, so
much so that two clearly defined groups were identified in this
research. One group is comprised of Internet gambling and betting
on sports and horse races. The other group is comprised of slots/
VLTs, raffles, lotteries, and bingo. These activities vary on a
number of dimensions, but the group two activities generally tend
to be low-wager activities (i.e., the amount that can be wagered on
any single outcome is relatively low), relative to group one, which
typically allows for much higher wagers, and this may partially
account for the high problem gambling score for the former group
relative to the latter. Even though problem gambling scores were
lower for group two, there was one activity within this group—
slots/VLT— that was associated with higher rates of problem
gambling (as will be discussed subsequently). Note also that the
gender difference (i.e., more women in group two and men in
group one) is consistent with other reports of gender differences in
gambling preferences (e.g., Petry, 2003).
Second, consistent with the difference in problem gambling
scores for the two gambling groups, there were differences be-
tween gambling activities in terms of their conversion rates. Spe-
cifically, conversion rates for Internet gambling and sports betting
(both sport select and betting with bookies) were far higher than
they were for the other activities. The percentage of people who
tried gambling on the Internet or sports betting and continued to
gamble frequently on the Internet or sports was very high. There is
no doubt, however, that conversion rates are influenced by gam-
bling availability. It is not possible to gamble frequently on an
activity if that activity is not available or is difficult to access. But
to a certain extent that is just the point. Increased availability will
Table 5
Distribution of Gambling Subtypes and Mean (SD) Problem
Gambling Severity Index (PGSI) Scores for Gambling Activities
That Are Played Frequently (at Least Weekly)
Activity (n)
Gambling subtype, %
risk Problem
Lottery (287) 69 18 9 4 1.0
Raffle (114) 77 13 7 3 .82
Horses (30) 63 23 10 3 1.0
Bingo (258) 63 22 12 3 1.19
Slots/VLT (68) 34 20 28 18 4.50
Sport Select (141) 54 30 12 4 1.36
Internet (24) 42 38 17 4 1.96
Bookie (8) 0 63 13 25 4.75
Note. Means without a superscript in common are significantly different
at p.05 using Dunnett’s C. PGSI SDs are presented in parentheses.
Mean of number of different games
Figure 1. Number of different games played as a function of problem
gambling subtype.
Table 6
Multiple Regression Analysis: Problem Gambling Severity Index
(PGSI) Scores Predicted by Frequency of Gambling Activity
Activity rBt
Lottery .152
.08 7.00
Raffle .004 .03 3.02
Horses .128
.05 4.06
Bingo .161
.11 9.53
Slots/VLT .240
.18 15.82
Sport select .147
.08 7.26
Internet .107
.07 6.30
Bookie .141
.10 8.43
result in increased gambling. For example, nothing is more avail-
able than the Internet; a player doesn’t even need to leave home.
And although the difference was not significant in the present
study, other researchers have reported elevated problem gambling
scores (on the SOGS) for people who gamble on the Internet (Ladd
& Petry, 2002).
Third, the relationship between problem gambling (as assessed
with the PGSI) and gambling activity varied as well. The major
difference was this. People who play raffles more frequently than
other games had significantly lower problem gambling scores than
people who preferred other games. This occurred when problem
gambling was treated as a categorical variable and when it was
treated as a continuous variable. And when problem gambling was
treated as a continuous variable, every gambling activity, except
raffles, contributed positively, independently, and significantly, to
problem gambling scores. In other words, more frequent playing of
any game was associated with increased problem gambling scores.
Fourth, some of the analyses suggest that frequent slots/VLTs
play is associated with increased problem gambling. Almost two-
thirds of the people who played slots/VLT on a weekly basis were
low risk or greater. And in the multiple regression analysis, slots/
VLTs had the largest beta weight of all gambling activities. These
data are consistent with the general findings of the Productivity
Commission (1999) regarding the enhanced addictive potential of
video gambling. Note that in the present research the slots/VLT
category is broader than the video gambling category considered
by the Productivity Commission (1999). Note also that the rela-
tionship between problem gambling and slots/VLTs occurred for
the frequency measures but not the favorite activity measure.
Problem gambling (low risk or greater) is heightened for people
who play slots/VLTs frequently (independent of whether it is their
most frequently played activity). And of course this is consistent
with the finding that problem gambling scores were positively
correlated with the number of different activities played.
Still, there must also be something about slots/VLTs that is
contributing to problem gambling, because for no other activity
was there such a large discrepancy in problem gambling rates
between the favored and frequently played analyses. Likely vari-
ables in this regard include playing speeds and payout interval
(Griffiths, 1993, 1999). The payout interval for slots is very short,
almost instantaneous; for raffles and lotteries it is much longer.
Even for casino games such as craps and blackjack the payout
interval tends to be longer because of the presence of other people.
Short intervals facilitate chasing, one of the defining features (if
not the most defining feature) of problem gambling. Quick pay-
outs, combined with fast playing speeds, may facilitate behavior
characteristic of problem gambling. Subsequent research attempt-
ing to identify those features of gambling activities that play
substantial roles in problem gambling is warranted.
The present research documented the existence of differing rates
of frequent gambling (conversion rates) and problem gambling for
different gambling activities. As with all correlational research,
however, the causal direction is unknown. People are not randomly
assigned to play different gambling activities. As a result, there
may be certain types of people inclined to participate in certain
activities, and it is that inclination that is critical rather than
anything about the activities themselves. But that seems too simple
as well. Gambling activities differ in their affordances, in what
they provide for those who choose to play them. Some activities
(bingo) offer a chance to socialize, other activities (craps) a chance
for intense and focused excitement. And people choose to play and
to continue to play those activities that mesh well with their
personalities (in addition to the activity’s availability, affordabil-
ity, and so on). So it is largely a Person Situation interaction that
will account for these patterns, an interaction that has yet to be
investigated in any systematic way.
Anderson, G., & Brown, R. (1984). Real and laboratory gambling:
Sensation-seeking and arousal. British Journal of Psychology, 75, 401–
Bagby, R. M., Vachon, D., Bulmash, E. L., Toneatto, T., Quilty, L. C., &
Costa, P. T. (2007). Pathological gambling and the five-factor model of
personality. Personality and Individual Differences, 43, 873– 880.
Blaszczynski, A., & Nower, L. (2002). A pathways model of problem and
pathological gambling. Addiction, 97, 487– 499.
Breen, R. B., & Zimmerman, M. (2002). Rapid onset of pathological
gambling in machine gamblers. Journal of Gambling Studies, 18, 31– 43.
Cameron, B., & Myers, J. (1966). Some personality correlates of risk
taking. The Journal of General Psychology, 74, 51– 60.
Dowling, N., Smith, D., & Thomas, T. (2005). Electronic gaming ma-
chines: Are they the “crack cocaine” of gambling? Addiction, 100,
33– 45.
Ferris, J., & Wynne, H. (2001). The Canadian problem gambling index:
Final report. Ottawa: Canadian Centre on Substance Abuse.
Gemini Research (1994). Social gaming and problem gambling in British
Columbia. Report to the British Columbia Lottery Corporation.
Northampton, MA: Author.
Gilley, W. F., & Uhlig, G. E. (1993). Factor analysis and ordinal data.
Education, 14, 258 –264.
Griffiths, M. D. (1993). Fruit machine gambling: The importance of
structural characteristics. Journal of Gambling Studies, 9, 101–120.
Griffiths, M. D. (1999). Gambling technologies: Prospects for problem
gambling. Journal of Gambling Studies, 15, 265–283.
Holtgraves, T. (1988). Gambling as self-presentation. Journal of Gambling
Behavior, 4, 78 –91.
Holtgraves, T. (in press). Evaluating the Problem Gambling Severity Index.
Journal of Gambling Studies.
Ipsos-Reid & Gemini Research. (2003). British Columbia problem gam-
bling prevalence study. Victoria, BC: Ministry of Public Safety and
Solicitor General.
Joreskog, K. G., & Moustaki, I. (2000). Factor analysis of ordinal vari-
ables: A comparison of three approaches. Multivariate Behavioral Re-
search, 36, 347–387.
Kessler, R. C., Hwang, I., LaBrie, R., Petukhova, M., Sampson, N. A.,
Winters, K. C., et al. (2008). The prevalence and correlates of DSM-IV
pathological gambling in the national comorbidity survey replication.
Psychological Medicine, 38, 1351–1360.
Kuley, N. B., Jacobs, D. F. (1988). The relationship between dissociative-
like experiences and sensation seeking among social and problem gam-
blers. Journal of Gambling Behavior, 4, 197–207.
Ladd, G. T., & Petry, N. M. (2002). Disordered gambling among
university-based medical and dental patients: A focus on Internet gam-
bling. Psychology of Addictive Behaviors, 16, 76 –79.
Ladouceur, R., & Mayrand, M. (1984). Evaluation of the “illusion of
control”: Type of feedback, outcome sequence, and number of trials
among regular and occasional gamblers. Journal of Psychology, 117,
37– 46.
Market Quest Research Group Inc. (2005). Newfoundland and Labrador
gambling prevalence study. Prepared for the Department of Health and
Community Services. St. John’s, Newfoundland: Government of New-
foundland and Labrador.
Mizerski, D., Jolley, B., & Mizerski, K. (2002). Disputing the “crack
cocaine of gambling” label for electronic gaming machines. In A.
Blaszczynski (Ed.), Culture and the gambling phenomenon. Proceedings
of the 11th Conference of the National Association of Gambling Studies
(pp. 276 –283). Melbourne, Australia: National Association for Gam-
bling Studies.
Patton, D., Brown, D., Dhaliwal, J., Pankratz, C., & Broszeit, B. (2002).
Gambling involvement and problem gambling in Manitoba. Manitoba,
Canada: Addictions Foundation of Manitoba.
Petry, N. M. (2003). A comparison of treatment-seeking pathological
gamblers based on preferred gambling activity. Addiction, 98, 645– 655.
Productivity Commission. (1999). Australia’s gambling industries. Can-
berra, Australia: AusInfo.
Shaffer, H. J., M. N. Hall, & J. Vander Bilt. (1997). Estimating the
prevalence of disordered gambling behavior in the United States and
Canada: A meta-analysis. Boston, MA: Harvard Medical School Divi-
sion on Addictions.
Slowo, D. (1998). Are all gamblers the same? An exploration of person-
ality and motivational characteristics of individuals with different gam-
bling preferences. In G. Coman, B. Evans, & R. Wooten (Eds.), Respon-
sible gambling: A future winner. Proceedings of the 8th conference of
the National Association for Gambling Studies (pp. 339 –351). Adelaide:
National Association for Gambling Studies.
Slutske, W. S., Caspi, A., Moffitt, T. E., & Poulton, R. (2005). Personality
and problem gambling: A prospective study of a birth cohort of young
adults. Archives of General Psychiatry, 52, 769 –775.
Smith, G. J., & Wynne, H. J. (2004). VLT Gambling in Alberta: A
preliminary analysis. Retrieved December 16, 2007, from https://
Smith, G. J., & Wynne, H. J. (2002). Measuring gambling and problem
gambling in Alberta using the Canadian Problem Gambling Index
(CPGI): Final report. Edmonton: Alberta Gaming Research Institute.
Volberg, R. A. (1997). Gambling and problem gambling in Oregon. Report to
the Oregon Gambling Addiction Treatment Foundation. Retrieved Decem-
ber 16, 2007, from
Wiebe, J., Mun, P., & Kauffman, N. (2006). Gambling and problem
gambling in Ontario 2005. Toronto: Responsible Gambling Council
Wiebe, J., Single, E., & Falkowski-Ham, A. (2001). Measuring gambling
and problem gambling in Ontario. Toronto: Canadian Centre on Sub-
stance Abuse and Responsible Gambling.
Wong, I. L., & So, E. M. (2003). Prevalence estimates of problem and
pathological gambling in Hong Kong. American Journal of Psychiatry,
160, 1353–1354.
Wynne, H. J. (2002). Gambling and problem gambling in Saskatchewan.
Ottawa: Canadian Center on Substance Abuse.
Received May 21, 2008
Revision received September 15, 2008
Accepted September 17, 2008
... Lotteries, scratch cards, and slot machines are the commonest types of gambling products in the United Kingdom (Gambling Commission, 2017), whereas poker, bingo, sports betting, casinos, lotteries, animal races, and electronic gambling machines are popular in North America (Holtgraves, 2009). In Australia, sports betting, horse betting, electronic gambling machines, and casinos are the most common forms of gambling (McCarthy et al., 2018). ...
... Studies have reported that men typically prefer strategic games (e.g., betting on sports or cards, casino games), whereas women tend to prefer non-strategic types of gambling that are determined by luck (e.g., lotteries, lotto, electronic gambling machines; Holdsworth et al., 2012;Holtgraves, 2009;Petry, 2003). Data from cross-sectional studies in Finland (Salonen et al., 2017) and Norway (Hanss et al., 2014) showed that men hold more favourable attitudes towards gambling than women do. ...
... This could be due to the popularity of European football leagues among Nigerians and because the majority of our respondents were men. Previous studies revealed that gambling players tend to prefer sports betting, which they see as a strategic game compared with other gambling platforms (Holdsworth et al., 2012;Holtgraves, 2009;Petry, 2003). In addition, a high rate of unemployment and inadequate government social welfare services could be factors that contribute to gambling as a way to make ends meet or a means to escape poverty among youth wagers. ...
Full-text available
In this cross-sectional survey study, we investigated gambling characteristics (''quick'' money syndrome, frequency of gambling, preferred gambling products and platforms) and demographic (age and gender) differences as determinants of attitudes towards gambling among youths in Lagos. We used a purposive snowball technique to recruit 179 respondents in Lagos (men: n = 165 [83.8% of participants]; women: n = 32) with a history of gambling activities. Gambling characteristics, demographics, and attitudes towards gambling were measured by using a gambling characteristics profile, the bio-data of the respondents, and the short form of the Attitudes Towards Gambling Scale (ATGS-8), respectively. The results showed that respondents believed that to make quick money, one needed to gamble once a week or more (67%), wager on sports betting (37.1%), and bet online (65.5%). The mean ATGS-8 composite score (28.2 ± 4.75) indicated overall positive attitudes among respondents. Independent sample t tests showed a significant difference between older (24-34 years) and younger (18-23 years) youths in their attitudes towards gambling (t =-2.30, p o .05) but no significant gender differences (t = 0.06, p 4 .05). One-way analysis of variance revealed a significant difference in attitudes towards gambling based on gambling frequency, F(3, 196) = 6.86, p o .05, with those who gamble monthly having the highest score. Participants displayed the belief that the easiest way to quick money is to gamble at least once weekly and to bet online. Younger participants and those who gamble at least once a month reported more positive attitudes towards gambling. Youths need to be made aware that gambling is not a viable source of income.
... McManus & Graham [16] analyse perceptions of gambling and the relationships between viewing horse races and gambling on these events, highlighting the variation in attitudes to gambling and the variety of experiences among gambling participants. Holtgraves [17] analyse the general issue of problem gambling across 5 domains including sports and horse racing. Wardle et al. [18] surveyed the gambling habits of [16][17][18][19][20][21][22][23][24] year olds in the UK and found that 10.8% had bet online on sports and 2.3% in a bookmakers in the last year, and 7% had bet online on horse racing and 5.8% in a bookmakers in the last year. ...
... Holtgraves [17] analyse the general issue of problem gambling across 5 domains including sports and horse racing. Wardle et al. [18] surveyed the gambling habits of [16][17][18][19][20][21][22][23][24] year olds in the UK and found that 10.8% had bet online on sports and 2.3% in a bookmakers in the last year, and 7% had bet online on horse racing and 5.8% in a bookmakers in the last year. We analyse gambling and its perceptions as a driver for horse racing engagement. ...
... The evidence regarding the relationship between problem gambling severity and gambling attitudes was somewhat conflicting in the studies included in the present synthesis. Considering the association between gambling frequency and gambling attitudes, it would be conceivable to expect that problem gambling severity would be associated with more positive attitudes as noted by four included studies (Donaldson et al., 2015;Gavriel-Fried, 2015;Salonen et al., 2014), seeing as those who struggle with gambling are likely to have a high gambling frequency (Holtgraves, 2009;Pallesen et al., 2020). The positive association could also be related to gambling motives, as problem gamblers are more likely to gamble to cope with negative emotions or experiences, to socialize, as well as use gambling as a way to fix financial issues compared to non-problem gamblers (Stewart & Zack, 2008;Tabri et al., 2022). ...
Full-text available
Several studies have investigated attitudes toward gambling using the Attitudes Towards Gambling Scale (ATGS), however, their findings have not previously been synthesized or systematically reported. Thus, we conducted a systematic literature review on studies employing the ATGS to summarize the current evidence. Database searches were conducted in January 2022 in Cinahl, Embase, PsycInfo, Pubmed, Web of Science, GreyNet, and Google Scholar. Papers were included if they presented data based on the ATGS and were published in a European language. Twenty-six papers presenting the results from 23 unique studies met the inclusion criteria. Two reviewers independently extracted the data and assessed the risk of bias. Most of the studies were cross-sectional and used the short (8-item) version of ATGS. The synthesis indicates an overall incline towards negative attitudes. More positive attitudes were associated with being male, younger age, and higher gambling frequency. Studies were divergent in findings concerning problem gambling and gambling attitudes, which could be due to variance in problem gambling severity in the samples. The current evidence base is encumbered by limitations in study quality and designs. Future research should emphasize longitudinal designs, include non-western samples, and investigate the directionality and causality of variables associated with attitudes towards gambling. ARTICLE HISTORY
... Vyšší hodnoty odpovídají vyšší míře negativních důsledků. Nástroj je na podkladě CFA (estimator DWLS) jednodimenzionální: CFI = 0,992, TLI = 0,992, (19), skríningový nástroj pro hodnocení výskytu intenzivního užívání konopných látek v populaci se skládá z 6 otázek, stupnice 1 (nikdy) -5 (často), vyšší hodnoty indikují vyšší frekvenci užívaní konopí; proměnná byla dichotomizovaná: užívá konopí ano/ne (CAST-D); • PGSI -Problem Gambling Severity Index (11,12), 9položkový nástroj pro odhad rizikového hráčství během posledních 12 měsíců, respondent odpovídá na stupnici 1 (nikdy) -4 (vždy); vyšší hodnota indikuje větší problém; proměnná byla dichotomizovaná: hraje hazardní hry ano/ne (PGSI-D). ...
Full-text available
Záměr: Cílem studie byla deskripce asociace počtu hodin strávených na internetu a symptomatologie "závislosti na internetu" korigované sociodemografickými údaji, psychickou labilitou, užíváním alkoholu a konopí, hráčstvím a subjektivním odhadem tělesného i duševního zdraví na podkladě odpovědí českého reprezentativního souboru respondentů v době pandemie COVID-19. Soubor a procedura: Soubor tvořilo 2 602 osob (1 206 mužů, 1 396 žen), průměrný věk 44,61 roku, SD = 15,82, rozsah 15-85 let náhodně vybraných kvótním výběrem na podkladě věku, pohlaví, vzdělání a regionu. Data byla zpracována hierarchickou multivariační lineární regresní analýzou (OLS). Závisle proměnnou byly údaje z testu Excessive Internet Use. Nezávisle proměnnými byly věk, pohlaví, čistý příjem, rodinný stav, vzdělání, subjektivní odhad duševního a tělesného zdraví, údaje z nástrojů MHI-5, CAGE, CAST a PGSI. Výsledky: Výsledky jsou v souladu s předchozími výzkumy, které se shodují na tom, že rozvoj závislosti na internetu je podmíněn především časem stráveným na internetu, riziko je v inverzním vztahu k věku a mírně vyšší u mužů. Vliv dalších proměnných z oblasti látkových závislostí, hazardního hraní a duševního zdraví podporuje hypotézy o společné etiologii různých typů závislosti a jejich souvislost s duševním zdravím. Omezení studie: Studie je založena na autoreferenčních údajích, má deskriptivní, empirický charakter a neopírá se o předem for-mulovanou teorii. Objectives: the aim of the study was description of the association between the number of hours spent on the internet and symptom-atology of Internet addiction, corrected by sociodemographic data, mental instability, alcohol and cannabis use, gambling, and subjective estimation of physical and mental health using responses from a representative sample of Czech respondents during COVID-19 pandemic. Sample and setting: The group consisted of 2 602 people (1 206 men, 1 396 women), average age 44.61 years, SD = 15.8223, range 15-85 years randomly selected by quota selection based on age, gender, education and region. Data were processed by hierarchical multivariate linear regression analysis (OLS). The dependent variable was data from the Excessive Internet Use test. Independent variables were age, gender, net income, marital status, education, subjective estimation of mental and physical health, data from MHI-5, CAGE, CAST and PGSI tools. Results: the results are in line with previous research, which indicates that the development of internet addiction is mainly due to time spent on the Internet, the risk is inversely related to age and slightly higher in men. The influence of other variables measuring substance use, gambling and mental health supports hypotheses about common etiology of various types of addiction and their association with mental health. Study limitation: the study is based on self-referential data, has a heuristic, empirical character and does not rely on a pre-formulated theory.
... Concerning the factors that might predict gambling severity, we found that preferring VLT as gambling activity corresponds to moderate and severe levels of intensity. This finding is in line with previous studies indicating that VLT players are at greater risk to develop problematic gambling ( Holtgraves, 2009 ), especially when depression co-occurs ( Lévesque et al., 2018 ). The onset of the COVID-19 pandemic could be pointed to as an additional reason for the increase of psychiatric symptoms among gamblers: the fear and anxiety related to the risk of being infected were generally really high among the population ( Serafini et al., 2020 ), so it is not so far-fetched to think that individuals with an already fragile mental health could be affected in a significant way. ...
Full-text available
The coronavirus pandemic affected the life of those suffering from addic- tive behaviors often confined to prolonged periods of self-isolation. To explore the variation of symptoms related to gambling, 46 outpatients of the mental health services in the Trento Province were invited to take part in a phone interview at the start of the national lockdown. Although only 2.17% increased gambling activity during this period, half of the sample (50.00%) experienced irritability, mood fluctuation (43.48%) and anxiety (39.13%). Follow-up studies should assess modifications in their behaviors that occurred after the reopening of gambling venues.
... Indeed, Turner (2008) argues that increased volatility can increase losses, depending on the type of game chosen and the player's betting strategy, potentially contributing to problematic gambling. In support of this view, Holtgrave's (2009) analysis of adult survey data in Canada by noting that problem gambling was particularly pronounced for moderately volatile games, such as slots, although it is unclear whether it might be some other features of slots driving this correlation. ...
Full-text available
Volatility refers to the variability of bet outcomes in gambling and has been recognized as a potentially important influence on behavior. The research literature has developed competing ideas for how different behavioral responses to volatility might influence player risk. However, few empirical studies have investigated how volatility influences player behavior in a live-play, real-money environment. This paper studies 4,281 regular online slot players from two operators in the UK-one casino-focused, one bingo-focused. Longitudinal panel regressions analyze variation in players' daily session time, financial loss and declined deposits as they switched among slots games with different volatilities relative to their usual play. The findings indicate that the relationship between game volatility and player behavior is complex and often non-linear. For slots players in the casino-focused sample, lower levels of volatility than usual were typically associated with lower than average losses, declined deposits and session time. However, significant relationships were not detected in the bingo-focused operator sample. Collectively, these findings suggest that while volatility may be an important influence on behavior, this influence is not necessarily uniform or easily generalized from one population of players to another.
... Výzkumy ukazují, že výskyt problémového hráčství v populaci v dané zemi roste (alespoň dočasně) s mírou herní participace v obecné populaci (např. 6,20,21). ...
Gambling brings excitement, which is a part of entertainment, but also a basis of operant conditioning, which, in conjunction with other biological and psychological factors, leads to the loss of control over the player's behaviour. This gambling disorder (problem gambling) is characterized by a high intensity and episodic character of gambling and a high amount of staked money, with negative consequences to players and their close ones. Problem gamblers have a high rate of psychiatric comorbidity, as well as suicide. Various gambling products pose a different risk of problem gambling. The game's risk is determined by the arousal for players, the social nature of the game, or the degree of skill required for gambling. It is an effect of so-called structural and situational characteristics, such as amount and variability of bets, structure and probability of win, jackpot, game speed, near wins, audio and visual effects, etc. The game risk increases also with its accessibility or with substance use while playing. In the Czech Republic, there is a high availability of electronic gaming machines (EGM) compared to the neighbouring countries; the availability of on-line games has increased dramatically in recent years. There is also an observed increase in participation in on-line gambling, not just odds and live betting, but also technical games or casino games as reported in population surveys. Estimated up to 5.7% of the adult population is at risk of problem gambling (approx. 510 thousand persons), of which 1.2-1.4% are at a high risk (approx. 80-120 thousand persons). EGMs represent the highest risk, but also casino or on-line games including odds betting, especially live betting shows high risk of problem gambling. Most of the problem gamblers are men; especially young men are highly vulnerable. Although the new Gambling Act has introduced a number of new preventive measures since 2017 according to basic types of games, the assessment of the risk potential of particular gambling products is not a part of their licensing. This contributes to increasing availability of high-risk games, especially on-line.
... A study by Holtgraves (2009) analyzed all data from population-based surveys conducted in Canada between 2001 and 2005 comprising 21,374 participants (including 12,229 who had gambled in the past year). Using the Problem Gambling Severity Index to assess problem gambling, the study found that horserace gamblers had the lowest prevalence rates of problem gambling along with those that played bingo and bought raffle tickets (3%). ...
Full-text available
Despite the popularity of horserace gambling around the world, there is surprisingly little in-depth research on the topic. Additionally, studies suggest that motives for gambling are an important proximal factor related to problematic gambling among young people and adults. The present study investigated reasons for gambling among Norwegian horse bettors utilizing questions based on the Reasons for Gambling Questionnaire. The Norwegian gambling operator Rikstoto tracks all players’ behavior across all game types on the internet as well as land-based gambling and provided the data for the study. Consequently, the responses to the questions were correlated with actual gambling behavior. The authors were given access to an anonymized dataset of 3627 players (934 females and 2693 males) from the Norwegian horse betting operator Rikstoto who all completed a short survey. The reasons for gambling most endorsed by horserace bettors were to win big prizes and for excitement. The least endorsed reasons for gambling were to impress other people and to decrease tension. Gambling for money and gambling for recreation and coping were the most highly correlated with self-reported problem gambling. Age was significant and negatively correlated with self-reported gambling problems. The number of bets made, the amount of money consciously bet (i.e., players choosing the horse(s) compared to letting a random generator choose), as well as the monthly loss limit were significant and positively correlated. To the best of the authors’ knowledge, the present study is the first to investigate (i) motivations to gamble combining self-report data with data from a real-world setting, (ii) horserace betting with actual player data, and (iii) correlations between self-reported information about gambling problems with actual gambling behavior and self-reported motivation to play. Consequently, the findings are of high existential value to the gambling studies field.
Full-text available
The Philippine gambling industry, and particularly online gaming, has recently emerged as a major driver of the country’s economy. The rise of this controversial industry has become a cause of concern for many Filipino citizens. For the past two decades, Business Process Outsourcing (BPO) has been a major driver of the Philippine economy. However, a new sector of the economy has emerged that appears poised to take over traditional BPOs: online gambling. With the emergence of online gambling, any citizen can simply log into a website and they can play the usual casino games like poker, roulette, and slot machines or even sports betting. Online gambling is just like having a wallet in your pocket right now and just waiting for the money to come in, you just pull out your cellphone and you gamble already right away, that easy. Technically, engaging in online gambling is legal unless operated by a Philippine-licensed offshore company. Unlike other countries, the Philippines does not have strict mandates against gambling. The research aims to cover if consumer attitude and government regulations are significantly proportion with the effect of online gambling behavior. Adding to that, the researcher has added an intervening variable which is promotional ads if it has a direct impact adopting online gambling behavior. By using a quantitative analysis, the study recorded 100 respondents residing in the Philippines and studying their opinion towards this new addictive behavior relating to money. The materials and resources collected have concluded that consumer attitude and government regulations have significant impact towards this unlikely behavior, while promotional ads remain irrelevant.
Full-text available
The operation of video gambling terminals (VGTs) in Illinois has been increasing since they were legalized in 2009, then implemented in 2012. Past research has found that while gambling expenditures are positively correlated with income, they are also regressive in that, as a proportion of individual income, they are negatively correlated with income. Very little research has focused on the effects of VGTs, particularly those located in non-casino establishments. Unlike the state lottery, another form of widespread gambling outside of casinos, VGTs, also known as slot machines, are a form of gambling where substantial amounts of money can be lost quickly. We seek to provide evidence on how their proliferation and revenues relate to income and other neighborhood characteristics. Using detailed VGT revenue data from the Illinois Gaming Board from 2017 and demographic data from the 2016 5-year American Community Survey, this study investigates the relationships of the number of VGTs in operation, and the expenditures on VGTs with the local poverty rate and other socioeconomic variables at the zip code level. Our results indicate that a 1 percentage point higher poverty rate is associated with 1.47% higher VGT expenditures per capita, and 1.17% more VGTs per 10,000 population. The OLS results are not significantly biased by sample selection or spatial auto-correlation. These findings suggest that VGT gambling is negatively correlated with neighborhood income which implies that the taxation of VGT revenue is a regressive tax policy.
Full-text available
Theory and methodology for exploratory factor analysis have been well developed for continuous variables. In practice, observed or measured variables are often ordinal. However, ordinality is most often ignored and numbers such as 1, 2, 3, 4, representing ordered categories, are treated as numbers having metric properties, a procedure which is incorrect in several ways. In this article we describe four approaches to factor analysis of ordinal variables which take proper account of ordinality and compare three of them with respect to parameter estimates and fit. The comparison is made both in terms of their relative methodological advantages and in terms of an empirical data example and two generated data examples. In particular, we discuss the issue of how to test the model and to measure model fit.
Full-text available
One of the attractions of gambling is the opportunity to present to oneself and to others a desired identity. Thus, a consideration of gambling as a type of self-presentation can contribute to our understanding of how and why people gamble. In this paper a self-presentational view of gambling is outlined in terms of both general identities and specific situated identities. Relevant literature is reviewed and interpreted in terms of self-presentation and preliminary data are reported.
The hippocampus and frontal lobes both contribute to episodic memory performance. In the present study, the authors evaluated the relative contributions of hippocampus, frontal lobes, anterior temporal cortex, and posterior cortex to memory performance in neurodegenerative patients and normal older controls. Subjects (n = 42) were studied with structural MRI and a memory paradigm that measured delayed recall, semantic clustering during recall, recognition discriminability, and recognition response bias. Data were analyzed with multiple regression. Consistent with the authors' hypotheses, hippocampal volumes were the best predictor of delayed recall and recognition discriminability, whereas frontal volumes were the best predictor of semantic clustering and response bias. Smaller frontal volumes were associated with less semantic clustering during recall and a more liberal response bias. Results indicate that hippocampal and frontal contributions to episodic memory can be dissociated, with the hippocampus more important for memory accuracy, and frontal structures more important for strategic processing and decision making. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Three studies examined factors affecting the acquisition and maintenance of gambling behavior within the framework of E. J. Langer's theory (1975) of the "illusion of control." Exp I, using 150 university students, evaluated the effect of partial and continuous feedback on the perception of control in different sequences of the outcomes. Exp II, using 64 undergraduates, assessed the type of feedback and the number of trials as facilitating factors of the illusion of control. Exp III, with 21 undergraduates, used regular gamblers to evaluate the effect of sequence of outcomes. Results indicate that Ss generally did not report an illusion of control toward the experimental task. These conclusions do not replicate Langer's findings, which showed that under certain conditions Ss attribute to their personal skills the outcomes of gambling. (4 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Background: The present study was designed to assess olfactory function in severely polydipsic/hyponatremic patients with schizophrenia who also had intermittent water intoxication. Methods: The University of Pennsylvania Smell Identification Test and an olfactory acuity battery were administered to three groups of male subjects: 9 patients with schizophrenia and severe polydipsia/hyponatremia, 9 control nonpolydipsic/normonatremic patients with schizophrenia, and 9 normal controls. Results: Male patients with severe polydipsia/hyponatremia and intermittent water intoxication had marked olfactory acuity and identification deficits when compared to the patient control group of similar age and age at illness onset, and to normal controls. Conclusions: The finding of deficient acuity (detection threshold) in the polydipsic/hyponatremic group but not the nonpolydipsic, normonatremic group suggests that for this subgroup, abnormalities of olfactory sensory function may occur in a pattern previously reported for other brain disorders such as Alzheimer's disease.
The present study examined the relationships between dissociative experiences, sensation seeking scores, and gambling behavior. On the basis of the frequency of their gambling behavior and responses to the Gamblers Anonymous Twenty Questions, subjects were designated as either problem gamblers (N=30) or social gamblers (N=30). Those designated as problem gamblers responded “yes” to an average of 12.17 questions on the Gamblers Anonymous Twenty Questions as compared to the social gamblers who averaged 1.90 “yes” responses. Responses on the Twenty Questions correlated strongly with the frequency of gambling behavior and dissociative experiences. Problem gamblers reported a significantly greater number of dissociative experiences than social gamblers (p<.01). Problem gamblers scored significantly higher than social gamblers on the Total Sensation Seeking Scale, and the Boredom Susceptibility, Experience Seeking, and Disinhibition subscales. Clinical implications and recommendations for future research are discussed.
Determinants of the decision to gamble not only include the gambler's biological and psychological constitution but also the structural characteristics of the gambling activity itself. Such characteristics may be responsible for reinforcement, may satisfy gambler's needs, and facilitate excessive gambling. Showing the existence of such relationships has great practical importance. Not only could potentially dangerous forms of gambling be identified but effective and selective legislation could be formulated. This paper outlines a history of the importance of structural characteristics in fruit machine gambling and then discusses the role of a number of distinct characteristics including pay out interval, multiplier potential, better involvement, skill, win probability, pay out ratio, suspension of judgement, symbol ratio proportions, the near miss, light, colour, and sound effects and naming. These are all examined in relation to the gambler's behaviour and/or cognitions. It is shown that structural characteristics of fruit machines have the potential to induce excessive gambling regardless of individuals' biological and psychological constitution and that such insights may help in decreasing fruit machine gambling's addictiveness potential and help in formulating effective gambling policy.