RESEARCH SYNOPSIS

RESEARCH SYNOPSIS
Arthur, D., Tong, W. L., Chen, C. P., Hing, A., Y., Sagara-Rosemeyer, M., Kua, E. H., Ignacio, J.
(2008). The validity and reliability of four measures of gambling behaviour in a sample of
Singapore university students. Journal of Gambling Studies, 24, 451-462.
RESEARCH QUESTIONS
Are the Diagnostic Statistical Manual-IV (DSM-IV)
criteria for Pathological Gambling (PG), South Oaks
Gambling Screen (SOGS), Gamblers Anonymous 20
(GA-20), and Canadian Problem Gambling Index
(CPGI) valid (i.e., accurate) and reliable (i.e.,
consistent) measures of problem gambling in
university students in Singapore?
PURPOSE
Among the studies on problem gambling, there are
four diagnostic and screening instruments (i.e., DSMIV, GA-20, SOGS and CPGI) with convincing
psychometric properties. The primary purpose of the
present study was to determine the reliability and
validity of these four instruments for detecting
problem gambling in university students in Singapore.
HYPOTHESIS
None stated.
PARTICIPANTS
Participants were 148 university students (47% male)
21 years of age and older in Singapore.
PROCEDURE
In order to select respondents from a representative
cross-section of faculties across campus, research
assistants visited common areas frequently visited by
potential participants such as Student Services
Centre, Health and Wellness Centre, Entrance to
Computer Centre, Entrance to Co-op Bookshops. In
addition to the four measures of interest, the
questionnaire also collected demographic and
gambling activity information.
MAIN OUTCOME MEASURES
The DSM-IV criteria assess pathological gambling as
an impulse control disorder. It focuses on the
psychological motivations that underlie problem
gambling and consists of ten diagnostic criteria which
are presented as a checklist. People meeting 5 out of
10 criteria are labeled as practicing pathological
problem gamblers or potential pathological gamblers.
The GA-20 assessing the severity of an actual
gambler’s problems related to gambling behaviour.
The 20 dichotomous items on this questionnaire are
primarily descriptions of typical pathological gamblers
and mainly focus on financial aspects of gambling
behaviour. Scoring is such that participants who score
1–6 points are considered problem gamblers, and
those who score 7 or more points are classified as
compulsive gamblers. The SOGS assess gambling
problems. Respondents circle a ‘‘yes’’ or no’’ in
response to initial items regarding gambling
behaviour and in later items, choose from a list of
responses pertaining to frequency of gambling and
amount of money spent on gambling. Respondents
who score 3 or above are classified as problem
gamblers and those who score 5 or more are
classified as probable pathological gamblers. The
CPGI is a multi-component measure with specific
information on type of gambling. It is made up of 42
questions requiring various responses such as
yes/no, frequency and degree of agreement, actual
amount and actual time. Four domains comprise the
questions in the CPGI: gambling involvement,
problem gambling behaviour, adverse consequences
and problem gambling correlates. The studies using
the CPGI classify persons with a score from 3 to 7 as
‘‘moderate problem gambling’’ and a score of 8 or
higher as ‘‘severe problem gambling.”
KEY RESULTS
All measures demonstrated moderate to high
reliability. The CPGI demonstrated the highest
reliability and there was strong convergent validity
(i.e., measured same construct) demonstrated
especially between SOGS and CPGI. The findings
suggested that CPGI was useful for identifying, not
only problem gambling, but other issues such as
depression, suicide, drinking and financial problems
related to gambling. Of the four measures, only the
CPGI demonstrated construct validity (i.e., measures
what intended to). The majority of students who
reported themselves as gamblers differed in numbers
depending on the instrument used. Gamblers (32%)
were easily identified by SOGS and the numbers of
gamblers varied from 14% to 32% depending on the
questionnaire used.
LIMITATIONS
Participants routinely asked ‘what do you mean
gambling behaviour?’, and ‘I have played Mahjong
with my family, but it was only during Chinese New
Year’ will this be considered as gambling behaviour?’
The high rejection rate of the CPGI was significant
while for the other instruments it was negligible. While
the CPGI is longer than the others there clearly may
have been an order effect as the CPGI was the last of
the four instruments to be administered. Participants
were reluctant to expose their gambling history.
CONCLUSIONS
The findings suggest
screening instrument
strong psychometric
dimensional structure
sample.
that CPGI is the most valid
for problem gambling having
features and a solid unifor use in a local Singapore
KEYWORDS: gambling, validity, reliability, DSM-IV, GA-20,
SOGS, CPGI, psychometric property
URL: http://dx.doi.org/10.1007/s10899-008-9103-y