Journal of Gambling Studies

, Volume 26, Issue 2, pp 189–204

Prevalence of Adolescent Problem Gambling, Related Harms and Help-Seeking Behaviours Among an Australian Population

Authors

    • Waverley Action for Youth Services (WAYS)
  • Shab Mireskandari
    • Psychosocial Research Group, Department of Medical OncologyPrince of Wales Hospital, Randwick
  • Kymbra Clayton
    • Psychology DepartmentMacquarie University
  • Alex Blaszczynski
    • Gambling Research Unit, School of Psychology (A18)The University of Sydney
Original Paper

DOI: 10.1007/s10899-009-9169-1

Cite this article as:
Splevins, K., Mireskandari, S., Clayton, K. et al. J Gambl Stud (2010) 26: 189. doi:10.1007/s10899-009-9169-1

Abstract

Epidemiological studies have consistently reported prevalence rates ranging between 0.9 and 23.5% for problem gambling among young people. With such a large range reported in the literature, it is clear that more research in this area would be of value. The current study investigated the prevalence rate of adolescent gambling and problem gambling and explored types of harm-related and help-seeking behaviours associated with gambling specific to this population in an Australian setting. A self-administered battery of questionnaires was distributed to 252 students aged 12–18 years, attending four private schools in the Eastern suburbs of Sydney, Australia. The battery included a self-administered socio-gambling demographic questionnaire, the Diagnostic and Statistical Manual Fourth Edition Multiple Response Juvenile (DSM-IV-MR-J) diagnostic instrument to assess problem gambling status, the Gambling Attitudes Scale, and questionnaires using a Likert scale to measure gambling-related harms and help-seeking behaviours. The prevalence rate among this group was found to be 6.7%. The study found further support for previous findings suggesting that a significant proportion of young people meet criteria for problem gambling, that males are at-risk and that few adolescents are able to recognise when gambling is problematic or access mental health professionals for assistance.

Keywords

AdolescentGamblingPrevalenceProblem gamblingRisk-taking

Introduction

Gambling is a popular pursuit among young people, with research in Canada (Derevensky and Gupta 2000; Gupta and Derevensky 1998a), the United States (Shaffer et al. 1994), the UK (Fisher 1993, 1999), and Australia (Delfabbro et al. 2005; Delfabbro and Thrupp 2003), reporting that between 60 and 99% of young people aged 12–20 years have gambled. The figure is comparable to that reported for adults (Petry 2005). It has been argued that the high risk of gambling behaviours in young people warrants further research in the area (Messerlian et al. 2007; Molde et al. 2009) and that the number of adolescents gambling is on the increase (Messerlian et al. 2007). Furthermore, adolescent and young adult gamblers may be the highest risk group for problem gambling (Messerlian et al. 2007). Epidemiological surveys report rates of adult pathological gambling, formally defined by DSM-IV (American Psychiatric Association 2000), at between 0.8 and 5.1% (National Research Council 1999; Petry 2005). Of significant concern, however, is the consistent finding that the comparable rates for pathological gambling among adolescents appear to be two to four times higher than their adult counterparts. These rates, measured by various instruments developed specifically for adolescents, range from 4.4 to 7.4% (Fisher 1999; Gupta and Derevensky 2000; National Research Council 1999; Shaffer et al. 1994; Shaffer and Hall 1996). Fisher’s (2000) 12-item Diagnostic and Statistical Manual Fourth Edition Multiple Response Juvenile (DSM-IV-MR-J) scale, a derivative of her original 1992 nine item DSM-J instrument, is becoming increasingly used to classify adolescent pathological gamblers.

The term ‘problem gambling’ is often used as a synonym for pathological gambling, but more accurately encompasses a wider network of individuals suffering gambling-related harms of varying nature and intensity who may or may not meet formal diagnostic criteria for pathological gambling (Productivity Commission 1999). Extending the definition to include problem gambling results in a wider range of reported prevalence rates from 0.9 to 23.5% (Sullivan 2001; Volberg and Moore 1999).

The importance of understanding the characteristics of adolescent gambling is underscored by data indicating that the majority of adult problem gamblers experienced problem gambling behaviours prior to age 20 (Gupta and Derevensky 1997, 1998a; National Research Council 1999) with some starting as young as age 8 years old (Fisher 1993). This is supported by Hanson and Rossow (2008), who found that 78.5% of participants in their study of adolescent gambling and problem gambling had engaged in gambling behaviours during the past year, with 3.1% of participants meeting criteria for problem gambling (Hanson and Rossow 2008). In addition, Magoon and Ingersoll’s (2006) study of parental modelling, attachment and supervision as moderators of adolescent gambling reported 26% of participants met problem gambling criteria. Furthermore, 91% of participants had gambled at least once in their life, with 36.2% gambling weekly and 19% gambling daily. It has been reported that males are more likely than females to gamble at problematic rates (Gupta and Derevensky 1998b; Ladouceur et al. 1994), and research suggests a direct link between the age of onset, and severity of future gambling problems (Fisher 1993), with evidence that the progression from social to problem gambling may be rapid for adolescents (Gupta and Derevensky 1998b). Given differences in maturational and developmental status between early and late adolescence, combined with legal access to gambling opportunities and motivational factors (gambling for fun among family/friends as opposed to viewing gambling as a potential source of income), it is imperative to apply proper consideration to relevant childhood factors that contribute to consolidation and persistence in gambling across early formative years into adulthood.

Research has indicated that gambling behaviours may in part, be maintained by attitudes and beliefs held about gambling. Studies have indicated that adolescents perceive gambling to involve a considerable element of skill (Derevenskym et al. 1996; Gupta and Derevensky 1998a) with a divergence shown between understanding objective odds and specific irrational beliefs related to the gambler’s personal perception of gambling (Delfabbro et al. 2005). Misconceptions about the economic returns of gambling and the adolescent tendency to be overly optimistic about winning may also be crucial in the development and maintenance of gambling problems (Delfabbro et al. 2005). Furthermore, there appears to be a functional relationship between involvement in gambling and a positive attitude toward gambling displayed by friends and family (Browne and Brown 1994; Delfabbro and Thrupp 2003). For adolescents, problem gambling appears to be associated with significantly higher rates of delinquent behaviours, truancy, substance abuse (Fisher 1999, 1992) petty crime (Gupta and Derevensky 1998b; Yeoman and Griffiths 1996), heightened risks of suicidal ideation and suicide attempts (Gupta and Derevensky 1998b), lowered self-esteem (Gupta and Derevensky 1998a) and depression (Gupta and Derevensky 1998a, b).

Adolescent problem gamblers also have significantly lower levels of psychological adjustment (Petry 2005). However, in the context of claims of elevated prevalence rates of gambling and severe gambling-related harm and symptomatology, few adolescents seek treatment, and few teachers or parents report awareness of the presence of problem gambling among young people (Ladouceur et al. 2004). The paucity of research on adolescents and help-seeking behaviours in relation to problem gambling makes it difficult to determine whether this lack of awareness is related to the fact that gambling-related harm in this population is of a low severity, simply not recognised and therefore not reported by either adolescents or parents, or hidden due to shame. Adult problem gamblers have reported shame, denial and social factors to be the most significant barriers to seeking help, rather than a lack of knowledge or dislike of treatment agencies (Evans and Delfabbro 2005). Similarly, Gupta and Derevensky (2000) argue that in adolescents, denial, fear of identification, negative perceptions of therapy, guilt and a belief that gambling can be self-managed, may account for the lack of help-seeking behaviours. Other reasons include the possibility that adolescents may grow out of their gambling problems, that they may be bailed out of trouble by parents and thus never reach a crisis leading them to seek external help (Griffiths 2002) or that they do not perceive themselves to have gambling problems (Hardoon et al. 2003). A recent study by Delfabbro et al. (2005) found adolescents turned to friends for help, then, in order of preference; counsellors, family, help-lines and school.

The primary objective of this study, therefore, was to assess the prevalence, attitudes and nature of gambling and problem gambling among a cohort of adolescents aged 12–18 years in order to obtain a detailed description of the types of gambling-related harm and help-seeking behaviours experienced by adolescent problem gamblers. The study is the first of its kind to be conducted in Sydney.

Method

Participants and Recruitment Procedures

Participants were drawn from a population of secondary school students attending four private schools (two Anglican, one Catholic and one Jewish) in the Eastern suburbs of metropolitan Sydney. Twenty private schools in the Eastern suburbs area of Sydney were randomly selected and approached. Of those, seven school principals gave initial verbal agreement to participate in the study. However, due to time limitations and school holidays, only four schools finally approved access to students, giving an overall school response rate of 20%.

In accordance with standard ethics approval protocols required for the conduct of research in independent schools, principals provided written consent to participate in the study. An information pack was then distributed to all students. This pack included a parental information sheet and consent form, a student information sheet and consent form, and a revocation of consent form. Under standard ethics protocols students could assent to the research if their parents provided consent for them to participate. Eighty-five percent of students returned signed parental and student consent forms and thus were considered eligible to participate in the study.

Class teachers allocated a period of time, 1 week after distribution of information packs, for eligible students to complete the questionnaire. All questionnaires were administered in a classroom or school hall setting under the supervision of the researchers. Prior to commencing the questionnaire, duty of care obligations detailed in the information packs were again verbally stated to students. Students were informed that the researchers would notify the school principal and/or relevant authorities of any student to be found at risk of harm; gambling-related or otherwise, as a consequence of responses given to the questionnaire. The questionnaire took approximately 15 minutes to complete.

Measures

Participants completed a self-administered questionnaire that elicited personal and gambling demographics, attitudes to gambling, a scale assessing probable problem gambling, and harm and help-seeking behaviours related to problem gambling.

Demographics

Information regarding demographic details: age, gender, school grade, country of birth, number of years residency in Australia, and living arrangements were obtained.

Gambling Characteristics

Participants were instructed that for purposes of the study, gambling was defined as betting money, property or something else of value on an activity with an uncertain outcome. It does not include friendly bets or challenges where nothing is won or lost. To assess patterns of gambling behaviour, data on frequency of gambling on specific commercial and non-commercial forms of gambling, net expenditure (defined as the difference between money taken to, and remaining, after each gambling session), changes in expenditure over time, and an estimate of income were elicited. Information on how participants obtained lottery tickets or scratch cards, reactions to their first experience of gambling and the main reasons for gambling were also obtained. Social factors influencing propensity to gamble included items related to the first person they gambled with, whether parents replaced money lost during gambling, whether playing with parents shaped interest in gambling and parental/guardian problem gambling status.

DSM-IV-MR-J (Fisher 2000)

The nine-item modified version of DSM-IV-MR-J adapted for adolescents was used to assess for the presence of probable problem gambling. The previous 12 month time-frame anchor-point was used. The DSM-IV-MR-J uses multiple responses along a four point Likert scale with options ranging from ‘Never’ to ‘Often’ (Fisher 2000). The scale has acceptable construct validity, internal consistency, reliability (alpha = 0.75), and a demonstrated ability to discriminate between problem and non-problem gamblers (Fisher 2000). The readability of the DSM-IV-MR-J has been assessed to be at high fourth grade reading level, allowing its use with adolescents with a wide range of reading abilities (Fisher 2000). A score of 4 or more on the scale is considered indicative of problem gambling (Fisher 2000).

Gambling Attitudes

The current study used the Gambling Attitudes Scale (Delfabbro and Thrupp 2003) to measure attitudes and economic perception of gambling held by participants. Participants were required to rate their agreement with nine statements such as: ‘Gambling is a risky activity’, ‘You can lose all your money gambling’, and ‘Gambling is a good way to get rich quickly’, on a Likert scale ranging from Strongly agree (1) to Strongly disagree (5). This scale has been found to have good internal consistency (alpha = 0.83; Delfabbro and Thrupp 2003). Participants were also given a yes/no response option on whether they believed gambling was an easy way to make money.

In addition, attitudes and beliefs surrounding the relative contributions of chance and skill were assessed. A 10-point Likert scale adapted from Delfabbro et al. (2005), assessed how much skill versus chance was involved in a selection of gambling activities, ranging from Chance (0), Some skill (5), All skill (10). Participants were also asked to rate their perception of their own gambling skill in comparison to that of a person of equal age and gender on a 10-point Likert scale ranging from Much worse than average (0) to Much better than average (10).

Participants were also asked for the main reasons they gambled, with options including: mixing with friends, for excitement, for a challenge/risk, as a hobby or to distract from everyday problems.

Gambling-Related Harms

Items from the DSM-IV-MR-J (Fisher 2000) were used to gain information on gambling-related harms including arguments with friends and family, lying, stealing money and missing school.

Other study-specific items were adapted from the South Oaks Gambling Screen (Lesieur and Blume 1987) to investigate whether participants had been criticised for gambling behaviours, hidden gambling-related goods, sold personal property to gamble or committed illegal acts. These items were scored using a 4-point Likert-scale: Never (0) to Often (3). Participants were also asked to indicate (with options ranging from At least once a day (0) to never (5)), whether they had been involved in other risk-taking behaviours: smoking, self-harm, drug/alcohol use or involvement with police.

Help-Seeking Behaviours

Participants were asked if they were aware of friends manifesting gambling-related problems and if so, the likely source of assistance they would pursue. Similarly, if they recognized the presence of a gambling problem in themselves, to whom they would turn for assistance.

Data Analyses

Data was analysed using SPSS 13.0 (Statistical Program for the Social Sciences, Version 13.0). Descriptive statistics (frequencies, means and medians) were used to summarize the data. In addition, bivariate analyses, including Spearman’s Rank correlations and Pearson’s Chi square tests were undertaken. Missing data or items that did not apply to respondents account for the variability in reported degrees of freedom.

Results

Demographics

Table 1 contains summary characteristics of participants by gender. The sample included a total of 252 participants with a mean age of 14.6 years (SD = 1.5; range = 12–18). This sample included 184 males (M = 14.6 years; SD = 1.5) and 68 females (M = 14.5 years; SD = 1.2).
Table 1

Summary characteristics of study participants by gender

Variable

Male (n = 184)

Female (n = 68)

Total (N = 252)

School year

 Years 7–10

129

62

191

 Years 11–12

55

6

61

Country of birth

 Australia

154

59

213

 Other

30

8

38

Living arrangements

 Two biological parents

145

54

199

 Other

39

14

53

The majority (n = 213; 84.5%) were born in Australia. The mean age of residency in Australia for those born overseas was 6.7 years. The majority of participants (n = 199; 79%) lived with both parents, 15 (6%) with a biological and step-parent, 26 (10%) a sole mother, 2 (0.8%) a sole father, and 1 (0.4%) with foster parents. Nine (4%) lived in other unspecified domestic situations.

One hundred and ninety-one (76%) participants (n = 129, 67% male; n = 62, 33% female) were in Years 7–10, and 61, (24%) participants, (n = 55, 90% males; n = 6, 10% females) in Years 11–12.

Problem Gambling Prevalence

A total of seventeen participants (6.7%) met the criteria for probable problem gambling, (a score of four or more on the DSM-IV-MR-J). All of these participants were male, with no females meeting the criteria for probable problem gambling. Table 2 shows the proportion of participants endorsing items on this scale. The mean number of items endorsed for the total sample was 0.86 (SD = 1.7; range = 0–11). The mean age of problem gamblers was 15.06 (SD = 1.02; range 13–17). Of these, nine (52.9%) were Year 9, two (11.8%) Year 10, four (23.5%) Year 11, and two (11.8%) were Year 12 students.
Table 2

Proportion of participants endorsing DSM-IV-MR-J items (n = 252)

Number of DSM-IV-MR-J items endorsed

n (%)

0

167 (66.3)

1

28 (11.1)

2

20 (7.9)

3

13 (5.2)

4

17 (6.7)

Missing

7 (2.8)

Total

252 (100)

Gambling Characteristics

Eighty-one percent (n = 205) of the total sample reported gambling within the past 12 months with slightly less than half (n = 109; 43%) having commenced prior to age 11. The median age for onset of gambling was 13 years.

Gender was significantly related to gambling with more males reporting having gambled (F(1,249) = 23.95; P = .0005), and a significantly greater proportion of males reported starting to gamble between the ages of 5–11 than females (t = 15.045, P = .005). Males appeared to have found their first experience of gambling significantly more exciting than females (F(1,201) = 15.598; P < .0005). Males were also significantly more likely to be preoccupied with thoughts about gambling than females (F(1,249) = 24.29; P = .0005) and expressed a need to increase expenditure to maintain levels of excitement (F(1,248) = 5.49; P = .020).

When looking at the relationship between gambling and the effect on friendships or family relationships, males were more likely to report having arguments about gambling with family and friends than females (r = −.179, P = .004).

Forms of Gambling

In order of most popular, coin tossing, sports betting, and lotteries were the three most frequent forms of gambling associated with weekly and/or daily play. The majority of participants had never gambled via the Internet, with only one person reporting high frequency gambling of this form.

Finances

As expected within a school population, pocket money and part-time jobs formed the main source of gambling funds, with no differences found according to gambling status (χ² = 1.114, df = 1, P = .373).

Social Influences

Problem gamblers were significantly more interested in gambling than non-problem participants if their first gambling experience was with their parents (F(1, 204) = 7.48; P < .005).

Gambling Attitudes

There was a significant difference between problem and non-problem gamblers in their beliefs about the amount of skill and chance involved in sports betting [F(1,227) = 5.767; P = .017], roulette [F(1,223) = 5.867; P = .016], and internet gambling [F(1,227 = 4.416; P = .037], with problem gamblers believing more skill was required for all of these forms of gambling.

Significantly more problem gamblers rated themselves as skilful at their favourite gambling activity compared to others of the same age and gender (F(1,234) = 12.80, P < .001), while non-problem gamblers perceived their skills to be the same as their peers.

Five percent of the total sample (n = 13) reported a belief that gambling was a very easy way of making money. Participants who had never gambled held significantly stronger beliefs that gambling was risky (F(1,249) = 6.99, P = .009), an easy way to lose money (F(1,247) = 4.33, P = 0.04), a waste of money (F(1,248) = 8.64, P = .004) or throwing away money (F(1,249) = 16.06, P < .000).

Negative attitudes toward gambling were also related to gender, with females significantly more likely than males to regard gambling as risky (F(1,249) = 6.987, P = .009), a waste of money (F(1,248) = 8.637, P = .004), throwing away of money (F(1,249) = 16.06, P < .000), or a way of losing money (F(1,247) = 4.333, P = .04). Males were significantly more likely to agree that individuals could make a living gambling (F(1,248) = 6.48, P = .01).

Perceptions of Reasons for Gambling

Table 3 shows the differences between problem and non-problem gamblers in their reported reasons for gambling. Compared to non-problem gamblers, problem gamblers endorsed gambling for its excitement, challenge, risk and their drive to win money as their salient motivation to a significantly greater level.
Table 3

Differences between problem and non-problem gambling respondents who gambled in the last 12 months in reasons given for gambling

Item

Non-PG (n = 188)

PG (n = 17)

F

P

n (%)

n (%)

To win money

68 (36)

13 (76)

10.97

.001

To mix with friends

44 (23)

3 (18)

0.317

.574

Out of curiosity

49 (26)

4 (24)

0.041

.839

Other reason

15 (8)

0 (0)

1.451

.230

My friends gamble

8 (4)

2 (12)

1.528

.218

I like the challenge/risk

46 (24)

12 (71)

17.28

.000

Hobby

5 (3)

1 (6)

0.582

.446

For entertainment/fun

113 (60)

11 (65)

0.113

.737

Excitement

58 (31)

11 (65)

8.13

.005

Distraction from everyday problems

5 (3)

0 (0)

0.455

.501

Note: participants were able to endorse multiple items

Gambling-Related Harms

Problem gamblers reported significantly more difficulties such as arguments with family or friends or problems at school (F(1,183) = 32.77; P = .0005). Similarly, correlation analyses showed these arguments were significantly (r = .223, P = .002) related to participants hiding their gambling behaviours. Problem gamblers were also more likely to have missed school as a result of their gambling compared to non-problem gamblers (F(1,250) = 18.46; P = .0005).

Significantly more problem gamblers than non-problem gamblers had sold personal property in order to gamble (F(1,182) = 18.574; P = .0005) and committed illegal acts to obtain money (F(1,182) = 4.402; P = 0.037). Correlation analyses revealed that the need to increase the amount of money spent gambling was related to the excitement of gambling and the need to increase this excitement level (r = .210, P = .004).

Table 4 depicts problem and non-problem gamblers’ experiences in relation to other risky behaviours, such as smoking and drug-taking. Problem gamblers were significantly more likely to have tried the behaviours at least once compared to those classified as non-problem gamblers, with only the use of illegal drugs such as ecstasy and legal medicinal drugs resulting in non-significant differences.
Table 4

Comparison between problem and non-problem gamblers for harmful behaviours

Behaviour

Comparison groups

M

df

F

P

Smoked cigarettes

PG

0.23

1

14.552

.000

Non PG

0.94

230

  

Cut or burned yourself

PG

0.31

1

6.846

.009

Non PG

0.87

228

  

Drank alcohol

PG

1.00

1

25.098

.000

Non PG

2.38

227

  

Used marijuana

PG

0.13

1

13.692

.000

Non PG

0.56

229

  

Used illegal drugs such as ecstasy/speed

PG

0.03

1

0.315

.575

Non PG

0.06

229

  

Used bulbs, paint thinners, glue or petrol

PG

0.07

1

10.940

.001

Non PG

0.50

230

  

Used legal drugs such as painkillers/Panadol for non-medicinal purposes

PG

0.18

1

0.714

.399

Non PG

0.06

230

  

Recognition of Problem Gambling

Despite a proportion of participants meeting the criteria for diagnosis of a probable problem gambling condition, there were no significant differences between problem and non-problem gamblers in self-reported presence of a gambling problem (F(1,246) = 3.66; P = .057). Only 1.2% (n = 3) of the whole sample reported having a gambling problem. The majority of problem and non-problem gambling adolescents reported experiencing no problems. Of the 17 participants classified as probable problem gamblers according to the DSM-IV-MR-J, only n = 1 (6.3%) recognised they had a problem with gambling. Of the non-problem gamblers, n = 2 (0.9%) reported that they had a gambling problem. There were no significant differences between problem and non-problem gamblers in the number of peers they identified as having problems with gambling (F(1,245) = 2.901; P = .09).

Help-Seeking Behaviours

The problem gambling and non-problem gambling participants differed significantly in respect to where they would go to seek assistance if their friend had a gambling problem (F(1,245) = 4.61; P = .030). In respect to whom they would approach for help if their friend had a gambling problem, both groups reported parents, counsellors and friends as their top three choices. However, non-problem gamblers were significantly more likely to approach their parents than problem gamblers (F(1,245) = 6.723; P = .010). In order of preference, problem gamblers reported they would turn to friends, counsellors, parents, help lines, siblings and teachers for help for a friend. Non-problem gamblers reported turning firstly to parents, then counsellors, friends, help lines, teachers, relatives and siblings for help for a friend.

When participants were asked about their own gambling behaviours, significantly more non-problem gamblers than problem gamblers reported that they would seek assistance if they had a gambling problem (F(1,245) = 10.958; P = .001). In examining who participants would approach for help for themselves, the only significant difference appeared to be that non-problem gamblers were significantly more likely to seek the help of a counsellor than were problem gamblers (F(1,243) = 4.41; P = .034). In order of preference, problem gamblers reported turning to friends and parents first, then siblings, help lines, counsellors and relatives if they felt they had a problem. Non-problem gamblers reported turning firstly to parents, and then friends, counsellors, help lines, siblings, teachers and relatives if they felt they had a gambling problem.

Discussion

The results of the present study are consistent with previous research indicating that gambling is a common activity among adolescents (Delfabbro et al. 2005; Shaffer et al. 1994), that gambling begins at a young age (Gupta and Derevensky 1997, 1998a), and that male problem gamblers are over-represented in this population (Gupta and Derevensky 1998b; Ladouceur et al. 1994). The problem gambling rate of 6.7% found in the current study is almost twice the 3% generally found in other Australian states (Delfabbro and Thrupp 2003; Delfabbro et al. 2005; Moore and Ohtsuka 1997) and more comparable with adolescent pathological gambling rates ranging from 4.4 to 7.4% reported in North American studies (Shaffer and Hall 1996). This discrepancy may be accounted for by the different populations measured and range of assessment tools used in previous studies.

It is also conceivable that instructions given to students prior to questionnaire completion informing them that school authorities would be notified if anyone was found to be at risk of gambling-related or other harms may have influenced the accuracy of self-reports. This would result in an underestimate of the true rate of problem gambling. However, this is unlikely given that, as noted above, the obtained rates were double those reported for Australian adolescents, and were comparable to North American rates.

Evidence is emerging for the need to conceptually clarify and refine the construct of problem gambling in the face of data that suggests substantive discordance between psychometric and clinically-derived instruments (Ladouceur et al. 2005). Setting these issues aside, it is apparent that rates of problem gambling found among adolescents are at least twice that found in adult populations (Delfabbro et al. 2005).

In terms of the types of gambling preferred by this sample, the main forms associated with weekly and/or daily play were coin tossing, sports betting and lotteries. At least one of these preferred forms—lottery playing—legally requires the involvement of a person over the age of 18, perhaps indicating that participation involves or is mediated by an older individual, either parent or peers. Given that participants reported parents, siblings, and relatives as primary providers of lottery tickets suggests a cultural acceptability of this form of gambling as innocuous and not a risk-factor or gateway for the development of subsequent gambling related problems (Moore and Ohtsuka 1997), however, it is not possible to draw a direct link between lotteries and problem gambling as there is little representation in the literature regarding the consequences and effects of lottery gambling (Griffiths and Wood 1999). Nonetheless, this holds implications for the potential need to shift misguided community attitudes that certain forms of gambling represent harmless behaviours. Additionally, it suggests a broader awareness of social learning theory, which argues that attitudes and beliefs are transmitted from adult models to child (Bandura 1986; Browne and Brown 1994).

Consistent with the notion that parents play an important role in exposing young people to gambling (Wood and Griffiths 1998), the current study found problem gamblers were significantly more likely to report that playing with their parents had initiated their primary interest in gambling and that the receipt of lotteries as gifts was more influential when given by parents and siblings than by more distant relatives or through self purchase. This may be particularly relevant to lottery playing which allows children to see the process and outcome of parents’ gambling as a socially acceptable activity (Fisher 1999; Wood and Griffiths 1998). Whether these tickets are bought on the initiative of the adult or in response to a young person’s request requires more investigation.

In line with the findings of Delfabbro et al. (2005), attitudes towards gambling were a differentiating factor, with non-problem gamblers significantly more likely to express negative beliefs toward gambling as an economic activity while problem gamblers endorsed a more positive and optimistic attitude toward it as a means of supplementing income. Similarly, young people who reported having never gambled were significantly more likely to hold negative views towards gambling as a strategy to gain money than those who had gambled at least once.

There was also a gender difference in gambling attitudes, with females being significantly more likely to see gambling as a risky activity which was not economically sound. Given both problem gamblers and non-problem gamblers perceive gambling as risky, the differentiating factor appears to relate to attitudes toward money as a valued commodity. Further investigation is needed into whether such attitudes are important in determining whether young people explore gambling as an acceptable activity, or whether the individual experience of gambling aids the development of such beliefs.

While problem gamblers had unrealistic beliefs about the economic returns of gambling, neither problem gamblers nor non-problem gamblers believed gambling to be an easy way to make money. The current study found that significantly more problem gamblers rated themselves as more skilful at their favourite gambling activity than someone else of the same age and gender.

It may be that problem gamblers believe winning is not easy, but if you are skilful it is possible to get rich quickly. Research on gambling has shown that young people hold erroneous views about the amount of skill and luck involved in gambling (Delfabbro et al. 2005). If a perception of skill is involved in gambling, then winning, as well as being positively reinforcing, would lead to feelings of being skilful and boost self-esteem.

This would be consistent with Jacob’s General Theory of Addictions (1986), which postulates gambling as a means of both escaping from and coping with emotional pain and distress. Jacobs (1986) argues that the capacity of gambling to offer an escape from emotionally painful realities and experience ego-enhancing fantasies of importance and success is the essential reinforcing quality which maintains the addictive patterns of behaviour. However, in the current study there were no significant differences between sub samples in reporting gambling being a way to escape everyday problems, with no problem gamblers reporting such motivations.

The current study found significantly more problem gamblers reported gambling for excitement, challenge, risk, and drive to win money, as salient motivating factors. Significantly more problem gamblers also found their first gambling experience exciting suggesting gambling may have a positively reinforcing quality. Together these findings suggest that experiential factors are perhaps more relevant in establishing interest in gambling than context or co-participants.

As previous research has suggested, it may be that along with other high risk behaviours, gambling provides excitement (Gupta and Derevensky 1998a) and can be a healthy part of adolescent development (Moore and Ohtsuka 1997), a rite of passage into adulthood (Derevensky et al. 1996; Hardoon and Derevensky 2001; Moore and Ohtsuka 1997) and a sense of group belonging (Moore and Ohtsuka 1999). However, not all adolescents maintain safe levels of gambling (Fisher 1999) and the group behaviours may also support unproductive norms, for example the belief that gambling is profitable.

Research has shown an association between gambling and other at-risk behaviours, such as alcohol consumption (Delfabbro et al. 2005; Jackson 1999). In the current study, problem gamblers were significantly more likely to smoke cigarettes, cut themselves, drink alcohol, and use marijuana or inhalants than non-problem gamblers. While it can be argued that such risk-taking is relevant to their developmental age and desire for experimentation and independence, Fisher (1993, 1999) argues that gambling often coincides with negative forms of risk taking.

While this study suggests young people experience gambling-related harms in terms of missed school, social problems and crime, previous research (Gupta and Derevensky 2000) shows that few adolescents seek assistance for gambling-related problems. The present study suggests this might be due to an inability to recognize gambling as problematic. Despite a proportion meeting criteria, the majority of problem gamblers failed to recognize or at least report the presence of gambling problems. One explanation for this could be that problem gamblers are more likely to believe that gambling is not a harmful activity (Jackson 1999).

The adolescent tendency to minimize and underestimate severity of gambling problems may also partly account for the few youths in treatment (Gupta and Derevensky 2000; Ladouceur et al. 2004). However, Fisher (1992) found that young people who played fruit machines in the UK, and problem gamblers in particular were aware of the possible negative consequences of play, and worried they played too much. Perhaps playing on commercial forms of gambling, such as fruit machines, is viewed as having the potential to be more problematic than, for example, coin tossing. This may particularly be the case if young people are making judgments about severity of their gambling habits based on adult indicators of gambling problems.

Indicators of adult distress such as debt, marital relationship breakdowns and job loss are not applicable to young people who may be unaware their gambling is problematic until they see severe consequences (Gupta and Derevensky 2000). Similarly, evidence suggests few teachers or parents report adolescent problem gambling (Ladouceur et al. 2004). They too may be looking for adult signs of problem gambling and miss youth indicators such as delinquent behaviours, loss of social relationships and petty theft. The qualitative difference in adolescent and adult gambling-related harms requires further investigation.

Problem gamblers reported being significantly more likely than non-problem gamblers to turn to their parents for help if a friend had a gambling problem but significantly less likely than non-problem gamblers to turn to a counsellor if they felt they had a problem themselves. The difference in choice of aide in relation to self or other’s gambling problem requires further investigation but may be due to factors such as shame which prevent a young person seeking help from someone who may judge them rather than an independent professional. However, both groups reported that a counsellor would be a second option for a friend in need, suggesting that it is not a lack of knowledge about existing services accounting for the few young people accessing them.

The reasons for this require further investigation but results from this study indicate that while problem gamblers experience harms, the need for help is rarely recognised by themselves or others and so is not sufficiently attended to. Furthermore, problem gamblers are at increased risk, as they are even less likely than their non-problem gambling counterparts to seek help, even if they were aware they had a problem.

Limitations

While the current study adds to the literature on adolescent gambling, the limitations of this study need to be acknowledged. This was a cross sectional study involving a population of private school students, thus generalisability to other populations is limited and no causative links can be made. It is relevant to note that there were multiple statistical comparisons raising the possibility of finding chance differences. Future prospective studies are needed to confirm the reliability of findings. In addition, replication of this research with other populations, such as students attending non-religious or state schools would be beneficial. It is also important to note that participants in this study were aware that any identified problem gamblers would be reported to the principal of the school. This may have influenced participants’ responses regarding their gambling behaviours and other behaviours such as substance misuse. Further, there is a possibility of a clustering effect, as a consequence of recruiting from a population of students attending a small number of schools. Whilst this study has generated some useful data, these limitations mean that it is important that future research in this area seeks to address these issues before any hypotheses or theories are expounded on the basis of these findings.

In terms of the distribution of the sample, there was a bias towards younger participants, with an average of 47.75 participants per year in Years 7–10 and an average of 27.5 participants in Years 11–12. Future studies could address this inequity by including more participants in the older age range. Future prospective studies are needed to decipher causation between problem gambling and harms/help-seeking behaviours and the progression of gambling behaviour in this population. While the problem gamblers in the current study were all found to be male, the relatively small numbers of female participants make it impossible to tell whether problem gambling is a male-only problem or a consequence of sample size. Further investigation and studies with larger samples sizes are required to clarify this. Use of focus groups might also be beneficial in gaining more in-depth information about harms related to gambling and help-seeking behaviours.

Conclusion

The current prevalence study is the first investigation of youth gambling in Sydney. Given that gambling begins at a young age, and the cognitive stage of development is such that it supports erroneous beliefs about gambling, it would appear that young people between the ages of 5 and 13 represent a target group for the prevention of excessive gambling. Teachers and parents would also benefit from having increased awareness of adolescent gambling and its related harms, and the role the family may play in transmitting positive attitudes about gambling to young people by modelling behaviours or flouting legal age limits to bet on their behalf.

In Australia, schools address issues such as sex and sexuality, drugs and alcohol and social issues as part of the curriculum. Compulsory psycho-social education on gambling could be embedded in personal development education. Adolescents could be educated on cognitive errors, risk factors and available services if they have gambling related concerns. This could be a step towards reducing the suffering for young people who are already involved in problem gambling, as well as help reduce the number of adolescents initiating gambling in the future.

Acknowledgements

The authors would like to thank Star City Casino for funding this research and Waverley Action for Youth Services (WAYS) for resources and access to schools in the Eastern suburbs of Sydney. Thanks to Dr Terri Said and Claire McGuire for their administrative support and most importantly to the schools and students who agreed to take part in this study.

Copyright information

© Springer Science+Business Media, LLC 2010