- Email: [email protected]
Validation of the reasons for gambling questionnaire (RGQ) in a British population survey
Addictive Behaviors 45 (2015) 276–280
Contents lists available at ScienceDirect
Addictive Behaviors
Short Communication
Validation of the reasons for gambling questionnaire (RGQ) in a British population survey Natale Canale a, Massimo Santinello a, Mark D. Griffiths b,⁎ a b
Department of Developmental and Social Psychology, University of Padova, Italy International Gaming Research Unit, Psychology Division, Nottingham Trent University, UK
H I G H L I G H T S • • • • •
The RGQ is a valid scale to assess gambling motives among the general population. The five subscales of the RGQ were confirmed among gender and age subgroups. Mixed-mode gamblers had higher scores for enhancement, recreation and money motives. Mixed-mode gambler males reported higher enhancement and recreation motives. Mixed-mode gamblers aged 16–34 years reported higher enhancement, recreation and money motives.
a r t i c l e
i n f o
Available online 24 February 2015 Keywords: Gambling motives Reasons for gambling Confirmatory factor analysis Online gambling Subgroups
a b s t r a c t Introduction: The aims of the study are to validate and further psychometrically test the five-dimensional structure of the Reasons for Gambling Questionnaire (RGQ), and to test the differences between different types of gamblers (i.e., offline gamblers who gambled in-person only vs. mixed-mode gamblers who gambled both online and offline) on the five dimensions of the RGQ. Methods: Data from the 2010 British Gambling Prevalence Survey (BGPS) were used. The analyzed data comprised 5677 individuals (52.7% female; mean age = 47.64 years; SD = 17.82). Confirmatory factor analysis and independent-sample t-tests were applied. Results: The five-dimensional structure of the RGQ was confirmed in the general sample and among gender and age subgroups. Furthermore, mixed-mode gamblers (MMGs) who gambled both online and offline had higher scores for enhancement, recreation and money motives than offline gamblers that gambled in-person only (IPGs). In addition among males, there was a significant difference in the scores for enhancement and recreation motives across MMGs and IPGs. Among past-year gamblers aged 16–34 years, MMGs had higher scores for enhancement, recreational and monetary motives than IPGs while among past-year gamblers aged 35–55 years, MMGs had higher scores for enhancement and recreational motives than IPGs. Conclusions: The results are consistent with a previous test of the RGQ and the findings indicate that the RGQ is a valid instrument to assess gambling motives among the general population. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction The most recent British Gambling Prevalence Survey (BGPS) published in 2011 showed that problem gambling in Great Britain had risen from 0.6% to 0.9% (i.e., a 50% increase) compared to the previous BGPS published in 2007 (Wardle, Moody, Griffiths, Orford, & Volberg, 2011a; Wardle et al., 2011b) using the DSM-IV criteria American Psychiatric Association (APA) (1994). One of the most important aims of the survey was to examine reasons for gambling that, in turn, shape ⁎ Corresponding author at: International Gaming Research Unit, Psychology Division, Nottingham Trent University, Burton Street, Nottingham NG1 4BU, UK. E-mail address: mark.griffi[email protected] (M.D. Griffiths).
http://dx.doi.org/10.1016/j.addbeh.2015.01.035 0306-4603/© 2015 Elsevier Ltd. All rights reserved.
gambling behavior. In order to develop prevention programs aimed at promoting responsible gambling, researchers need to have a sound knowledge based on empirical evidence of the reasons as to why people participate in gambling. This is important for any research that aims to uncover determinants of varying levels of gambling involvement (Binde, 2009). Studies suggest that motives for gambling are an important proximal factor related to problematic gambling among young people and adults (e.g., Griffiths, 2011; Stewart, Zack, Collins, & Klein, 2008). Despite the existence of some psychometric instruments developed to assess gambling motives and related constructs — such as the Gambling Motivation Scale (GMS: Chantal, Vallerand, & Vallieres, 1994) and the Gambling Motives Questionnaire (GMQ: Stewart & Zack, 2008) — the
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
Reasons for Gambling Questionnaire (RGQ: Wardle et al., 2011b) was the first standardized measure of gambling motives to be included in a large-scale national prevalence survey. The 15-item RGQ was specifically designed to reflect broad motivations for gambling evident among gamblers in general, after determining that the available scales (i.e., GMS and GMQ) had some gaps in the range of motives identified in previous studies. For example, the GMQ fails to capture motives related to money, because items are derived from the alcohol literature (Dechant & Ellery, 2011). While people generally do not drink alcohol for monetary gain, winning money has frequently been reported among the primary reasons for gambling (e.g., Hodgins, 2008). Based on these considerations, Wardle et al. (2011b) developed the RGQ to measure the five gambling motives implied: social motives (e.g., because it is something that is engaged in with friends and family), monetary motives (e.g., for the chance of winning large amounts of money), enhancement motives (e.g., for the excitement), recreational motives (e.g., to fill time), and coping motives (e.g., to relieve tension). Thus, the RGQ has a number of distinct advantages in that it: (i) can be applied to the general population (including those with gambling problems), (ii) includes a wide range of gambling motives, (iii) can measure general gambling motives, rather than relative to a specific gambling activity (Wardle et al., 2011b). To date, the RGQ has been developed and analyzed in the large-scale BGPS population study (Wardle et al., 2011b) and was also employed in the second Social and Economic Impact Study (SEIS) of Gambling in Tasmania study conducted in 2011 in Australia (Francis, Dowling, Jackson, Christensen, and Wardle 2014). In the Australian study, the most commonly endorsed reasons for gambling were for fun, for the chance of winning big money, because it is something to do with friends and family, to be sociable, and for the excitement. The objectives of this study were to test, validate, and further psychometrically analyze the RGQ in the same sample that was used in the original BGPS study (Wardle et al., 2011b). The first aim was to confirm the five dimensional structures (i.e., social, monetary, enhancement, recreation, and coping) and other measurement properties such as model fit, item loadings, and internal consistencies. The second aim was to demonstrate whether the RGQ performs as well across gender (male vs. female) and age band (i.e., 16–34 years, 35–55 years and 55 + years). The third aim was to demonstrate concurrent validity by testing the differences between different types of gamblers (more specifically, those who gamble in person only gamblers [i.e., offline] vs. those who gamble both in person and online [mixed-mode]) on the five dimensions of the RGQ. This is because Wardle et al. (2011b) found that the reasons for gambling vary significantly between different groups of gamblers (problem gamblers, regular gamblers). In addition, previous studies have shown that online gamblers and non-online gamblers display different motivations to gambling related to the online specific characteristics (e.g. privacy, greater variety of games) (Gainsbury, Wood, Russell, Hing, & Blaszczynski, 2012; McCormack & Griffiths, 2012; Parke & Griffiths, 2011). Considering the motivations for gambling evident among gamblers in general, Lloyd et al. (2010) demonstrated that among internet gamblers, gambling to regulate mood, gambling for monetary motives, and gambling for enjoyment were higher in individuals at a heightened risk of problematic gambling. In a more recent study, Dowling, Lorains, and Jackson (2015) found that reasons for gambling among internet gamblers were more likely to be for a challenge, for positive feelings, for fun and/or excitement, or to relieve boredom compared non-internet gamblers. Thus in this study, it was hypothesized that mixed-mode gamblers (those who gambled both online and offline), were more likely to gamble for enhancement, coping, and money reasons, than those who gambled offline only.
277
2. Method 2.1. Participants Data were extracted from the 2010 British Gambling Prevalence Survey and further methodological details are provided in the originally published study (Wardle et al., 2011b). A total of 7756 individuals participated in the BGPS. In order to specifically address gambling motives, the 2039 individuals (26.29%) who had not gambled in the 12 months prior to participating in the survey were excluded from analysis, in addition to 40 individuals (0.5%) that were excluded because they failed to answer more than one question related to gambling motives. The analyzed data therefore comprised 5677 individuals (52.7% female). The mean age of the final sample was 47.64 years (SD = 17.82). 2.2. Instruments 2.2.1. Gambling motives All participants who had gambled in the 12 months prior to participating in the survey were administered the RGQ (Wardle et al., 2011b). All 15 items of the RGQ are statements concerning the frequency of gambling for five distinct dimensions (i.e., enhancement, recreation, social, coping and monetary). Participants were asked to consider all the times they had gambled in the past 12 months and to indicate how often they had gambled for each given motive. Items were answered on a 4-point Likert scale ranging from “never” (coded as 1) to “always” (coded as 4). 2.2.2. Gambling participation Participants were required to indicate whether they had taken part in any of 16 different gambling activities for money in the previous 12 months. Of these 16 activities, nine could be gaged in both online and offline. For activities where different modes of participation were available, participants were asked to report whether they had undertaken the activity in-person (i.e., offline), online, or both. Consistent with a previous categorization using BGPS data (i.e., Wardle et al., 2011a) gamblers were considered ‘offline gamblers’ when they reported gambling in-person only and ‘mixed-mode gamblers’ when they reported gambling both online and offline for at least one activity in the past year. 2.3. Analysis First, given the a priori knowledge of a factor structure1, a confirmatory factor analysis (CFA) using robust weighted least squares for ordinal items (e.g., Likert-type scales) was used to test the structure of the RGQ. The R program (R Core Team, 2013) was used with the “lavaan” library (Rosseel, 2012). To evaluate the overall model fit, the following were considered: goodness of fit index (GFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA) [90% confidence interval (CI)]. Cronbach's alpha (and CIs) assessed internal consistencies of the dimensions. To determine the equivalence of factor structure in different subgroups, a multi-group CFA was also performed to examine measurement invariance of the RGQ across gender (males vs. females), and age (i.e., the age band groups 16–34 years, 35– 55 years, and 55+ years). A hierarchical approach was taken by successively constraining model parameters and comparing changes in model fit (Steenkamp & Baumgartner, 1998). Three models (i.e., configural, metric, and scalar) were estimated and are considered prerequisites for meaningful across-group comparisons based on factor scales. The use of Δ X2 values has been criticized because of their sensitivity to sample size (Cheung & Rensvold, 2002). For this reason, testing for 1 In the original analysis of the RGQ, a final principal component factor analysis was performed on 14 items. The item “because I'm worried about not winning if I don't play” was not included in this analysis. Further details are provided in the original published study (Wardle et al., 2011b).
278
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
Table 1 Results of a confirmatory factor analysis (standardized item loadings), mean frequencies, and internal consistencies of individuals' motivations for gambling on the Reasons for Gambling Questionnaire (N = 5667). Gambled for these reasons in the past 12 months…
Five-factor model RGQ Enh
For the mental challenge or to learn about the game or activity To compete with others (e.g. bookmaker, other gamblers) Because of the sense of achievement when I win Because it's exciting As a hobby or a pastime To escape boredom or to fill my time Because it's fun To relax Because it's something that I do with my friends or family To be sociable Because it helps when I'm feeling tense To impress other people For the chance of winning big money To make money Internal consistency Cronbach's alpha (CI 95%)
Recr
Soc
Cop
.72 .75 .63 .84 .88 .78 .83 .85 .64 .98
.66 (.65–.68)
.68 (.67–.70)
.59 (.57–.61)
% endorsementa
1.23 (.60) 1.10 (.40) 1.63 (.93) 1.76 (.94) 1.66 (.99) 1.27 (.64) 2.53 (1.13) 1.32 (.70) 1.74 (1.01) 1.50 (.85) 1.07 (.33) 1.05 (.28) 2.79 (1.16) 2.24 (1.25)
15.8 6.6 39.8 49.3 37.7 18.4 77.9 21.6 43.7 32.1 5.4 3.4 82.9 58.8
Mon
.84 .77 .73 .78
.69 (.67–.70)
Mean (SD)
.68 (.66–.70)
Note: All factor loadings are significant (p b .001). Model fit: X2 = 1368.36, df = 67, CFI = 0.98, GFI = 0.99, RMSEA = .058 (.056–.061). Mean frequency based on a scale of: 1 = never, 2 = sometimes, 3 = often, 4 = always. Enh = Enhancement; Rec = Recreation; Soc = Social; Cop = Coping; Mon = money. a % endorsement for an item represents the percentage of participants who responded at least 2 = sometimes to that item.
invariance was examined through the practical perspective (Byrne & Stewart, 2006), which recommends that invariance can be based on two criteria: (a) the multigroup factor model exhibits an adequate fit to the data and (b) the change in values for fit indices (e.g., ΔCFI, ΔRMSEA) is negligible. A ΔCFI of larger than 0.01 and a change larger than .015 in RMSEA is indicative of non-invariance (Cheung & Rensvold, 2002; Gilson et al., 2013). To confirm the concurrent validity of the RGQ, independent-sample t-tests to compare gambling motives in gambling subgroups were performed. Due to the large sample size, an alpha of 0.001 was used and effect sizes are reported for all t-tests. 3. Results 3.1. Confirming the factor structure The results of the CFA revealed high and homogenous item loadings (see Table 1). For each dimension, item loadings were 0.50 or higher. Greatest endorsement was shown for the chance of winning large amounts of money and gambling being fun. There were substantial endorsements for monetary and recreational motives. In contrast, coping motives were infrequently endorsed, with the “to impress other people” item showing the lowest endorsement compared to other items. The fit indices revealed a good model fit with CFI and GFI values of 0.98 and 0.99, respectively. Similarly, the RMSEA value was .058 (.056–.061). The internal consistency of all five gambling motive factors was just within conventionally accepted limits, varying from α = .59 (CI = .57–.61) to .69 (CI = .67–.70).
1496.6, CFI = .977, RMSEA = .058 [.055–.060]). All item loadings related to each factor, were constrained to equality. Fit statistics showed that this model (compared to the configural model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .001) suggesting that the five-factor RGQ assesses similar underlying factors across both males and females. Scalar invariance was tested by constraining the intercept of each item while maintaining constraints on the factor loadings (X2(166) = 1577.1, CFI = .976, RMSEA = .055 [.052–.057]). Fit statistics showed that this model (compared to the metric model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .003). The CFA was also performed separately across different age band groups. The model fit was excellent within each year group: 16–34 years (X2(67) = 588.6, CFI = .975, RMSEA = .072 [.067–.077]), for 35–54 years (X2(67) = 566.1, CFI = .975, RMSEA = .059 [.055–.064]) and for 55 + years (X2(67) = 372.4, CFI = .984, RMSEA = .047 [.042–.052]). Fit indices showed that the configural model had adequate fit for the data (X2(201) = 1527.1, CFI = .978, RMSEA = .059 [0.056–0.062]) again suggesting that the factor structure is similar across groups. Fit statistics showed that the metric model (X2(219) = 1577.4, CFI = .977, RMSEA = .057 [.055–.060]) (compared to the configural model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .002) suggesting that the five-factor RGQ assesses similar underlying factors across age band groups. Fit statistics showed that scalar model (X2(265) = 2029.1, CFI = .971, RMSEA = .059 [.057–.062]) (compared to the metric model) did not result in a significant degradation of fit (ΔCFI = .006; ΔRMSEA = −.002). 3.3. Mean differences in gambling motives across gambling subgroups
3.2. Measurement invariance across gender and age Before measurement invariance testing, the five-factor RGQ model was estimated separately in both male and females. Results (see Table 2) demonstrated that the model fit was adequate to excellent for both men (X2(67) = 709.4, CFI = .98, RMSEA = .060 [.056–.064]) and women (X2(67) = 731.6, CFI = .97, RMSEA = .058 [.054–.061]). A configural model was first established as a baseline model, and all parameters were freely estimated (unconstrained) across gender. Fit indices showed that this model had adequate fit for the data (X2(134) = 1441.1, CFI = .978, RMSEA = .059 [.056–.061]) suggesting that the factor structure is similar across groups. A subsequent metric model that tested for invariance of all factor loadings was established (X2(143) =
Independent-sample t-tests to compare gambling motives in gambling subgroups were performed to determine the concurrent validity of the RGQ (see Table 3). Among the past-year gamblers, a total of 4614 (81.1%) were classified as in person (i.e., offline) only gamblers (IPGs) and 1075 (18.9%) were classified as mixed-mode gamblers (MMGs) as they gambled both online and offline. MMGs had higher scores for enhancement, recreational, and monetary motives when compared to IPGs. In addition, other independent-sample t-tests were conducted to compare gambling motives in gambling subgroups across gender and age. Among males, there was a significant difference in the scores for enhancement and recreation motives across MMGs and IPGs. Among past-year gamblers aged 16–34 years, MMGs had higher
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
279
Table 2 Fit indices for measurement invariance tests on the Reasons for Gambling Questionnaire (N = 5667). Model Gender Men Women Configural invariance Metric invariance Scalar invariance Age 16–34 35–54 55+ Configural invariance Metric invariance Scalar invariance
n
X2
df
p
CFI
ΔCFI
RMSEA (90% CI)
ΔRMSEA
2686 2991 5677 5677 5677
709.4 731.6 1441.1 1496.6 1577.1
67 67 134 143 166
b.001 b.001 b.001 b.001 b.001
.980 .970 .978 .977 .976
– – – .001 .001
.060 (.056–.064) .058 (.054–.061) .059 (.056–.061) .058 (.055–.060) .055 (.052–.057)
– – – .001 .003
1505 2113 2059 5677 5677 5677
588.6 566.1 372.4 1527.1 1577.4 2029.1
67 67 67 201 219 265
b.001 b.001 b.001 b.001 b.001 b.001
.975 .975 .984 .978 .977 .971
– – – – .001 .006
.072 (.067–.077) .059 (.055–.064) .047 (.042–.052) .059 (.056–.062) .057 (.055–.060) .059 (.057–.062)
– – – – .002 −.002
scores for enhancement, recreational and monetary motives than IPGs while among past-year gamblers aged 35–55 years, MMGs had higher scores for enhancement and recreational motives than IPGs. 4. Discussion In addition to the psychometric properties of the RGQ reported by Wardle et al. (2011b), the present study extends the information about this new measure of gambling motives by using more robust statistical analyses to confirm the five-dimensional structure of the RGQ. More specifically, the present study, provided other measurement properties not investigated in the original analysis by Wardle et al. (2011b) such as model fit, internal consistencies, and measurement invariance of the RGQ across gender (males vs. females), and age (i.e., the age band groups 16–34 years, 35–55 years, and 55+ years). More specifically, results of the present study indicate that the hypothesized fivefactor structure of the RGQ (Wardle et al., 2011b) provided a good fit to the data. Most dimensions had acceptable internal consistency with the expectancy dimension coming close to the required α = 0.70
threshold (Nunnally & Bernstein, 1994). The relative rates of endorsement — with monetary and recreational motives being most strongly endorsed — were highly consistent with findings of a recent study conducted in a large-scale survey of an Australian population (i.e., Francis et al., 2014) that reported a five-factor structure that is a little different from that reported in the BGPS (i.e., for money, for regulating internal state, to gain positive feelings, for social reasons, and for a challenge). Furthermore, the invariance of the RGQ across different gender and age groups was confirmed. According to Lloyd, Doll, Hawton, et al. (2010) mixed-mode gamblers that gambled online have higher scores on enhancement, recreational and monetary motives than offline (inperson) only gamblers. In contrast, there was no difference in coping motives. This result could be considered alongside the results of Stewart and Zack (2008) where coping motives were unique independently covarying with gambling severity than gambling behavior in pathological gamblers. In addition, some of these differences were present among male gamblers and among past-year gamblers aged 16– 34 years and 35–55 years. More specifically regarding age, the difference between MMGs and IPGs concerning monetary motives was
Table 3 Comparison of mean factor scores for in-person (offline) only gamblers (IPGs) and mixed-mode (both online and offline) gamblers (MMGs) according to gender and age (M = 5667) on the Reasons for Gambling Questionnaire.
Total IPGs (n = 4584) MMGs (n = 1071) t(5653); p value; d Gender Male IPGs (n = 2098) MMGs (n = 579) t(2675); p value; d Female IPGs (n = 2486) MMGs (n = 492) t(2976); p value; d Age 16–34 IPGs (n = 1072) MMGs (n = 430) t(1500); p value; d 35–55 IPGs (n = 1642) MMGs (n = 468) t(2108); p value; d 55+ IPGs (n = 1870) MMGs (n = 173) t(2041); p value; d
Money
Recreation
Enhancement
Coping
Social
2.48 (1.06) 2.64 (.97) 4.38; b.001; .16
1.67 (.62) 1.80 (.63) 5.80; b.001; .21
1.39 (.51) 1.57 (.62) 9.82; b.001; .32
1.05 (.24) 1.08 (.31) 3.70; .002; .10
1.63 (.83) 1.64 (.75) .66; .505; .02
2.58 (1.06) 2.70 (.97) 2.46; .014; .12
1.71 (.64) 1.87 (.67) 5.50; b.001; .24
1.44 (.56) 1.71 (.70) 9.44; b.001; .43
1.06 (.27) 1.11 (.36) 3.02; .003; .16
1.63 (.81) 1.70 (.76) 1.84; .06; .09
2.39 (1.06) 2.56 (.97) 3.16; .002; .17
1.63 (.61) 1.70 (.57) 1.98; .04; .12
1.35 (.45) 1.41 (.47) 2.63; .009; .13
1.04 (.22) 1.05 (.24) .80; .42; .04
1.61 (.85) 1.57 (.72) −1.08; .28; .05
2.33 (.99) 2.55 (.95) 3.88; b.001; .23
1.66 (.56) 1.84 (.63) 5.25; b.001; .31
1.51 (.55) 1.72 (.64) 6.37; b.001; .35
1.08 (.30) 1.12 (.39) 2.29; .02; .11
1.66 (.79) 1.73 (.75) 1.50; .13; .09
2.51 (1.04) 2.67 (.96) 2.88; .004; .16
1.61 (.57) 1.76 (.62) 5.02; b.001; .25
1.38 (.48) 1.49 (.59) 4.10; b.001; .20
1.05 (.22) 1.06 (.26) 1.27; .20; .04
1.57 (.78) 1.58 (.72) .36; .72; .01
2.54 (1.11) 2.76 (1.02) 2.55; .011; .20
1.73 (.69) 1.77 (.68) .68; .50; .06
1.34 (.50) 1.42 (.57) 1.97; .04; .15
1.04 (.22) 1.05 (.22) .449; .65; .04
1.63 (.89) 1.54 (.79) −1.26; .20; .11
Note: Due to the large sample size, an alpha of 0.001 was used.
280
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
significant only in the younger age group, suggesting that for these gamblers, the prospect of winning money represents a more important determinant of motivation than older gamblers. Thus, the difference between gambling subgroups may be explained in part by motivational factors (Griffiths, Wardle, Orford, Sproston, & Erens, 2011). The present study is not without its limitations. Firstly, the study did not include usual concurrent validity that generally tests the links between motives and problematic behaviors (e.g., multiple regression analyses) because the problem gambling prevalence rates in the BGPS were very low (0.9%). Outcomes for problem gambling and gamblingrelated problems should be included in future studies in order to gain a more comprehensive analysis of the concurrent validity. Secondly, the ‘coping’ dimension with only two items shows low Cronbach's alpha. Adding more items to the scale could be recommended for future research. Thirdly, as with all survey data, the data collected were based on self-report and subject to attendant problems of this mode of collection (e.g., social desirability biases, memory recall biases, etc.). Fourthly, although the overall sample size was large, the base sizes for some of the past-year gambling subgroups were small. Therefore, the subgroups presented in the present study were not definitive. It is plausible that further subgroups exist and other analytical techniques could perhaps be used to examine this more thoroughly. 5. Conclusion Examining the findings as a whole, the present study is one of a few studies (i.e., Francis et al., 2014; Wardle, et al., 2011a; Wardle et al., 2011b) that have provided detailed evidence for the utility and usefulness of the RGQ. The results confirmed the overall five-dimensional structure of the RGQ, and across gender (males vs. females), and age (16–34 years, 35–55 years and 55+ years). Role of funding sources No external funding source for this study has been received. Contributors Author MDG designed the study and wrote the protocol. All authors conducted literature searches and provided summaries of previous research studies. Author NC conducted the statistical analysis. Author NC wrote the first draft of the manuscript and all authors contributed to and have approved the final manuscript. Conflict of interest Mark Griffiths has received funding for a number of research projects in the area of gambling education for youth, social responsibility in gambling and gambling treatment from the Responsibility in Gambling Trust, a charitable body who funds its research program based on donations from the gambling industry. The author also undertakes consultancy for various gaming companies in the area of social responsibility in gambling. All other authors declare that they have no conflicts of interest.
References American Psychiatric Association (APA) (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington DC: American Psychiatric Association.
Binde, P. (2009). Gambling motivation and involvement: A review of social science research. http://www.responsiblegambling.org/articles/binde-gambling-motivationreview.pdf (Accessed 23 December 2014) Byrne, B. M., & Stewart, S. M. (2006). The MACS approach to testing for multigroup invariance of a second-order structure: A walk through the process. Structural Equation Modeling, 13, 287–321. Chantal, Y., Vallerand, R. J., & Vallieres, E. F. (1994). Construction et validation l'echelle de motivation relative aux jeux de hasard et d'argent. On the development and validation of the Gambling Motivation Scale (GMS). Society and Leisure, 17, 189–212. Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9, 233–255. Dechant, K., & Ellery, M. (2011). The effect of including a monetary motive item on the gambling motives questionnaire in a sample of moderate gamblers. Journal of Gambling Studies, 27, 331–344. Dowling, N. A., Lorains, F. K., & Jackson, A. C. (2015). Are the profiles of past-year internet gamblers generalizable to regular internet gamblers? Computers in Human Behavior, 43, 118–128. Francis, K. L., Dowling, N. A., Jackson, A. C., Christensen, D. R., & Wardle, H. (2014). Gambling motives: Application of the Reasons for Gambling Questionnaire in an Australian population survey. Journal of Gambling Studies. http://dx.doi.org/10.1007/ s10899-014-9458-1 (Epub ahead of print). Gainsbury, S., Wood, R., Russell, A., Hing, N., & Blaszczynski, A. (2012). A digital revolution: Comparison of demographic profiles, attitudes and gambling behavior of Internet and non-Internet gamblers. Computers in Human Behavior, 28, 1388–1398. Gilson, K. M., Bryant, C., Bei, B., Komiti, A., Jackson, H., & Judd, F. (2013). Validation of the Drinking Motives Questionnaire (DMQ) in older adults. Addictive Behaviors, 38, 2196–2202. Griffiths, M. D. (2011). Adolescent gambling. In B. B. Brown, & M. Prinstein (Eds.), Encyclopedia of Adolescence. Vol. 3. (pp. 11–20). San Diego: Academic Press. Griffiths, M. D., Wardle, H., Orford, J., Sproston, K., & Erens, B. (2011). Internet gambling, health, smoking and alcohol use: Findings from the 2007 British Gambling Prevalence Survey. International Journal of Mental Health and Addiction, 9, 1–11. Hodgins, D. C. (2008). What we see depends mainly on what we look for (commentary). Addiction, 103, 1118–1119. Lloyd, J., Doll, H., Hawton, K., Dutton, W. H., Geddes, J. R., Goodwin, G. M., et al. (2010). How psychological symptoms relate to different motivations for gambling: An online study of internet gamblers. Biological Psychiatry, 68, 733–740. McCormack, A., & Griffiths, M. D. (2012). Motivating and inhibiting factors in online gambling behaviour: A grounded theory study. International Journal of Mental Health and Addiction, 10, 39–53. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Parke, A., & Griffiths, M. D. (2011). Effects on gambling behaviour of developments in information technology: A grounded theoretical framework. International Journal of Cyber Behaviour, Psychology and Learning, 1(4), 36–48. R Core Team (2013). R: A language and environment for statistical computing [Computer software manual]. Vienna, Austria Available from http://www.R-project.org/. Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement invariance in cross-national consumer research. Journal of Consumer Research, 25, 78–90. Stewart, S. H., & Zack, M. (2008). Development and psychometric evaluation of a three‐dimensional Gambling Motives Questionnaire. Addiction, 103, 1110–1117. Stewart, S. H., Zack, M., Collins, P., & Klein, R. M. (2008). Subtyping pathological gamblers on the basis of affective motivations for gambling: Relations to gambling problems, drinking problems, and affective motivations for drinking. Psychology of Addictive Behaviors, 22, 257–268. Wardle, H., Moody, A., Griffiths, M., Orford, J., & Volberg, R. (2011a). Defining the online gambler and patterns of behaviour integration: Evidence from the British Gambling Prevalence Survey 2010. International Gambling Studies, 11, 339–356. Wardle, H., Moody, A., Spence, S., Orford, J., Volberg, R., Jotangia, D., et al. (2011b). British gambling prevalence survey 2010. London: The Stationery Office.
Contents lists available at ScienceDirect
Addictive Behaviors
Short Communication
Validation of the reasons for gambling questionnaire (RGQ) in a British population survey Natale Canale a, Massimo Santinello a, Mark D. Griffiths b,⁎ a b
Department of Developmental and Social Psychology, University of Padova, Italy International Gaming Research Unit, Psychology Division, Nottingham Trent University, UK
H I G H L I G H T S • • • • •
The RGQ is a valid scale to assess gambling motives among the general population. The five subscales of the RGQ were confirmed among gender and age subgroups. Mixed-mode gamblers had higher scores for enhancement, recreation and money motives. Mixed-mode gambler males reported higher enhancement and recreation motives. Mixed-mode gamblers aged 16–34 years reported higher enhancement, recreation and money motives.
a r t i c l e
i n f o
Available online 24 February 2015 Keywords: Gambling motives Reasons for gambling Confirmatory factor analysis Online gambling Subgroups
a b s t r a c t Introduction: The aims of the study are to validate and further psychometrically test the five-dimensional structure of the Reasons for Gambling Questionnaire (RGQ), and to test the differences between different types of gamblers (i.e., offline gamblers who gambled in-person only vs. mixed-mode gamblers who gambled both online and offline) on the five dimensions of the RGQ. Methods: Data from the 2010 British Gambling Prevalence Survey (BGPS) were used. The analyzed data comprised 5677 individuals (52.7% female; mean age = 47.64 years; SD = 17.82). Confirmatory factor analysis and independent-sample t-tests were applied. Results: The five-dimensional structure of the RGQ was confirmed in the general sample and among gender and age subgroups. Furthermore, mixed-mode gamblers (MMGs) who gambled both online and offline had higher scores for enhancement, recreation and money motives than offline gamblers that gambled in-person only (IPGs). In addition among males, there was a significant difference in the scores for enhancement and recreation motives across MMGs and IPGs. Among past-year gamblers aged 16–34 years, MMGs had higher scores for enhancement, recreational and monetary motives than IPGs while among past-year gamblers aged 35–55 years, MMGs had higher scores for enhancement and recreational motives than IPGs. Conclusions: The results are consistent with a previous test of the RGQ and the findings indicate that the RGQ is a valid instrument to assess gambling motives among the general population. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction The most recent British Gambling Prevalence Survey (BGPS) published in 2011 showed that problem gambling in Great Britain had risen from 0.6% to 0.9% (i.e., a 50% increase) compared to the previous BGPS published in 2007 (Wardle, Moody, Griffiths, Orford, & Volberg, 2011a; Wardle et al., 2011b) using the DSM-IV criteria American Psychiatric Association (APA) (1994). One of the most important aims of the survey was to examine reasons for gambling that, in turn, shape ⁎ Corresponding author at: International Gaming Research Unit, Psychology Division, Nottingham Trent University, Burton Street, Nottingham NG1 4BU, UK. E-mail address: mark.griffi[email protected] (M.D. Griffiths).
http://dx.doi.org/10.1016/j.addbeh.2015.01.035 0306-4603/© 2015 Elsevier Ltd. All rights reserved.
gambling behavior. In order to develop prevention programs aimed at promoting responsible gambling, researchers need to have a sound knowledge based on empirical evidence of the reasons as to why people participate in gambling. This is important for any research that aims to uncover determinants of varying levels of gambling involvement (Binde, 2009). Studies suggest that motives for gambling are an important proximal factor related to problematic gambling among young people and adults (e.g., Griffiths, 2011; Stewart, Zack, Collins, & Klein, 2008). Despite the existence of some psychometric instruments developed to assess gambling motives and related constructs — such as the Gambling Motivation Scale (GMS: Chantal, Vallerand, & Vallieres, 1994) and the Gambling Motives Questionnaire (GMQ: Stewart & Zack, 2008) — the
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
Reasons for Gambling Questionnaire (RGQ: Wardle et al., 2011b) was the first standardized measure of gambling motives to be included in a large-scale national prevalence survey. The 15-item RGQ was specifically designed to reflect broad motivations for gambling evident among gamblers in general, after determining that the available scales (i.e., GMS and GMQ) had some gaps in the range of motives identified in previous studies. For example, the GMQ fails to capture motives related to money, because items are derived from the alcohol literature (Dechant & Ellery, 2011). While people generally do not drink alcohol for monetary gain, winning money has frequently been reported among the primary reasons for gambling (e.g., Hodgins, 2008). Based on these considerations, Wardle et al. (2011b) developed the RGQ to measure the five gambling motives implied: social motives (e.g., because it is something that is engaged in with friends and family), monetary motives (e.g., for the chance of winning large amounts of money), enhancement motives (e.g., for the excitement), recreational motives (e.g., to fill time), and coping motives (e.g., to relieve tension). Thus, the RGQ has a number of distinct advantages in that it: (i) can be applied to the general population (including those with gambling problems), (ii) includes a wide range of gambling motives, (iii) can measure general gambling motives, rather than relative to a specific gambling activity (Wardle et al., 2011b). To date, the RGQ has been developed and analyzed in the large-scale BGPS population study (Wardle et al., 2011b) and was also employed in the second Social and Economic Impact Study (SEIS) of Gambling in Tasmania study conducted in 2011 in Australia (Francis, Dowling, Jackson, Christensen, and Wardle 2014). In the Australian study, the most commonly endorsed reasons for gambling were for fun, for the chance of winning big money, because it is something to do with friends and family, to be sociable, and for the excitement. The objectives of this study were to test, validate, and further psychometrically analyze the RGQ in the same sample that was used in the original BGPS study (Wardle et al., 2011b). The first aim was to confirm the five dimensional structures (i.e., social, monetary, enhancement, recreation, and coping) and other measurement properties such as model fit, item loadings, and internal consistencies. The second aim was to demonstrate whether the RGQ performs as well across gender (male vs. female) and age band (i.e., 16–34 years, 35–55 years and 55 + years). The third aim was to demonstrate concurrent validity by testing the differences between different types of gamblers (more specifically, those who gamble in person only gamblers [i.e., offline] vs. those who gamble both in person and online [mixed-mode]) on the five dimensions of the RGQ. This is because Wardle et al. (2011b) found that the reasons for gambling vary significantly between different groups of gamblers (problem gamblers, regular gamblers). In addition, previous studies have shown that online gamblers and non-online gamblers display different motivations to gambling related to the online specific characteristics (e.g. privacy, greater variety of games) (Gainsbury, Wood, Russell, Hing, & Blaszczynski, 2012; McCormack & Griffiths, 2012; Parke & Griffiths, 2011). Considering the motivations for gambling evident among gamblers in general, Lloyd et al. (2010) demonstrated that among internet gamblers, gambling to regulate mood, gambling for monetary motives, and gambling for enjoyment were higher in individuals at a heightened risk of problematic gambling. In a more recent study, Dowling, Lorains, and Jackson (2015) found that reasons for gambling among internet gamblers were more likely to be for a challenge, for positive feelings, for fun and/or excitement, or to relieve boredom compared non-internet gamblers. Thus in this study, it was hypothesized that mixed-mode gamblers (those who gambled both online and offline), were more likely to gamble for enhancement, coping, and money reasons, than those who gambled offline only.
277
2. Method 2.1. Participants Data were extracted from the 2010 British Gambling Prevalence Survey and further methodological details are provided in the originally published study (Wardle et al., 2011b). A total of 7756 individuals participated in the BGPS. In order to specifically address gambling motives, the 2039 individuals (26.29%) who had not gambled in the 12 months prior to participating in the survey were excluded from analysis, in addition to 40 individuals (0.5%) that were excluded because they failed to answer more than one question related to gambling motives. The analyzed data therefore comprised 5677 individuals (52.7% female). The mean age of the final sample was 47.64 years (SD = 17.82). 2.2. Instruments 2.2.1. Gambling motives All participants who had gambled in the 12 months prior to participating in the survey were administered the RGQ (Wardle et al., 2011b). All 15 items of the RGQ are statements concerning the frequency of gambling for five distinct dimensions (i.e., enhancement, recreation, social, coping and monetary). Participants were asked to consider all the times they had gambled in the past 12 months and to indicate how often they had gambled for each given motive. Items were answered on a 4-point Likert scale ranging from “never” (coded as 1) to “always” (coded as 4). 2.2.2. Gambling participation Participants were required to indicate whether they had taken part in any of 16 different gambling activities for money in the previous 12 months. Of these 16 activities, nine could be gaged in both online and offline. For activities where different modes of participation were available, participants were asked to report whether they had undertaken the activity in-person (i.e., offline), online, or both. Consistent with a previous categorization using BGPS data (i.e., Wardle et al., 2011a) gamblers were considered ‘offline gamblers’ when they reported gambling in-person only and ‘mixed-mode gamblers’ when they reported gambling both online and offline for at least one activity in the past year. 2.3. Analysis First, given the a priori knowledge of a factor structure1, a confirmatory factor analysis (CFA) using robust weighted least squares for ordinal items (e.g., Likert-type scales) was used to test the structure of the RGQ. The R program (R Core Team, 2013) was used with the “lavaan” library (Rosseel, 2012). To evaluate the overall model fit, the following were considered: goodness of fit index (GFI), comparative fit index (CFI), and root mean square error of approximation (RMSEA) [90% confidence interval (CI)]. Cronbach's alpha (and CIs) assessed internal consistencies of the dimensions. To determine the equivalence of factor structure in different subgroups, a multi-group CFA was also performed to examine measurement invariance of the RGQ across gender (males vs. females), and age (i.e., the age band groups 16–34 years, 35– 55 years, and 55+ years). A hierarchical approach was taken by successively constraining model parameters and comparing changes in model fit (Steenkamp & Baumgartner, 1998). Three models (i.e., configural, metric, and scalar) were estimated and are considered prerequisites for meaningful across-group comparisons based on factor scales. The use of Δ X2 values has been criticized because of their sensitivity to sample size (Cheung & Rensvold, 2002). For this reason, testing for 1 In the original analysis of the RGQ, a final principal component factor analysis was performed on 14 items. The item “because I'm worried about not winning if I don't play” was not included in this analysis. Further details are provided in the original published study (Wardle et al., 2011b).
278
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
Table 1 Results of a confirmatory factor analysis (standardized item loadings), mean frequencies, and internal consistencies of individuals' motivations for gambling on the Reasons for Gambling Questionnaire (N = 5667). Gambled for these reasons in the past 12 months…
Five-factor model RGQ Enh
For the mental challenge or to learn about the game or activity To compete with others (e.g. bookmaker, other gamblers) Because of the sense of achievement when I win Because it's exciting As a hobby or a pastime To escape boredom or to fill my time Because it's fun To relax Because it's something that I do with my friends or family To be sociable Because it helps when I'm feeling tense To impress other people For the chance of winning big money To make money Internal consistency Cronbach's alpha (CI 95%)
Recr
Soc
Cop
.72 .75 .63 .84 .88 .78 .83 .85 .64 .98
.66 (.65–.68)
.68 (.67–.70)
.59 (.57–.61)
% endorsementa
1.23 (.60) 1.10 (.40) 1.63 (.93) 1.76 (.94) 1.66 (.99) 1.27 (.64) 2.53 (1.13) 1.32 (.70) 1.74 (1.01) 1.50 (.85) 1.07 (.33) 1.05 (.28) 2.79 (1.16) 2.24 (1.25)
15.8 6.6 39.8 49.3 37.7 18.4 77.9 21.6 43.7 32.1 5.4 3.4 82.9 58.8
Mon
.84 .77 .73 .78
.69 (.67–.70)
Mean (SD)
.68 (.66–.70)
Note: All factor loadings are significant (p b .001). Model fit: X2 = 1368.36, df = 67, CFI = 0.98, GFI = 0.99, RMSEA = .058 (.056–.061). Mean frequency based on a scale of: 1 = never, 2 = sometimes, 3 = often, 4 = always. Enh = Enhancement; Rec = Recreation; Soc = Social; Cop = Coping; Mon = money. a % endorsement for an item represents the percentage of participants who responded at least 2 = sometimes to that item.
invariance was examined through the practical perspective (Byrne & Stewart, 2006), which recommends that invariance can be based on two criteria: (a) the multigroup factor model exhibits an adequate fit to the data and (b) the change in values for fit indices (e.g., ΔCFI, ΔRMSEA) is negligible. A ΔCFI of larger than 0.01 and a change larger than .015 in RMSEA is indicative of non-invariance (Cheung & Rensvold, 2002; Gilson et al., 2013). To confirm the concurrent validity of the RGQ, independent-sample t-tests to compare gambling motives in gambling subgroups were performed. Due to the large sample size, an alpha of 0.001 was used and effect sizes are reported for all t-tests. 3. Results 3.1. Confirming the factor structure The results of the CFA revealed high and homogenous item loadings (see Table 1). For each dimension, item loadings were 0.50 or higher. Greatest endorsement was shown for the chance of winning large amounts of money and gambling being fun. There were substantial endorsements for monetary and recreational motives. In contrast, coping motives were infrequently endorsed, with the “to impress other people” item showing the lowest endorsement compared to other items. The fit indices revealed a good model fit with CFI and GFI values of 0.98 and 0.99, respectively. Similarly, the RMSEA value was .058 (.056–.061). The internal consistency of all five gambling motive factors was just within conventionally accepted limits, varying from α = .59 (CI = .57–.61) to .69 (CI = .67–.70).
1496.6, CFI = .977, RMSEA = .058 [.055–.060]). All item loadings related to each factor, were constrained to equality. Fit statistics showed that this model (compared to the configural model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .001) suggesting that the five-factor RGQ assesses similar underlying factors across both males and females. Scalar invariance was tested by constraining the intercept of each item while maintaining constraints on the factor loadings (X2(166) = 1577.1, CFI = .976, RMSEA = .055 [.052–.057]). Fit statistics showed that this model (compared to the metric model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .003). The CFA was also performed separately across different age band groups. The model fit was excellent within each year group: 16–34 years (X2(67) = 588.6, CFI = .975, RMSEA = .072 [.067–.077]), for 35–54 years (X2(67) = 566.1, CFI = .975, RMSEA = .059 [.055–.064]) and for 55 + years (X2(67) = 372.4, CFI = .984, RMSEA = .047 [.042–.052]). Fit indices showed that the configural model had adequate fit for the data (X2(201) = 1527.1, CFI = .978, RMSEA = .059 [0.056–0.062]) again suggesting that the factor structure is similar across groups. Fit statistics showed that the metric model (X2(219) = 1577.4, CFI = .977, RMSEA = .057 [.055–.060]) (compared to the configural model) did not result in a significant degradation of fit (ΔCFI = .001; ΔRMSEA = .002) suggesting that the five-factor RGQ assesses similar underlying factors across age band groups. Fit statistics showed that scalar model (X2(265) = 2029.1, CFI = .971, RMSEA = .059 [.057–.062]) (compared to the metric model) did not result in a significant degradation of fit (ΔCFI = .006; ΔRMSEA = −.002). 3.3. Mean differences in gambling motives across gambling subgroups
3.2. Measurement invariance across gender and age Before measurement invariance testing, the five-factor RGQ model was estimated separately in both male and females. Results (see Table 2) demonstrated that the model fit was adequate to excellent for both men (X2(67) = 709.4, CFI = .98, RMSEA = .060 [.056–.064]) and women (X2(67) = 731.6, CFI = .97, RMSEA = .058 [.054–.061]). A configural model was first established as a baseline model, and all parameters were freely estimated (unconstrained) across gender. Fit indices showed that this model had adequate fit for the data (X2(134) = 1441.1, CFI = .978, RMSEA = .059 [.056–.061]) suggesting that the factor structure is similar across groups. A subsequent metric model that tested for invariance of all factor loadings was established (X2(143) =
Independent-sample t-tests to compare gambling motives in gambling subgroups were performed to determine the concurrent validity of the RGQ (see Table 3). Among the past-year gamblers, a total of 4614 (81.1%) were classified as in person (i.e., offline) only gamblers (IPGs) and 1075 (18.9%) were classified as mixed-mode gamblers (MMGs) as they gambled both online and offline. MMGs had higher scores for enhancement, recreational, and monetary motives when compared to IPGs. In addition, other independent-sample t-tests were conducted to compare gambling motives in gambling subgroups across gender and age. Among males, there was a significant difference in the scores for enhancement and recreation motives across MMGs and IPGs. Among past-year gamblers aged 16–34 years, MMGs had higher
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
279
Table 2 Fit indices for measurement invariance tests on the Reasons for Gambling Questionnaire (N = 5667). Model Gender Men Women Configural invariance Metric invariance Scalar invariance Age 16–34 35–54 55+ Configural invariance Metric invariance Scalar invariance
n
X2
df
p
CFI
ΔCFI
RMSEA (90% CI)
ΔRMSEA
2686 2991 5677 5677 5677
709.4 731.6 1441.1 1496.6 1577.1
67 67 134 143 166
b.001 b.001 b.001 b.001 b.001
.980 .970 .978 .977 .976
– – – .001 .001
.060 (.056–.064) .058 (.054–.061) .059 (.056–.061) .058 (.055–.060) .055 (.052–.057)
– – – .001 .003
1505 2113 2059 5677 5677 5677
588.6 566.1 372.4 1527.1 1577.4 2029.1
67 67 67 201 219 265
b.001 b.001 b.001 b.001 b.001 b.001
.975 .975 .984 .978 .977 .971
– – – – .001 .006
.072 (.067–.077) .059 (.055–.064) .047 (.042–.052) .059 (.056–.062) .057 (.055–.060) .059 (.057–.062)
– – – – .002 −.002
scores for enhancement, recreational and monetary motives than IPGs while among past-year gamblers aged 35–55 years, MMGs had higher scores for enhancement and recreational motives than IPGs. 4. Discussion In addition to the psychometric properties of the RGQ reported by Wardle et al. (2011b), the present study extends the information about this new measure of gambling motives by using more robust statistical analyses to confirm the five-dimensional structure of the RGQ. More specifically, the present study, provided other measurement properties not investigated in the original analysis by Wardle et al. (2011b) such as model fit, internal consistencies, and measurement invariance of the RGQ across gender (males vs. females), and age (i.e., the age band groups 16–34 years, 35–55 years, and 55+ years). More specifically, results of the present study indicate that the hypothesized fivefactor structure of the RGQ (Wardle et al., 2011b) provided a good fit to the data. Most dimensions had acceptable internal consistency with the expectancy dimension coming close to the required α = 0.70
threshold (Nunnally & Bernstein, 1994). The relative rates of endorsement — with monetary and recreational motives being most strongly endorsed — were highly consistent with findings of a recent study conducted in a large-scale survey of an Australian population (i.e., Francis et al., 2014) that reported a five-factor structure that is a little different from that reported in the BGPS (i.e., for money, for regulating internal state, to gain positive feelings, for social reasons, and for a challenge). Furthermore, the invariance of the RGQ across different gender and age groups was confirmed. According to Lloyd, Doll, Hawton, et al. (2010) mixed-mode gamblers that gambled online have higher scores on enhancement, recreational and monetary motives than offline (inperson) only gamblers. In contrast, there was no difference in coping motives. This result could be considered alongside the results of Stewart and Zack (2008) where coping motives were unique independently covarying with gambling severity than gambling behavior in pathological gamblers. In addition, some of these differences were present among male gamblers and among past-year gamblers aged 16– 34 years and 35–55 years. More specifically regarding age, the difference between MMGs and IPGs concerning monetary motives was
Table 3 Comparison of mean factor scores for in-person (offline) only gamblers (IPGs) and mixed-mode (both online and offline) gamblers (MMGs) according to gender and age (M = 5667) on the Reasons for Gambling Questionnaire.
Total IPGs (n = 4584) MMGs (n = 1071) t(5653); p value; d Gender Male IPGs (n = 2098) MMGs (n = 579) t(2675); p value; d Female IPGs (n = 2486) MMGs (n = 492) t(2976); p value; d Age 16–34 IPGs (n = 1072) MMGs (n = 430) t(1500); p value; d 35–55 IPGs (n = 1642) MMGs (n = 468) t(2108); p value; d 55+ IPGs (n = 1870) MMGs (n = 173) t(2041); p value; d
Money
Recreation
Enhancement
Coping
Social
2.48 (1.06) 2.64 (.97) 4.38; b.001; .16
1.67 (.62) 1.80 (.63) 5.80; b.001; .21
1.39 (.51) 1.57 (.62) 9.82; b.001; .32
1.05 (.24) 1.08 (.31) 3.70; .002; .10
1.63 (.83) 1.64 (.75) .66; .505; .02
2.58 (1.06) 2.70 (.97) 2.46; .014; .12
1.71 (.64) 1.87 (.67) 5.50; b.001; .24
1.44 (.56) 1.71 (.70) 9.44; b.001; .43
1.06 (.27) 1.11 (.36) 3.02; .003; .16
1.63 (.81) 1.70 (.76) 1.84; .06; .09
2.39 (1.06) 2.56 (.97) 3.16; .002; .17
1.63 (.61) 1.70 (.57) 1.98; .04; .12
1.35 (.45) 1.41 (.47) 2.63; .009; .13
1.04 (.22) 1.05 (.24) .80; .42; .04
1.61 (.85) 1.57 (.72) −1.08; .28; .05
2.33 (.99) 2.55 (.95) 3.88; b.001; .23
1.66 (.56) 1.84 (.63) 5.25; b.001; .31
1.51 (.55) 1.72 (.64) 6.37; b.001; .35
1.08 (.30) 1.12 (.39) 2.29; .02; .11
1.66 (.79) 1.73 (.75) 1.50; .13; .09
2.51 (1.04) 2.67 (.96) 2.88; .004; .16
1.61 (.57) 1.76 (.62) 5.02; b.001; .25
1.38 (.48) 1.49 (.59) 4.10; b.001; .20
1.05 (.22) 1.06 (.26) 1.27; .20; .04
1.57 (.78) 1.58 (.72) .36; .72; .01
2.54 (1.11) 2.76 (1.02) 2.55; .011; .20
1.73 (.69) 1.77 (.68) .68; .50; .06
1.34 (.50) 1.42 (.57) 1.97; .04; .15
1.04 (.22) 1.05 (.22) .449; .65; .04
1.63 (.89) 1.54 (.79) −1.26; .20; .11
Note: Due to the large sample size, an alpha of 0.001 was used.
280
N. Canale et al. / Addictive Behaviors 45 (2015) 276–280
significant only in the younger age group, suggesting that for these gamblers, the prospect of winning money represents a more important determinant of motivation than older gamblers. Thus, the difference between gambling subgroups may be explained in part by motivational factors (Griffiths, Wardle, Orford, Sproston, & Erens, 2011). The present study is not without its limitations. Firstly, the study did not include usual concurrent validity that generally tests the links between motives and problematic behaviors (e.g., multiple regression analyses) because the problem gambling prevalence rates in the BGPS were very low (0.9%). Outcomes for problem gambling and gamblingrelated problems should be included in future studies in order to gain a more comprehensive analysis of the concurrent validity. Secondly, the ‘coping’ dimension with only two items shows low Cronbach's alpha. Adding more items to the scale could be recommended for future research. Thirdly, as with all survey data, the data collected were based on self-report and subject to attendant problems of this mode of collection (e.g., social desirability biases, memory recall biases, etc.). Fourthly, although the overall sample size was large, the base sizes for some of the past-year gambling subgroups were small. Therefore, the subgroups presented in the present study were not definitive. It is plausible that further subgroups exist and other analytical techniques could perhaps be used to examine this more thoroughly. 5. Conclusion Examining the findings as a whole, the present study is one of a few studies (i.e., Francis et al., 2014; Wardle, et al., 2011a; Wardle et al., 2011b) that have provided detailed evidence for the utility and usefulness of the RGQ. The results confirmed the overall five-dimensional structure of the RGQ, and across gender (males vs. females), and age (16–34 years, 35–55 years and 55+ years). Role of funding sources No external funding source for this study has been received. Contributors Author MDG designed the study and wrote the protocol. All authors conducted literature searches and provided summaries of previous research studies. Author NC conducted the statistical analysis. Author NC wrote the first draft of the manuscript and all authors contributed to and have approved the final manuscript. Conflict of interest Mark Griffiths has received funding for a number of research projects in the area of gambling education for youth, social responsibility in gambling and gambling treatment from the Responsibility in Gambling Trust, a charitable body who funds its research program based on donations from the gambling industry. The author also undertakes consultancy for various gaming companies in the area of social responsibility in gambling. All other authors declare that they have no conflicts of interest.
References American Psychiatric Association (APA) (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington DC: American Psychiatric Association.
Binde, P. (2009). Gambling motivation and involvement: A review of social science research. http://www.responsiblegambling.org/articles/binde-gambling-motivationreview.pdf (Accessed 23 December 2014) Byrne, B. M., & Stewart, S. M. (2006). The MACS approach to testing for multigroup invariance of a second-order structure: A walk through the process. Structural Equation Modeling, 13, 287–321. Chantal, Y., Vallerand, R. J., & Vallieres, E. F. (1994). Construction et validation l'echelle de motivation relative aux jeux de hasard et d'argent. On the development and validation of the Gambling Motivation Scale (GMS). Society and Leisure, 17, 189–212. Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9, 233–255. Dechant, K., & Ellery, M. (2011). The effect of including a monetary motive item on the gambling motives questionnaire in a sample of moderate gamblers. Journal of Gambling Studies, 27, 331–344. Dowling, N. A., Lorains, F. K., & Jackson, A. C. (2015). Are the profiles of past-year internet gamblers generalizable to regular internet gamblers? Computers in Human Behavior, 43, 118–128. Francis, K. L., Dowling, N. A., Jackson, A. C., Christensen, D. R., & Wardle, H. (2014). Gambling motives: Application of the Reasons for Gambling Questionnaire in an Australian population survey. Journal of Gambling Studies. http://dx.doi.org/10.1007/ s10899-014-9458-1 (Epub ahead of print). Gainsbury, S., Wood, R., Russell, A., Hing, N., & Blaszczynski, A. (2012). A digital revolution: Comparison of demographic profiles, attitudes and gambling behavior of Internet and non-Internet gamblers. Computers in Human Behavior, 28, 1388–1398. Gilson, K. M., Bryant, C., Bei, B., Komiti, A., Jackson, H., & Judd, F. (2013). Validation of the Drinking Motives Questionnaire (DMQ) in older adults. Addictive Behaviors, 38, 2196–2202. Griffiths, M. D. (2011). Adolescent gambling. In B. B. Brown, & M. Prinstein (Eds.), Encyclopedia of Adolescence. Vol. 3. (pp. 11–20). San Diego: Academic Press. Griffiths, M. D., Wardle, H., Orford, J., Sproston, K., & Erens, B. (2011). Internet gambling, health, smoking and alcohol use: Findings from the 2007 British Gambling Prevalence Survey. International Journal of Mental Health and Addiction, 9, 1–11. Hodgins, D. C. (2008). What we see depends mainly on what we look for (commentary). Addiction, 103, 1118–1119. Lloyd, J., Doll, H., Hawton, K., Dutton, W. H., Geddes, J. R., Goodwin, G. M., et al. (2010). How psychological symptoms relate to different motivations for gambling: An online study of internet gamblers. Biological Psychiatry, 68, 733–740. McCormack, A., & Griffiths, M. D. (2012). Motivating and inhibiting factors in online gambling behaviour: A grounded theory study. International Journal of Mental Health and Addiction, 10, 39–53. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Parke, A., & Griffiths, M. D. (2011). Effects on gambling behaviour of developments in information technology: A grounded theoretical framework. International Journal of Cyber Behaviour, Psychology and Learning, 1(4), 36–48. R Core Team (2013). R: A language and environment for statistical computing [Computer software manual]. Vienna, Austria Available from http://www.R-project.org/. Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. Steenkamp, J. E. M., & Baumgartner, H. (1998). Assessing measurement invariance in cross-national consumer research. Journal of Consumer Research, 25, 78–90. Stewart, S. H., & Zack, M. (2008). Development and psychometric evaluation of a three‐dimensional Gambling Motives Questionnaire. Addiction, 103, 1110–1117. Stewart, S. H., Zack, M., Collins, P., & Klein, R. M. (2008). Subtyping pathological gamblers on the basis of affective motivations for gambling: Relations to gambling problems, drinking problems, and affective motivations for drinking. Psychology of Addictive Behaviors, 22, 257–268. Wardle, H., Moody, A., Griffiths, M., Orford, J., & Volberg, R. (2011a). Defining the online gambler and patterns of behaviour integration: Evidence from the British Gambling Prevalence Survey 2010. International Gambling Studies, 11, 339–356. Wardle, H., Moody, A., Spence, S., Orford, J., Volberg, R., Jotangia, D., et al. (2011b). British gambling prevalence survey 2010. London: The Stationery Office.