Abstract

Introduction: Cognitive distortions are related to gambling frequency and gambling severity. Having a culturally sensitive measure to assess cognitive distortions will facilitate the early detection of people who might be at risk of developing problematic gambling behaviors. The Gamblers' Beliefs Questionnaire was translated into Spanish (GBQ-S) but no previous study explored the structure of the GBQ-S in a non-US sample with different levels of gambling involvement. Aim: The present study examined the factor structure of the GBQ-S in a community sample of gamblers from Argentina. It also analyzed the association between cognitive distortions and type of gambling activity and frequency of gambling behaviors and the predictive utility of the GBQ-S on gambling severity. Participants: 508 youth and adults completed the GBQ-S. Results: The CFA showed an overall acceptable fit to the data confirming the proposed two-factor model. Scores of the two GBQ sub-scales were positively and significantly correlated with scores on gambling severity. Cognitive distortions have a significant effect on gambling severity after controlling for frequency of engagement in gambling activities. Luck and perseverance, but not illusion of control, was positively related to gambling severity. Discussion: scores measured by the GBQ-S exhibit adequate psychometric properties for the accurate assessment of cognitive distortions across adults and youth from the general community of Argentina.

Keywords

Gamblers' Beliefs Questionnaire;Confirmatory factor analysis;Criterion validity;Gambling

1. Introduction

Gambling is a prevalent recreational activity in many countries and across diverse cultures (Clark, 2010; Frascella et al., 2010 ;  Ledgerwood et al., 2009). The large majority of people who gamble do not experience adverse consequences, however, a subset of players experience gambling-related problems and develop severe forms of gambling disorder (French et al., 2008 ;  Korman et al., 2006). Gambling opportunities are rapidly expanding in Argentina. For instance, in the state of Cordoba, where the present study was conducted, there are 19 casinos and 3600 slots (National Lottery of Cordoba, 2016). Unfortunately, there is a scarcity of studies assessing prevalence of recreational, problem or disordered gambling in this country. Some evidence, however, suggest that 60% of Argentinean college students gambled at least once in their lifetime while between 6 and 12% meet the cut score criteria for problem gambling based on SOGS' scores (Tuzinkievich et al., 2013a ;  Tuzinkievich et al., 2013b). These rates of gambling engagement are similar to those found in United States (Huang, Jacobs, Derevensky, Gupta, & Paskus, 2007) and Canada (Huang and Boyer, 2007 ;  Ladouceur et al., 1994) employing the DSM IV and SOGS criteria, respectively. The aims of the current study were to examine the behavior of Argentinian gamblers and enrich the psychometric validity of the Spanish Version of the Gamblers' Beliefs Questionnaire (GBQ-S).

Cognitive perspectives on gambling emphasize the presence of a number of cognitive distortions, including overestimation of personal ability to influence a win, superstitious beliefs and lucky rituals assumed to increase the chance of winning, and a misunderstanding of random sequences and the independence of turns (Fortune & Goodie, 2012). Cognitive distortions are prevalent among gamblers across the severity spectrum, from social gamblers to disordered gamblers, and even non-gamblers (e.g., Ladouceur, 2004 ;  Winfree et al., 2013). Strong confidence in these beliefs is presumed to maintain problematic gambling behavior despite repeated negative outcomes.

Two prevalent cognitive distortions reported in the literature are illusion of control and illusory correlation (Fortune & Goodie, 2012). Illusion of control underlies the belief that gambling is a game of skill, instead of a game of chance, leading the gamblers to believe their skills are determinant to win the game. In other words, illusion of control refers to “an expectancy of a personal success probability that is higher than the objective probability should warrant” (Fortune & Goodie, 2012, pp. 301). Illusory correlation is the process of relating two events, based on experience or perceptions, even when there is no association between them (Fortune & Goodie, 2012). This cognitive distortion underlies the belief that luck plays an important role in gambling outcomes and different superstitions in relation to gambling (Fortune & Goodie, 2012). Illusory correlation is the core element of the association between particular habits, thoughts and superstitions with winning (Fortune & Goddie, 2012).

A greater level of these cognitive distortions is related to greater severity of gambling, not only among clinical disordered gamblers (Michalczuk et al., 2011 ;  Winfree et al., 2015), but also among non-clinical adolescents (Donati et al., 2015 ;  Taylor et al., 2014) and college students (MacKillop et al., 2006 ;  Winfree et al., 2013). The illusion of control has become an important phenomenon in experimental studies (Stefan & David, 2013, meta-analysis), however, its relevance to explain gambling behavior and disordered gambling is limited by the fact that some forms of gambling involve a genuine component of skill (e.g. poker: Meyer et al. 2013). Traditionally, two different categories have been used to classify the broad range of gambling activities: chance based games versus skill based games (Myrseth et al., 2010) or strategic versus non-strategic games (Grant et al., 2012). Skill or strategic games include gambling activities where some level of knowledge or skills in the game (e.g. poker, sports betting) can influence the outcome, at least potentially. In games of chance or non-strategic games, the gambler has no control or influence on the gambling outcomes (e.g. slots machines, lottery, and bingo). There is evidence that cognitive distortions are differentially related to different game preferences and, for example, skill gamblers, compared to chance gamblers, showed higher scores on the illusion of control scale of the Gamblers' Beliefs Questionnaire (Myrseth et al., 2010).

The valid assessment of gambling-related cognitive distortions can support evidence-based treatments targeting these distorted cognitions. The Gamblers' Beliefs Questionnaire (GBQ; Steenbergh et al., 2002) was developed to assist in case conceptualization, treatment planning, cognitive restructuring, relapse prevention, and monitoring behavior change (Whelan, Steenbergh, & Meyers, 2007). Factor analytic data from a community sample supported a two-factor structure: Illusion of Control and Luck/Perseverance (Steenbergh et al., 2002). The Illusion of Control factor was comprised of items that shared a theme of overestimating the influence of ones skill orientation on the outcome of chance-determined games. The Luck/Perseverance factor was comprised of items that share a common theme of an overestimation of chances of winning, including beliefs that one is prone to good fortune (i.e., illusory correlation).

The GBQ showed adequate psychometric properties regarding internal structure, confiability and concurrent validity (Steenbergh et al., 2002). Recently, the GBQ was examined in a treatment-seeking sample of disordered gamblers (Winfree et al., 2015). As expected, GBQ scores significantly decreased following a brief cognitive-behavioral intervention for disordered gambling. Results indicated that the original two-factor model provided a better fit than a single-factor model. However, the original two-factor model did not provide an overall adequate fit to the data. This finding is likely a result of the treatment-seeking sample endorsing more overall distortions and endorsing them more strongly (Winfree et al., 2015).

A Spanish version of the GBQ has also shown promise. Winfree, Meyers, & Whelan (2013) evaluated a Spanish adaptation of the measure (GBQ-S) in a U.S. based adult Latino sample and found adequate psychometric evidence and replicated the factor structure of the original English version. These results, however, do not guarantee the adequacy and appropriateness of this version to assess cognitive distortions in other Spanish speaking samples. The International Test Commission (ITC) asks for a comprehensive examination of the potential linguistic and cultural differences among the population for whom the versions of an instrument are intended (ITC, 2010). To date, the measure has not been evaluated in a Spanish-speaking sample outside US and, therefore, further validation of this measure in other Spanish-speaking samples from other parts of the world is an important next step.

Most psychological constructs are highly dependent on the cultural aspects where the tests are used, therefore, bias in test construction and test adaptation have a profound and deleterious impact on decisions regarding treatment and intervention (ITC, 2010). Argentina is a South American country with a large prevalence of European descent, and this European immigration has influenced Argentinean culture. The socio-cultural background is, therefore, quite dissimilar to that found among Hispanic population in U.S., mostly characterized by Mexican ascendant (US Department of Commerce Economics and Statistics Administration, 2012). It is reasonable; therefore, not to assume in advance that psychometric tests will behave similarly across diverse cultural groups, despite they speak the same language. Additionally, there is a need to study psychological variables in more diverse cultural groups (Henrich et al., 2010).

Evaluating the GBQ-S in an Argentinian sample would provide valuable information regarding gambling behavior and cognitive distortion endorsement among the Argentinian community. Having a culturally sensitive measure to assess cognitive distortions will facilitate the early detection of people who might be at risk of developing problematic gambling behaviors. The present study examined the factor structure of the GBQ-S in a sample of gamblers from Argentina with different levels of gambling involvement. Additionally, we analyzed the association between cognitive distortions and type of gambling activity and frequency of gambling behaviors. Finally, we examined the predictive validity of the GBQ-S by assessing its association with level of gambling severity.

2. Methods

2.1. Sample

The studys sample comprised 616 (209 men, 407 women) participants. A majority of the participants (n = 339, 54.9%) completed a paper-and-pencil survey while 277 (45.1%) participants completed an online survey. Participants who completed the paper-and-pencil survey were recruited from psychology (40.4%), engineering (24.9%) and biology (7.1%) courses at the National University of Cordoba (Argentina). We approached all department chairs via email or phone; yet only those from these three departments accepted being part of the study, and, as such, only students enrolled in these classes were included in the study. Twenty-seven cases were eliminated due to missing > 20% of data. Participants who completed the online survey were recruited through advertisements on social network sites (i.e. Facebook and Twitter), and e-mailing lists. This advertisement asked for young adults and adults (ages between 18 and 60 years) from the general community who gambled within the previous twelve months. Participants did not receive any monetary compensation for their participation. Only participants who reported lifetime gambling were retained for the study. Eighty-one cases from the paper-and-pencil survey reported no lifetime prevalence of gambling and were not included in the study. The final sample was composed of 508 participants (65.6% females). The mean age of 25.50 years (SD = 9.46 years) was statistically similar across males and females. Of the total sample, 82.5% were between 18 and 30 years old and 17.5% between 31 and 60. The majority of the sample (70.9%) reported living in the state of Cordoba, however; only 46.5% indicated they were born in that state. The remaining participants were from (53.5%) and lived in (29.1%) other states of Argentina. None of the participants indicated a different nationality. Table 1 shows sex, age and lifetime gambling engagement as a function of method of data collection.

Table 1. Demographic characteristics and occurrence of gambling for each sub-sample (online and paper-and-pencil surveys).
Paper-and-pencil Online
Sex (%)
 Men 40.1 29.6
 Women 59.5 70.4
Age (M/SD) 20.9/4.5 29.4/10.7
 College student 100 52.7
 Workera 21.1 44.4
Lifetime gambling (%)
 Casino 45.7 84.1
 Lottery 58.6 88.4
 Online gambling 2.6 67.9
 Gambling with others 75.4 89.5
S-PGSI (%)b
 NPG (0) 47.8
 LLP (1–2) 34.8
 MLP (3–7) 13.0
 PG (≥ 8) 4.4

Data are presented as percentage of subjects that fell into each category. For continuous variables data are presented as means and standard deviation in each category.

a. Percentages do not sum to 100% because being a student or a worker are not mutually exclusive.

b. NPG: non-problem gambling; LLP: low level of problems with few or no identified negative consequences; MLP: moderate level of problems leading to some negative consequences; PG: problem gambling with negative consequences and a possible loss of control.

3. Measures

3.1. Spanish version of the Gamblers' Beliefs Questionnaire (GBQ-S, Winfree et al., 2013)

This is a 20-item self-report instrument designed to assess gambling related cognitive distortions (Steenbergh et al., 2002). Each item of the GBQ consists of a statement that represents a cognitive distortion commonly held by gamblers. Respondents rate their level of agreement with each statement on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). GBQ has two sub-scales: Luck/Perseverance (12 items) and Illusion of Control (8 items). Higher scores are indicative of higher levels of cognitive distortion. The Spanish-GBQ has demonstrated adequate internal consistency (from α = 0.87 to α = 0.97; Winfree et al., 2013). Both surveys (internet and paper-and-pencil) included the Spanish-Gamblers' Beliefs Questionnaire (see the instruments section).

3.2. Gambling involvement

Gambling was defined as the act of placing an amount of money on the uncertain prospect of a larger monetary outcome. One question assessed lifetime occurrence of different gambling activities (roulette, sports, blackjack, poker, bingo, lottery, scratch cards, dice, and slot machines) in different gambling settings (casino, internet, in groups with other people). Additionally, we developed an ad hoc questionnaire to assess frequency of gambling behaviors. Participants reported last year frequency (i.e. almost daily, 1–3 times a week, 1–3 times a month, 10–11 times per year, 4–8 times per year, 2–3 times per year, once a year, never in the previous year) of gambling in each of the mentioned gambling activities. This set of questions was similar to those used in previous studies (Dechant, 2013; Ginley et al., 2014 ;  MacLaren et al., 2012). Only participants who completed the online survey answered questions regarding frequency of gambling involvement. For the regression analyses, gambling activities were grouped based on their gambling setting (i.e. casino, internet) and their category (i.e. chance and skill). Therefore, the following categories of gambling activities were developed: chance based games at a casino, skill based games at a casino, chance based games on internet, skill based games on internet, chance based games with other people, skill based games with other groups, and lottery games [bingo, scratch].

3.3. Spanish-Problem Gambling Severity Index (S-PGSI)

In a previous study (Pilatti & Tuzinkievich, 2015) we translated the nine items of the PGSI from English into Spanish. Participants indicated how often they had experienced each of the nine items within the previous year. Response choices range from 0 (never) to 3 (almost always) and, therefore, total scores range from 0 to 27. Only participants who completed the online survey answered this measure. Scores on S-PGSI are interpreted as followed: 0 = non-problem gambling; 1–2 = low level of problems with few or no identified negative consequences; 3–7 = moderate level of problems leading to some negative consequences; 8 or more = problem gambling with negative consequences and a possible loss of control (Hodgins, Stea, & Grant, 2011). The S-PGSI showed adequate internal consistency in a previous study (α = 0.88) and predictive validity in terms of gambling motives (Pilatti & Tuzinkievich, 2015). The present study yielded a Cronbachs alpha of 0.87 for S-PGSI scores.

4. Procedure

4.1. Paper and pencil survey

The principal researcher and three advanced students in Psychology administered the survey collectively in a classroom. They explained how to answer the questions of the anonymous survey. They remained throughout the session until all participants had completed the survey. Researchers encouraged the participants to complete the whole questionnaire. They emphasized the voluntary nature of participation and guaranteed confidentiality of the participants. Participants provided verbal informed consent before scale administration. The administration of the survey took approximately 30 min. The collection of data took four months within the first half of 2013.

4.2. Online survey

We created a computerized questionnaire, developed with the LimeSurvey software (license of the National University of Cordoba), to collect data. Participants responded to advertisements posted in social network sites (i.e., Facebook and Twitter) and sent by e-mail. The advertisement contained a link to the survey that included an online consent form. This online procedure minimizes the likelihood of missing data by prompting messages for any missed items. Questions did not ask for personally identifying information. The collection of data took five months within the end of 2013 and the beginning of 2014.

Both surveys (internet and paper-and-pencil) included the Spanish-Gamblers' Beliefs Questionnaire (see the instruments section). The online survey included the Spanish-PGSI and questions regarding frequency of gambling behaviors. The University Internal Review Board approved all study procedures and the National Agency for Promotion of Science and Technology (FONCyT) reviewed the protocol.

4.2.1. Data analysis

4.2.1.1. Descriptive analyses

We employed descriptive analysis to depict the occurrence of gambling behaviors and severity of gambling involvement. We conducted Cramers V and Students t- to explore differences between men and women in the frequency and severity of gambling, respectively.

4.2.1.2. Confirmatory and reliability analyses

First, we screened for outliers and missing data (Tabachnick & Fidell, 2011). Normality of the distribution of each item was assessed through inspection of asymmetry and kurtosis scores. Values between ± 2 in asymmetry and kurtosis, which are considered appropriate in the literature (George & Mallery, 2003). Univariate atypical cases were identified by calculating the standard z-score for each variable (z-scores > 3.29 were considered atypical) (Kline, 2011). These analyses were performed separately for the total sample and then for the sub-samples who completed the online and the paper-and-pencil forms. Subsequently, we conducted a confirmatory factor analysis (CFA) with Mplus to examine the underlying factor structure of the GBQ-S. The model proposed two inter-correlated factors in which items load on their individual subscale factor. We employed robust weighted least squares (robust WLS) to estimate the model. This method better suites factor analyses with ordinal indicators ( Flora & Curran, 2004). We employed multiple indices of goodness of fit to evaluate the fit of the model: the Tucker–Lewis index (TLI), the Comparative Fit Index (CFI), and the root mean square error of approximation (RMSEA) and Weighted Root Mean Square Residual (WRMR). CFI and TLI values equal or > 0.90, RMSEA values between 0.05 and 0.08, and WRMR values below 1.00 (Yu & Muthén, 2002) indicate an excellent or acceptable model fit. To study internal consistency, the composite reliability (i.e., the total amount of true score variance in relation to the total scale score variance) was estimated for each subscale of the instrument. Values equal or greater than 0.70 were considered as acceptable ( Nunnally, 1978). Scores of each of the subscales were calculated by the sum of the items that are defined a priori by the instruments (GBQ-S and S-PGSI). To determine if data from both methods of data collection (online and paper-and-pencil) show similar structures to the proposed model, we ran CFA separately for each sub-sample of participants.

4.2.1.3. Evidence of validity

We conducted a set of analysis to provide evidence for the validity of the measure. First, we explored the correlation between scores of each of the GBQ subscales and the level of severity. Next, we screened the variables (total scores on each sub-scale) to detect multivariate outliers (Mahalanobis distance measure; p < 0.001) (Kline, 2011). We also examined the distribution of the variables through inspection of asymmetry and kurtosis scores. Finally, we conducted a hierarchical regression analysis (enter method) to determine the capacity of each scale of gambling distortions (Luck and Perseverance and Illusion of Control) to predict level of gambling severity (criterion variable). We controlled the effect of gambling frequency. These analyses were performed separately for the samples of men and women.

5. Results

5.1. Descriptive results

Table 2 shows frequency of gambling engagement in a broad range of gambling activities and its association with sex. It also displays mean scores on the S-PGSI as a function of sex.

Table 2. Frequency of gambling engagement and its association with sex.
Never in the last year Once a year 2–3 times per year 4–8 times per year 10–12 times per year 1–3 times per month 1–3 times per week Almost every day Cramers V/t
Casino
 Cards 77.6 10.8 5.1 1.4 2.2 0.4 1.1 1.4 0.35⁎⁎⁎
 Roulette 66.4 17.3 10.1 2.2 1.1 1.8 0.4 0.7 0.20
 Bingo 71.5 14.8 9.4 2.2 0.4 1.1 0.0 0.7 0.25⁎⁎
 Slots 45.8 27.1 15.2 5.1 1.8 2.9 0.7 1.4 0.28⁎⁎⁎
 Dice 94.2 2.9 1.4 0.7 0.0 0.7 0.0 0.0 0.19
 Sports 96.4 2.5 0.7 0.0 0.0 0.0 0.0 0.4 0.19
Lottery
 Lottery tickets 50.2 28.2 11.6 4.3 2.2 2.9 0.4 0.4 0.21
 Quiniela 39.0 24.9 17.7 5.4 3.2 4.0 2.5 3.2 0.23
 Scrapies 55.6 20.9 13.4 4.7 1.4 2.5 0.7 0.7 0.13
 Quini 6 44.4 18.1 11.9 7.9 5.8 5.8 5.8 0.4 0.19
 Bingo 62.1 19.5 7.6 6.1 1.4 2.2 0.4 0.7 0.23
Internet
 Pocker 72.6 4.3 5.8 2.9 4.0 5.1 1.1 4.3 0.36⁎⁎⁎
 Slots 88.8 4.3 1.8 2.2 0.7 0.4 0.4 1.4 0.13
 Table games 69.7 9.0 7.2 3.2 4.3 3.2 0.7 2.5 0.08
 Roulette 96.0 1.8 0.7 0.4 0.4 0.4 0.0 0.4 0.20
 Cards 90.6 3.6 1.4 2.2 0.7 1.1 0.0 0.4 0.17
 Sports 94.9 1.1 2.2 0.0 0.7 0.4 0.0 0.7 0.13
 Dice 93.9 2.2 1.8 0.7 0.4 0.7 0.0 0.4 0.23
 Bingo 98.6 0.4 0.4 0.0 0.4 0.0 0.0 0.4 0.14
In groups with others
 Pocker 61.7 11.9 8.7 4.7 4.0 5.4 1.4 2.2 0.33⁎⁎⁎
 Sports 77.3 9.7 7.6 2.2 2.2 0.7 0.0 0.4 0.20
 Table games 55.6 15.2 15.9 6.1 2.2 4.0 0.4 0.7 0.16
 Video games 74.7 8.7 7.2 2.9 2.2 2.9 0.4 1.1 0.22
 Roulette 89.9 6.9 1.8 0.0 0.7 0.4 0.0 0.4 0.20
 Pool 70.8 14.4 11.2 1.8 1.1 0.4 0.0 0.4 0.14
 Cards 49.5 14.1 18.4 7.6 4.0 4.7 0.7 1.1 0.20
 Dice 66.1 14.1 9.7 4.3 1.4 3.6 0.0 0.7 0.18
 S-PGSI (mean/SD)
 Male 2.7(4.5) 4.1⁎⁎⁎
 Female 1.1(1.9)

Note: Quiniela and Quini 6 are two common gambling activities in Argentina.

⁎. p ≤ 0.05.

⁎⁎. p ≤ 0.01.

⁎⁎⁎. p ≤ 0.001.

5.2. Confirmatory factor analysis and composite reliability

Because < 5% of the data were missing, missing values were imputed by mode substitution (Schafer, 1999). This imputation method imputes values within the seven response options (discrete values such as 1, 2, etc.) of the self-response scale, whereas other methods (e.g. estimation by maximization) tend to introduce continuous variables (e.g., 2.4, 3.8), changing the original distribution (Dominguez Lara, 2014). We then calculated asymmetry and kurtosis indices to examine the univariate distribution of the items assessing frequency of gambling, severity of gambling and cognitive distortions (Tabachnick & Fidell, 2011). Seventeen variables showed values greater than ± 2.0 which are considered unacceptable (George & Mallery, 2011). Non-normal distribution may affect the association between the predictor and the criterion variables. Therefore, we transformed these variables. Specifically, we first applied a rank transformation and then we obtained z scores (McDonald, 1999). These transformed values were used in the rest of the analysis.

CFA results confirmed the two-factor model. Specifically, CFA results showed an overall acceptable fit to the data for the total sample (CFI 0.93; TLI 0.92, RMSEA 0.085, WRMR 1.387) and for both, the online (CFI 0.92; TLI 0.90, RMSEA 0.090, WRMR 1.241) and the paper-and-pencil (CFI 0.96; TLI 0.95, RMSEA 0.081, WRMR 1.011) sub-samples. The factor loadings weight (p ≤ 0.05) for the Illusion of Control factor varied between 0.45 and 0.85, 0.52 and 0.90, 0.35 and 0.79 for the total, online and paper-and-pencil samples, respectively. The standardized regression weights for Luck/Perseverance varied between 0.67 and 0.83, 0.76 and 0.82, 0.59 and 0.79 for the total, online and paper-and-pencil samples. Composite reliability for Illusion of Control was ρ = 0.85, ρ = 0.80 and ρ = 0.89 for the total, online and paper-and-pencil samples, respectively. For Luck/Perseverance, the observed composite reliability was ρ = 0.94, ρ = 0.92, and ρ = 0.96 for each sample. Table 3 shows standardized factorial weights for each item on its corresponding factor, the composite reliability for each subscale and indices of goodness of fit for the total sample and separately as a function of method of data collection.

Table 3. Standardized regressions weights and composite reliability for each GBQ item. Results are presented for the total sample and for each sub-sample (online and paper-and-pencil surveys).
Total Paper-and-pencil Online
Factor loadings weight
IC1 0.49 0.54 0.43
IC2 0.45 0.52 0.35
IC3 0.71 0.70 0.71
IC4 0.64 0.76 0.50
IC6 0.74 0.81 0.66
IC8 0.49 0.55 0.44
IC9 0.85 0.89 0.79
IC10 0.79 0.90 0.66
CR 0.85 0.89 0.80
LyP5 0.73 0.77 0.68
LyP7 0.75 0.83 0.65
LyP11 0.75 0.83 0.68
LyP12 0.67 0.76 0.59
LyP13 0.83 0.85 0.83
LyP14 0.76 0.76 0.78
LyP15 0.76 0.81 0.73
LyP16 0.79 0.84 0.73
LyP17 0.70 0.78 0.63
LyP18 0.83 0.87 0.79
LyP19 0.80 0.83 0.78
LyP20 0.71 0.82 0.60
CR 0.94 0.96 0.92
CFI 0.93 0.96 0.92
TLI 0.92 0.95 0.90
RMSEA 0.09 0.08 0.09
WRMR 1.39 1.01 1.24

CFI: Comparative Fit Index; TLI: Tucker–Lewis index; RMSEA: root mean square error of approximation; WRMR: Weighted Root Mean Square Residual. CR: composite reliability.

5.3. Criterion-related validity

Scores of each GBQ-S sub-scales were positively and significantly correlated with scores on the S-PGSI. Specifically, a higher level of cognitive distortion measured by the Illusion of Control (r = 0.40, p ≤ 0.001) and the Luck/Perseverance (r = 0.45, p ≤ 0.001) scales was related to a greater level of gambling severity as measured by the S-PGSI. Next, we conducted the regression analysis. Based on the observed differences in gambling involvement and level of severity, we decided to conduct the regression analysis separately for the subsample of men and the subsample of women. We checked for multivariate assumptions of independence between residues, homoscedasticity, linearity, normality, and multicollinearity (Tabachnick & Fidell, 2011). Table 4 shows the main results of this analysis. In the first step of the regression, we included frequency of gambling on each gambling activity (chance and skill games at a casino, chance and skill games on internet, chance and skill games with other people and lottery games). In the second step, we entered the two scales of cognitive distortions (Luck and Perseverance and Illusion of Control). We conducted this analysis for women and men in two separate analyses.

Table 4. Predictive utility of gambling involvement and cognitive distortions on gambling severity. Results are presented separately for the sub-sample of women and the sub-sample of men.
Men Women
Step 1 Step 2 Step 1 Step 2
β β β β
Casino chance games 0.41⁎⁎⁎ 0.38⁎⁎⁎ 0.20⁎⁎ 0.21⁎⁎
Casino skill games 0.30 0.13 0.15⁎⁎ 0.12
Lottery tickets 0.03 0.01 − 0.02 − 0.09
On-line skill games 0.10 − 0.01 0.04 − 0.01
On-line chance games 0.11 0.13 − 0.14 − 0.11
Group skill games − 0.11 − 0.09 − 0.02 0.03⁎⁎
Group chance games − 0.09 − 0.09 0.33⁎⁎⁎ 0.20
Luck and perseverance 0.30⁎⁎ 0.39⁎⁎⁎
Control illusion 0.16 − 0.01
ΔR2 0.12⁎⁎⁎ 0.12⁎⁎⁎
R2 0.43⁎⁎⁎ 0.55⁎⁎⁎ 0.18⁎⁎⁎ 0.31⁎⁎⁎
R2 adjusted 0.37⁎⁎⁎ 0.49⁎⁎⁎ 0.15⁎⁎⁎ 0.27⁎⁎⁎

β = standardized regression coefficient; ΔR2 = R Square change; R2 = R Square; Adjusted R Square.

⁎⁎⁎. p < 0.001.

⁎⁎. p < 0.01.

⁎. p < 0.05.

5.4. Subsample of men

In the first step, gambling frequency explained 43% of the variance of the severity of gambling F (7, 61) = 7.60. Casino chance games (β = 0.41, t = 3.74, p < 0.001) and casino skill games (β = 0.30, t = 2.52, p < 0.05) were positively associated to severity of gambling. In the second step, the percentage of explained variance increased 12% [F (2, 69) = 9.36, p < 0.001] after entering Luck and Perseverance and Illusion of Control scales. The scale measuring Luck and Perseverance (β = 0.30, t = 2.80, p < 0.01) was positively associated to severity of gambling. Additionally, the effect of frequency of casino skill gambling was no longer significant, indicating this effect was completely mediated by cognitive distortions.

5.5. Subsample of women

Gambling frequency for each gambling activity explained 18% of the variability of the severity of gambling F (7, 183) = 5.87. Casino chance games (β = 0.20, t = 2.77, p < 0.01), casino skill games (β = 0.15, t = 2.10, p < 0.05) and group skill games (β = 0.33, t = 3.78, p < 0.001) had positive effects on scores of indicating severity of gambling. In the second step, the percentage of explained variance increased to 31% [F (2, 181) = 16.11, p < 0.001] after entering the two scales of cognitive distortions (Luck and Perseverance and Illusion of Control). The scale Luck and Perseverance (β = 0.39, t = 4.38, p < 0.001) was positively associated to the criterion variable. Also, we observed that the effects of frequency of casino skill games and group skill games were completely and partially mediated, respectively, by these cognitions. Specifically, in the second step the effect of frequency of casino skill gambling was no longer significant while the effect of frequency of gambling in groups descended to 0.20.

6. Discussion

The present study examined the psychometric properties of the Spanish version of the Gamblers' Beliefs Questionnaire in an Argentinean sample of youth and adults. This Spanish version, adapted from the original English version, has shown adequate psychometric properties with Hispanic gamblers living in the US (Winfree et al., 2013). Until now, however, no previous study evaluated the psychometric properties of the GBQ-S in Spanish-speaking sample outside U.S. Gambling is a prevalent activity among Argentinean youth and there is a need to develop culturally appropriate measures for the assessment of gambling behaviors and its risk factors in this population. In the present study, we seek to provide evidence of the validity of this measure for its use among Argentinean youth and adults. We addressed different sources of validity to provide further evidence of the adequacy of this Spanish version.

Consistent with previous studies (Winfree et al., 2013), results supported the Spanish GBQ as a valid tool to assess cognitive distortions in Spanish-speaking youth and adults from Argentina. Specifically, we confirmed and sustained the original two factor model proposed by Winfree et al. (2013). Importantly, and similarly to prior research (Winfree et al., 2013), both sub-scales showed high values of internal consistency. According to our results, cognitive distortions were similarly distributed regardless of the age of the respondents but men were more likely than women to hold more distorted beliefs about gambling (Raylu & Oei, 2004) and to display higher levels of severity (King et al., 2010). These results added more evidence regarding the validity of this measure.

Previous work supported the association between cognitive distortions and gambling frequency (Donati et al., 2015 ;  Raylu and Oei, 2004) and between cognitive distortions and gambling-related problems (Donati et al., 2015; Raylu and Oei, 2004; Steenbergh et al., 2002 ;  Winfree et al., 2013). Considering the strong relationship between gambling frequency and gambling severity (Bertossa et al., 2014; Holtgraves, 2009 ;  Rahman et al., 2012) a particular goal was to examine the impact of specific types of cognitive distortions on gambling severity while controlling the effect of gambling frequency. To our knowledge, this is the first study to analyze this.

Our results indicate that cognitive distortions explained a percentage of the variance of gambling severity over the effect of frequency of engagement in a broad range of gambling activities. Notably, gambling beliefs regarding luck and perseverance, but not those depicting illusion of control, were positively related to gambling severity, suggesting the former beliefs had an independent and unique impact on gambling related problems regardless of the frequency of gambling involvement. Similarly, Donati et al. (2015) did not find a significant association between GRCSs sub-scales assessing illusion of control and gambling severity, emphasizing the role of these specific distorted cognitions. Accordingly, interventions aimed at reducing problematic gambling behaviors would benefit from challenging these particular beliefs. Indeed, one of the central features of Cognitive-Behavioral Therapy (CBT) for problem gambling is challenging erroneous beliefs about gambling (to understand and identify erroneous thinking about his/her chances of winning) (Okuda et al., 2009).

Despite the evidence regarding the sustainability of the Spanish GBQ to assess cognitive distortions in Spanish-speaking youth and adults from Argentina, there are some limitations that need to be considered when interpreting these results. First, we employed two different methods for data collection, web-based survey and paper and pencil questionnaires. Previous work, however, indicates that quality of information is similar across both methods (Hohwü et al., 2013 ;  Touvier et al., 2010). Additionally, our sample had a higher percentage of women than men and most of the participants were 31 years old or younger. Future studies might explore our results in a more diverse sample regarding not only sociodemographic characteristics but also at the level of involvement in gambling activities.

Despite these limitations this is the first study that examines the psychometric properties of the Spanish version of the GBQ in a sample of Spanish-speaking youth and adults outside of the United States. Our results suggest that scores measured by this tool exhibit adequate psychometric properties (based on the CTT) for the accurate assessment of cognitive distortions across youth and adults from the general community.

References

  1. Bertossa et al., 2014 S. Bertossa, P. Harvey, D. Smith, A. Chong; A preliminary adaptation of the Problem Gambling Severity Index for Indigenous Australians: Internal reliability and construct validity; Australian and New Zealand Journal of Public Health, 38 (4) (2014), pp. 349–354
  2. Clark, 2010 L. Clark; Decision-making during gambling: An integration of cognitive and psychobiological approaches; Philosophical Transactions of the Royal Society of London B: Biological Sciences, 365 (1538) (2010), pp. 319–330
  3. Dechant, 2013 K. Dechant; Show me the money: Incorporating financial motives into the Gambling Motives Questionnaire; Journal of Gambling Studies, 30 (4) (2013), pp. 949–965
  4. Donati et al., 2015 M.A. Donati, F. Ancona, F. Chiesi, C. Primi; Psychometric properties of the Gambling Related Cognitions Scale (GRCS) in young Italian gamblers; Addictive Behaviors, 45 (2015), pp. 1–7
  5. Flora and Curran, 2004 D.B. Flora, P.J. Curran; An empirical evaluation of alternative methods of estimation for confirmatory factor analysis with ordinal data; Psychological Methods, 9 (2004), pp. 466–491
  6. Fortune and Goodie, 2012 E.E. Fortune, A.S. Goodie; Cognitive distortions as a component and treatment focus of pathological gambling: A review; Psychology of Addictive Behaviors, 26 (2) (2012), p. 298
  7. Frascella et al., 2010 J. Frascella, M.N. Potenza, L.L. Brown, A.R. Childress; Shared brain vulnerabilities open the way for nonsubstance addictions: Carving addiction at a new joint?; Annals of the New York Academy of Sciences, 1187 (1) (2010), pp. 294–315
  8. French et al., 2008 M.T. French, J.C. Maclean, S.L. Ettner; Drinkers and bettors: Investigating the complementarity of alcohol consumption and problem gambling; Drug and Alcohol Dependence, 96 (1) (2008), pp. 155–164
  9. George and Mallery, 2003 D. George, P. Mallery; SPSS for Windows step by step: A simple guide and reference. 1.0 update; (4th ed)Allyn and Bacon, Boston (2003)
  10. George and Mallery, 2011 D. George, M. Mallery; SPSS for Windows step by step: a simple guide and reference; Allyn & Bacon, Boston, MA (2011)
  11. Ginley et al., 2014 M.K. Ginley, J.P. Whelan, A.W. Meyers, G.E. Relyea, G.D. Pearlson; Exploring a multidimensional approach to impulsivity in predicting college student gambling; Journal of Gambling Studies, 30 (2) (2014), pp. 521–536
  12. Grant et al., 2012 J.E. Grant, B.L. Odlaug, S.R. Chamberlain, L.R. Schreiber; Neurocognitive dysfunction in strategic and non-strategic gamblers; Progress in Neuro-Psychopharmacology and Biological Psychiatry, 38 (2) (2012), pp. 336–340
  13. Henrich et al., 2010 J. Henrich, S.J. Heine, A. Norenzayan; Most people are not WEIRD; Nature, 466 (2010), p. 29
  14. Hodgins et al., 2011 D.C. Hodgins, J.N. Stea, J.E. Grant; Gambling disorders; Lancet, 378 (9806) (2011), pp. 1874–1884 http://doi.org/10.1016/S0140-6736(10)62185-X
  15. Hohwü et al., 2013 L. Hohwü, H. Lyshol, M. Gissler, S.H. Jonsson, M. Petzold, C. Obel; Web-based versus traditional paper questionnaires: A mixed-mode survey with a Nordic perspective; Journal of Medical Internet Research, 15 (8) (2013), Article e173
  16. Holtgraves, 2009 T. Holtgraves; Gambling, gambling activities, and problem gambling; Psychology of Addictive Behaviors, 23 (2) (2009), p. 295
  17. Huang and Boyer, 2007 J.H. Huang, R. Boyer; Epidemiology of youth gambling problems in Canada: A national prevalence study; The Canadian Journal of Psychiatry/La Revue Canadienne de Psychiatrie, 52 (2007), pp. 657–665
  18. Huang et al., 2007 J.H. Huang, D.F. Jacobs, J.L. Derevensky, R. Gupta, T.S. Paskus; Gambling and health risk behaviors among US college student-athletes: Findings from a national study; Journal of Adolescent Health, 40 (2007), pp. 390–397
  19. International Test Commission (2010), n.d International Test Commission (2010); International Test Commission Guidelines for translating and adapting tests; Retrieved from  http://www.intestcom.org April 20, 2015
  20. King et al., 2010 S.M. King, K. Abrams, T. Wilkinson; Personality, gender, and family history in the prediction of college gambling; Journal of Gambling Studies, 26 (3) (2010), pp. 347–359
  21. Kline, 2011 R.B. Kline; Principles and practice of structural equation modeling; (3rd ed.)Guilford, New York, NY (2011)
  22. Korman et al., 2006 L.M. Korman, T. Toneatto, W. Skinner; Pathological gambling; Practitioners guide to evidence-based psychotherapy, Springer US (2006), pp. 291–300
  23. Ladouceur, 2004 R. Ladouceur; Perceptions among pathological and nonpathological gamblers; Addictive Behaviors, 29 (3) (2004), pp. 555–565
  24. Ladouceur et al., 1994 R. Ladouceur, D. Dubé, A. Bujold; Prevalence of pathological gambling and related problems among college students in the Quebec metropolitan area; The Canadian Journal of Psychiatry/La Revue Canadienne de Psychiatrie, 39 (1994), pp. 289–293
  25. Lara, 2014 S.A.D. Lara; Análisis Psicométrico de la Escala de bienestar psicológico para Adultos en estudiantes universitarios de Lima: un enfoque de ecuaciones estructurales; Psychologia: avances de la disciplina, 8 (1) (2014), pp. 23–31
  26. Ledgerwood et al., 2009 D.M. Ledgerwood, S.M. Alessi, N. Phoenix, N.M. Petry; Behavioral assessment of impulsivity in pathological gamblers with and without substance use disorder histories versus healthy controls; Drug and Alcohol Dependence, 105 (1) (2009), pp. 89–96
  27. MacKillop et al., 2006 J. MacKillop, E.J. Anderson, B.A. Castelda, R.E. Mattson, P.J. Donovick; Convergent validity of measures of cognitive distortions, impulsivity, and time perspective with pathological gambling; Psychology of Addictive Behaviors, 20 (1) (2006), p. 75
  28. MacLaren et al., 2012 V.V. MacLaren, K.A. Harrigan, M. Dixon; Gambling motives and symptoms of problem gambling in frequent slots players; Journal of Gambling Issues (2012)
  29. McDonald, 1999 R.P. McDonald; Test theory: A unified approach; Lawrence Earlbaum, Mahwah, NJ (1999)
  30. Meyer et al., 2013 G. Meyer, M. von Meduna, T. Brosowski, T. Hayer; Is poker a game of skill or chance? A quasi-experimental study; Journal of Gambling Studies, 29 (3) (2013), pp. 535–550
  31. Michalczuk et al., 2011 R. Michalczuk, H. Bowden-Jones, A. Verdejo-Garcia, L. Clark; Impulsivity and cognitive distortions in pathological gamblers attending the UK National Problem Gambling Clinic: A preliminary report; Psychological Medicine, 41 (12) (2011), pp. 2625–2635
  32. Myrseth et al., 2010 H. Myrseth, G.S. Brunborg, M. Eidem; Differences in cognitive distortions between pathological and non-pathological gamblers with preferences for chance or skill games; Journal of Gambling Studies, 26 (4) (2010), pp. 561–569
  33. National Lottery of Cordoba, 2016 National Lottery of Cordoba; Statistics of gambling activities; (2016) Retrieved online on August, 24th 2016: https://www.loteriadecordoba.com.ar/
  34. Nunnally, 1978 J.C. Nunnally; Psychometric theory; McGraw-Hill Book Company (1978), pp. 86–113 190-255
  35. Okuda et al., 2009 M. Okuda, I. Balán, N.M. Petry, M. Oquendo, C. Blanco; Cognitive-behavioral therapy for pathological gambling: Cultural considerations; American Journal of Psychiatry (2009)
  36. Pilatti and Tuzinkievich, 2015 A. Pilatti, F.B. Tuzinkievich; Valoración de las propiedades psicométricas del Cuestionario de Motivos de Apuestas en jóvenes y adultos argentinos; Adicciones, 27 (1) (2015), pp. 17–26
  37. Rahman et al., 2012 A.S. Rahman, C.E. Pilver, R.A. Desai, M.A. Steinberg, L. Rugle, S. Krishnan-Sarin, M.N. Potenza; The relationship between age of gambling onset and adolescent problematic gambling severity; Journal of Psychiatric Research, 46 (5) (2012), pp. 675–683
  38. Raylu and Oei, 2004 N. Raylu, T.P. Oei; The Gambling Related Cognitions Scale (GRCS): Development, confirmatory factor validation and psychometric properties; Addiction, 99 (6) (2004), pp. 757–769
  39. Schafer, 1999 J.L. Schafer; Multiple imputation: A primer; Statistical Methods in Medical Research, 8 (1999), pp. 3–15
  40. Steenbergh et al., 2002 T.A. Steenbergh, A.W. Meyers, R.K. May, J.P. Whelan; Development and validation of the Gamblers' Beliefs Questionnaire; Psychology of Addictive Behaviors, 16 (2) (2002), p. 143
  41. Stefan and David, 2013 S. Stefan, D. David; Recent developments in the experimental investigation of the illusion of control. A meta-analytic review; Journal of Applied Social Psychology, 43 (2) (2013), pp. 377–386
  42. Tabachnick and Fidell, 2011 B.G. Tabachnick, L.S. Fidell; Multivariate analysis of variance (MANOVA); International encyclopedia of statistical science, Springer, Berlin Heidelberg (2011), pp. 902–904
  43. Taylor et al., 2014 R.N. Taylor, J.D. Parker, K.V. Keefer, P.H. Kloosterman, L.J. Summerfeldt; Are gambling related cognitions in adolescence multidimensional?: Factor structure of the gambling related cognitions scale; Journal of Gambling Studies, 30 (2) (2014), pp. 453–465
  44. Touvier et al., 2010 M. Touvier, C. Méjean, E. Kesse-Guyot, C. Pollet, A. Malon, K. Castetbon, S. Hercberg; Comparison between web-based and paper versions of a self-administered anthropometric questionnaire; European Journal of Epidemiology, 25 (5) (2010), pp. 287–296
  45. Tuzinkievich et al., 2013a Tuzinkievich, F., del Vera B. V., Caneto, F., Garimaldi, J., & Pilatti, A. (2013a). Poster: Ocurrencia de juegos de apuestas en adolescentes y su relación con distorsiones cognitivas y consumo de alcohol. Libro de Resúmenes VIII Congreso Latinoamericano Interdisciplinario del Adolescente (CLIOA). ISBN: 978-950-33-1078-6
  46. Tuzinkievich et al., 2013b F. Tuzinkievich, B.V. del Vera, F. Caneto, J. Garimaldi, A. Pilatti; Poster: Prevalencia de apuestas y su relación con distorsiones cognitivas e impulsividad en estudiantes universitarios; Revista Argentina de Ciencias del Comportamiento, Suplemento (Julio) (2013), pp. 368–369
  47. US Department of Commerce Economics and Statistics Administ US Department of Commerce Economics and Statistics Administration; US Census Bureau; (2012) Retrieved from https://www.census.gov/prod/cen2010/briefs/c2010br-04sp.pdf April 29, 2015
  48. Whelan et al., 2007 J.P. Whelan, T.A. Steenbergh, A.W. Meyers; Problem and Pathological Gambling; Hogrefe, Cambridge (2007)
  49. Winfree et al., 2013 W.R. Winfree, A.W. Meyers, J.P. Whelan; Validation of a Spanish translation of the Gamblers' Beliefs Questionnaire; Psychology of Addictive Behaviors, 27 (1) (2013), p. 274
  50. Winfree et al., 2015 W.R. Winfree, M.K. Ginley, J.P. Whelan, A.W. Meyers; Psychometric evaluation of the Gamblers' Beliefs Questionnaire with treatment-seeking disordered gamblers; Addictive Behaviors, 43 (2015), pp. 97–102
  51. Yu and Muthén, 2002 C.Y. Yu, B.O. Muthén; Evaluation of model fit indices for latent variable models with categorical and continuous outcomes (technical report); University of California at Los Angeles, Graduate School of Education & Information Studies, Los Angeles (2002)
Back to Top

Document information

Published on 26/05/17
Submitted on 26/05/17

Licence: Other

Document Score

0

Views 7
Recommendations 0

Share this document

claim authorship

Are you one of the authors of this document?