Testing a cognitive behavioural theory and treatment of problem gambling

Raylu, N. N. (2005). Testing a cognitive behavioural theory and treatment of problem gambling PhD Thesis, School of Psychology, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
THE18390.pdf Full text application/pdf 16.02MB 0
Author Raylu, N. N.
Thesis Title Testing a cognitive behavioural theory and treatment of problem gambling
School, Centre or Institute School of Psychology
Institution The University of Queensland
Publication date 2005-01
Thesis type PhD Thesis
Supervisor Oei, Tian
Duck, Julie Maree
Total pages 275
Collection year 2005
Language eng
Subjects L
380000 Behavioural and Cognitive Sciences
780108 Behavioural and cognitive sciences
Formatted abstract

The main aim of this thesis was to make advances in the assessment, theoretical framework and treatment of problem gambling (PG), especially in relation to the cognitive behavioural (CB) perspective. A review of the PG literature (chapter one) suggested that the lack of gambling related instruments had prevented researchers from adequately testing the extent to which these gambling related variables [such as gambling urges (GU) and gambling related cognitions (GRC)] existed among gamblers/PGs, the validity of these CB models presented in the gambling literature as well as the efficacy of cognitive behavioural therapy (CBT) in reducing these cognitions and urges. The thesis thus, aimed to devise and validate two measuring instruments to assess GU and GRC and use these and other measures to test a range of gambling variables among a conummity sample, test the validity of Sharpe's (2002) CB model of PG and consequently, present a pilot study that assessed the efficacy of a group and individual CBT program. 


The thesis also presented two studies that discussed data from BreakEven, a major treatment agency for problem gamblers (PGs) in Queensland, Australia, outlining the characteristics of gamblers presenting for treatment, treatment outcomes and possible factors that may predict treatment outcomes or methods of treatment cessation. Data from Break Even was presented to show the characteristics of gamblers presenting for treatment in an Australian treatment agency for PG compared to those in the community. They were also completed to compensate for the lack of studies conducted with Australian treatment agencies for PGs. These BreakEven studies also aimed to explore whether gambling may be in response to traimiatic events (e.g., abuse) in order to provide further evidence for the vulnerability aspect of Sharpe's (2002) model. 


Results (in chapter two) generally showed that although most characteristics of gamblers presenting for treatment were similar to the ones found in previous/American studies, some gender differences were found. Results showed that gamblers presenting for treatment were equally likely to be males or females and be single or married/in a relationship. However, they tended to report secondary as their highest level of education, have family members with gambling problems, have significant negative correlates of gambling (e.g., negative consequences related to financial, family, relationship, vocational, psychological, physical/health, accommodation and legal issues), and comorbidity (e.g., other excessive behaviours such as substance abuse, and psychological problems). Furthermore, most of the participants were Caucasians, in their early fourties and machine gamblers. The study also provided information on additional characteristics/comorbid problems (e.g., abuse issues) of PGs presenting for treatment, providing further evidence for the vulnerability aspect of the CB conceptualisation of PG. Significant number of individuals reported dealing with other issues such as child abuse (6.6%) and spouse assault (6.4%) or reported that they had a recent experience of family violence or intimidation in their domestic/living situation (17.1%). There were significant gender differences found in the demographic characteristics, gambling behaviours and the correlates/consequences of gambling. 

BreakEven data discussed in chapter three showed that generally outcomes (e.g., therapeutic success and drop out rates) were similar to the ones reported in other addiction populations. Results also showed that certain demographic factors (e.g., gross personal income), PG behaviours (e.g., preoccupation and tolerance to gambling, gambling to escape and chasing losses) and gambling correlates (other excessive behaviours, intrapersonal problems and abuse issues) could significantly predict resolution of certain gambling correlates or methods of treatment cessation.

Chapters four and five aimed to develop and validate two instruments to assess GU and GRC, respectively. A 6-item scale was developed to assess GU (The Gambling Urge Scale, GUS - chapter four) using data provided by 968 community participants Furthermore, a 23 item instrument was developed to assess 5 domains of GRC (The Gambling Related Cognitions Scale, GRCS - chapter five) using the same sample. Both of these scales were assessed using confirmatory factor analyses and had good reliabilities and validities. 


Using the GUS and the GRCS, a range of variables including gambling behaviours gambling specific variables (e.g., motivations towards gambling, GRC and GU) and gambling correlates (e.g., psychological states such as depression, anxiety and stress levels forms of coping and personality traits) were assessed using a community (N=969) sample (chapter six). Results showed significant differences between non-problem gamblers, at risk gamblers and probable problem gamblers for almost all of these variables. Significant interactions between gender and PG groups in relation to these variables were found for gambling urges, certain forms of coping styles and gambling motivations. 


Using the data presented in chapter six and the two questionnaires developed in chapters four and five, the validity of Sharpe's CB model was assessed via statistical means (chapter seven). Results showed that statistically the model was valid and thus, provided support for Sharpe's CB model. More importantly, it showed that the model fitted the two genders differently. This could either be attributed to the general bias in the gambling literature in using predominantiy male samples and/or the gender differences found both in this study as well as other studies in the PG literature. 

Chapter eight showed that CBT (compared to a 6-week waitiist group) significantly improves PG behaviours as well as gambling correlates such as quality/satisfaction of life and mood, and maintain these changes for at least a 6-month period. However, no significant differences were found between group and individual CBT. 


The strengths, weaknesses, implications and recommendations for each of the studies contained in this thesis were discussed in the relevant chapters and the general discussion presented in chapter nine. The studies contained in this thesis were unique investigations in Australian and international research in this area. They will not only add considerably to this growing area in the gambling and PG literature, but also assist in designing and conducting preventative and treatment programs as well as future studies.

Keyword Compulsive gambling
Compulsive behavior

Document type: Thesis
Collection: UQ Theses (RHD) - UQ staff and students only
Citation counts: Google Scholar Search Google Scholar
Created: Fri, 24 Aug 2007, 18:52:10 EST