2.1 Consumer Personality. Big Five

The Trait Theory is the most influential school of thought in personality psychology [62]. Specifically, one of the most well-known and widely used personality models in psychology to measure the five previously mentioned personality traits is the Big Five model [20], [32], [60], [62], [81], [82]. Indeed, behaviors are better predicted using a common framework such as the Big Five [83]. Recently evidence [73] reveals that personality traits could be reliable sources to describe consumers’ behaviors.

Following [32], [38], [62], [67], [82], personality traits are the extent to which consumers appraise themselves regarding five dimensions: (1) conscientiousness assesses one’s degree of organization, persistence, and motivation in goal-directed behaviour; (2) neuroticism assesses an individual prone to psychological distress, unrealistic ideas, excessive cravings or urges, and maladaptive coping responses; (3) extraversion assesses an individual’s quantity and intensity of interpersonal interaction and activity level; (4) agreeableness assesses an individual’s quality of interpersonal orientation along a continuum from compassion to antagonism in thoughts, feelings, and actions; and (5) openness assesses an individual’s proactive seeking and appreciation of experience for its sake, toleration for, and exploration of the unfamiliar.

In the context of video games and online games, personality traits have been related to the fact of playing or not, the style of playing, addiction to video games, and the motivations and benefits derived from playing (see Table 1). However, the link between the Big Five traits and self-efficacy playing has yet to be studied.

Table 1: Research on video gamers and Big Five personality traits 

2.2 Self-Efficacy Role

The theory of self-efficacy lies at the heart of Bandura’s Social Cognitive Theory [10]. Self-efficacy refers to peoples' judgment of their ability to organize and execute courses of action required to attain designated types of performance [12], [35]. It is domain specific and strongly tied to the behavior of interest. Self-efficacy is concerned not only with the skills one has but also with the confidence with which one uses those skills [14], [35], [91].

As with self-confidence or self-control [64], self-efficacy is an individual resource, according to psychology literature. However, unlike other self-traits such as self-esteem, self-efficacy refers to individuals’ self-assessment of their capacity to accomplish a given task and achieve a particular goal [13], [58]. Self-efficacy is a broader engagement of inner resources that encompasses feelings, cognitions, and behaviors [91].

On this conceptual basis, we can identify different self-efficacy domains. In fact, self-efficacy has been studied in different fields of knowledge, such as student learning [59], [101], employee behaviour [17], [84], [99], consumer reactions [20], [77], [91] and video gaming [57]. Different technologies have also been addressed in self-efficacy literature: computer self-efficacy [43], [82] or Internet self-efficacy [43], [45], [52], [61]. However, research has seldom studied self-efficacy applied to the video game industry [57] or online shopping [43].

In this work, we consider gaming self-efficacy as individuals’ self-assessment of their capacity to solve the challenges posed in a game and to become skilled in a game app. In the context of video games, we have only found the study by [57], which specifically uses video gaming self-efficacy and computer gaming self-efficacy. Similarly, online shopping self-efficacy refers to individuals’ self-confidence when searching for and acquiring products and services in online stores [36]. Following several authors, self-efficacy is specifically related to different contexts and concepts, giving rise to various kinds of self-efficacy apart from general self-efficacy [29], [52], [57]. As an example, [43] distinguished between general Internet self-efficacy and specific www self-efficacy as different concepts.

The predictive capability of self-efficacy is strongest and most accurate when determined by specific domain-linked measures rather than by general measures [82]. In this line, specific self-efficacy is one of the factors that proves a key driver vis-à-vis understanding individual behavior and adoption of new technologies [43], [45], [52], [61], [82]. For instance, [64] emphasize the importance of possessing expertise and self-control in order to perform shopping tasks successfully. While players who value video game success as part of their identity may reap benefits from video game playing, those who do not value such success may experience more harmful effects [92].

2.3 Personality-Self-Efficacy Link

High self-efficacy is attributed to individuals’ beliefs that they can manage and solve the problems at hand more successfully [91]. However, self-efficacy can be influenced by personality. Individuals’ personality can affect their level of self-efficacy [13]. Nevertheless, few studies have related personality traits with consumer self-efficacy [52], [58], [61], [77], [91]. Some studies have explored personality traits as drivers of gamer behavior [23], [30], [49], [85], although to the best of our knowledge, none have thus far related personality to self-efficacy in a video gaming context.

As regards conscientiousness, [77], [82] found that it is positively associated with task fulfillment and job performance and that it influences self-efficacy. In the context of gaming, it can be assumed that game players must be persistent if they are to achieve game goals and be successful. In other words, persistence and motivation allow players to achieve self-efficacy and become skilled players.

As for neuroticism, [82] find that neuroticism negatively influences self-efficacy. Individuals’ psychological distress may limit their ability to become skilled and thus obtain the expected achievements in a game. Indeed, one of the harmful effects of playing video games is stress and anxiety [69].

As for extraversion, in gamer behavior [26] argue that social influence, which explains how other people's attitudes or opinions may affect an individual's decision making, is a determining factor in social games. Extraversion positively influences self-efficacy [82]. Interacting with other people, coupled with a tendency to be sociable, may increase their knowledge of the game and their abilities and skills when playing it, as suggested by [100].

Results concerning the effects of agreeableness on self-efficacy are mixed. While [62] finds a significant relationship between agreeableness and brand loyalty in the case of video games, [60] observe a negative effect of agreeableness and Internet usage, and [82] find that agreeableness negatively influences self-efficacy. These authors argue that there is no reason to propose a positive relationship between agreeableness and self-efficacy since the former is related to higher computer anxiety and less use of mobile phones. In [55], agreeableness was not significantly related to self-efficacy.

Finally, individuals displaying openness are more willing to try something new and also possess good imagination. An open mind is usually an advantage when solving problems at work and in life; consequently, open-minded individuals would also have a positive evaluation of their efficacy [77]. Openness increases loyalty to video games [62] and, in the same line, [66] show that openness relates to brand loyalty to products which have a hedonic value. Also, [82] find a positive influence of openness on self-efficacy. In the context of gaming, players’ openness involves searching for new experiences as well as imaginative and innovative games. Eventually, the search for new experiences might increase their abilities in the game.

Gamers are thus expected to display greater self-efficacy if they are organized and persistent (conscientious); not worried, nervous and emotionally stable (not neurotic); sociable, optimistic and talkative (extrovert); unforgiving, cold and skeptical (not agreeable) and curious, creative and imaginative (open). Following the previous reasoning, the hypotheses proposed reflect the effect of personality traits on gaming self-efficacy.

H1: Conscientiousness positively influences gaming self-efficacy.

H2: Neuroticism negatively influences gaming self-efficacy.

H3: Extraversion positively influences gaming self-efficacy.

H4: Agreeableness negatively influences gaming self-efficacy.

H5: Openness positively influences gaming self-efficacy.

2.4 Proposed Transfer between Gaming Self-Efficacy and Online Shopping Self-Efficacy

As [45] affirms, Internet has potentially and irrevocably impacted on the way business is conducted and has redefined the parameters of business in general. Indeed, the Internet has grown to become one of the leading sales channels today, with game apps being no exception [65]. Prior research provides evidence about the link between mobile gaming and online shopping behavior [11]. [93]’s study has found that richer online formats and recommendations throughout video games enhance consumers’ perceptions about experience products and positively affect consumers' intentions to purchase products. Furthermore, [44] found that app-based gaming has created complex models to generate purchase (in and out the game). Models that comprises explicit exchange of money for digital items within each game or economic systems based on virtual assets (e.g., Worldwide Asset eXchange and OPSkins market) and no explicit -nor direct- purchase related to the game (i.e., through branding freemium games or ads). Recent data [34], [89] suggest that video gamers conduct in-game purchases, such as upgrades, additional lives, currency, personalized avatars, an ad-free experience, unrestricted playing time or special items (i.e., famous football players’ avatars or skins that modifies the appearance of a character or item), and consumer spending on in-game purchases will grow until 32 billion in 2020. In fact, 4% of global iOS device users and 3% of Android users are making in-game purchases per month and [89].

In this context, online shoppers must demonstrate adequate self-efficacy as a precedent of Internet shopping and when buying game-related products. As [52] point out, Internet self-efficacy is not as important in offline contexts as in online contexts because consumers learn how to make an offline purchase at an early stage. In fact, consumers exhibiting low self-efficacy are uncertain and feel uncomfortable with Internet shopping [45]. Internet or computer self-efficacy is felt to be one of the best predictors of online shopping [43], [59].

To the best of our knowledge, the only study to specifically use self-efficacy during online shopping (online shopping self-efficacy as a concept) is [36]. Online shopping self-efficacy, or the belief in one’s ability to organize and perform the Internet actions required to produce given attainments, is a potentially important factor when buying online.

Following Rogers’ Theory of Innovation Diffusion, people are more likely to adopt an innovation they are comfortable with and that is compatible with other technologies they already use [79]. Following [13], [52], perceived self-efficacy will tend to transfer across two tasks within the same domain based on the degree of similarity between the qualitative features of the two tasks and the skill they require. Recently, [100] found evidence of transfer of technological game knowledge between different individuals interacting in the same group.

As [43] suggests, when users decide to adopt one activity they tend to adopt another similar one. According to his study, people are more likely to adopt information technologies that are functionally compatible with those previously adopted. More specifically, [43] proposes that prior use of the telephone is positively related to adopting online shopping and that Internet self-efficacy has a positive relationship with the adoption of online shopping. Following his arguments, and applying them to our research context, when playing with game apps and when purchasing online game-related products, when users complete transactions over the Internet they must transfer private information over the network creating a similar mode of action between the two contexts. Moreover, gaming and online shopping require a similar amount of skill and are contingent on the same premise (fulfilling users’ goals successfully and facing the challenge of winning the game or receiving high-quality game-related products). These arguments are coherent with what [79] calls technological clustering, which occurs when adopters of a given innovation perceive another new idea as being closely related. Therefore, when consumers exhibit self-efficacy playing a video game, they will probably display this ability when purchasing related products. Therefore, we propose that video gamers acquire several skills when playing, the fact that allow them to feel more self-confident and effective when using online shopping platforms to buy games or accessories.

This self-efficacy transfer argument also fits in with the evidence regarding the benefits of video games as teaching tools [25], [100]: they are used for military training, for teaching in schools, or for teaching perceptual skills (such as attending to more visual cues). “By playing video games the individuals learn scripts, organized sets of knowledge that define situations and guide behaviors” [25] (p. 368) and how to solve problems. They conclude that video games can teach, regardless of their content and regardless of the gamer’s intended use.

Thus, the following hypothesis tests self-efficacy transfer from one technological field (video gaming) to another (online shopping). Summing up, this hypothesis contains a novel contribution to the literature in this study and involves personality traits determining gamer self-efficacy and these skills being transmitted to other activities, specifically to online shopping.

H6: Gaming self-efficacy positively influences the self-efficacy of online purchases of game-related products.

2.5 The Influence of Self-Efficacy on the Intention to Purchase Online

One of the main goals pursued by most companies is to attract and retain customers, which is also applicable to online game-related shopping. Purchase intention is defined as the players' willingness to purchase virtual items for use in the mobile game [8]. Players may need to buy additional complements and items if they want to progress quickly in the game (extra time, lives, moves, virtual items to access higher levels, etc.). In this sense, one way to achieve such a goal is self-efficacy. In fact, some authors have emphasized the importance of self-efficacy in behavior and adoption [29], [36], [82]. Internet and computer self-efficacy are two of the best predictors of online shopping [52], [59], [91]. [43] posits that self-efficacy influences overall adoption of electronic shopping. Consumers engaged in Internet shopping transactions should demonstrate adequate self-efficacy which precedes perceptions strongly associated with Internet shopping [45]. In line with these findings and suggestions, the final hypothesis reflects the idea that if gamers display self-efficacy when purchasing online game-related products, they will later acquire more games and accessories online.

H7: Online shopping self-efficacy positively influences the intention to purchase game-related products online.

Figure 1 reflects the proposed structural equation model (SEM).

Figure 1: Proposed model 

3.1 Sampling and Measures

Information was collected from a sample of 366 Spanish game app players. The questionnaire was handed out personally to players of video games who have bought online video game related products/services (e.g., extras, premium game versions, music or games), and regarding the last app game played. On average 4.5 gamers in the sample play the same game. To collect data, we used snowball sampling. We contacted a group of players each of whom was asked to answer a questionnaire and to collect two additional questionnaires from other players they knew. Participants ranged in age from 18 to 65 (M=23, SD=7.81) and dedicated an average of 4.4 hours per week to playing game apps. The sample shows certain similarities with the average Spanish video game player. 65% is young (between 15 and 24 years old) and play an average of 6.6 hours per week [3]. Last available data from the gaming study show that 57.6% of females prefer to play with mobile game apps while 61.1% of males prefer consoles [4]. Gamer categories in Spain [54] include 56% of males, 21% between 6 and 24 years old; 35% of males between 25- and 64-years old play with games apps; 44% of females are gamers, 19% aged between 6 and 24 years old and 25% of females aged between 25 and 64 years old play game apps. Table 2 provides further details about the sample profile.

Table 2: Sample characteristics 

Note: younger individuals (<18 years old) must be discarded since it is necessary an official permission for being interviewed.

All of the items used to develop the measurement instrument were adapted to the context of the study based on the indicators proposed in the literature. The short version of the Big-Five Inventory (BIF) used here is based on the studies of [38], [67]. This scale contains all of the dimensions of the Five-Factor Model of personality [67], [68] and measures five orthogonal factors (Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness), one for each personality trait [7], [42]. In [48], it is suggested to include all the dimensions, since most individual differences in human personality can be classified into these five broad personality domains. The selection of specific subscales from the index to fit the research purpose was proposed by [7], [38], [47]. They have been effectively operationalized as reflective scales in recent studies on gaming and mobile purchase behavior [18], [42], [82].

Self-efficacy scales were likewise taken from prior literature so as to improve their validity content. We use the scales proposed in [13], [28], [35] to measure self-efficacy. Prior studies have tested the unidimensionality and validity of the reflective self-efficacy scale across several contexts, including mobile purchase and online behavior [82], [84]. In line with the recommendations of [21], selected self-efficacy items are the ones that best capture individual beliefs regarding specific tasks under research. To measure purchase intention, we employed a three-item reflective scale that refer to the individuals’ intention to invest additional money in the game app [8] (for item details see Table 3).

The translation and reverse translation processes were completed by two native-English-speaking researchers to guarantee the readability and clarity of the survey [24], [70]. Afterward, a pilot test with 15 players was conducted to pinpoint and correct shortcomings in the questionnaire. Five-point Likert scales were used, ranging from strongly disagree (1) to strongly agree (5).

3.2 Analysis and Results

Data analysis was conducted using LISREL 8.7. Confirmatory factor analysis (CFA) shows that all standardized loadings are significant at .001. After re-specification (items suppressed due to a low R2), results indicate that the measurement model provided a good fit to the data since all the fit indices met the threshold criteria recommended [5], [50]. Specifically, the model estimation showed a χ2(468) of 837.78 (p=.00), a Comparative Fit Index (CFI) of .93, a Goodness of Fit Index (GFI) of .90, and a Root Mean Square Error of Approximation (RMSEA) of .047. Likewise, all Cronbach alpha (α) values are greater than or equal to .7, except in the cases of agreeableness and extraversion, which were slightly below .7. However, since all of the remaining coefficients in our study fulfill the requirement, this might be due to the number of items comprising those constructs [47]. Furthermore, although the Cronbach alpha () is the most widely used estimator of scale reliability, it is thought to underestimate reliability. A better option to test reliability is composite reliability (CR) [78]. All (CR) values exceed .7, and average variance extracted (AVE) values exceed .5 [9]. Table 3 presents elementary descriptive statistics (mean and standard deviation) for the observable variables included in the measurement model, scales validity and the significant standardized coefficients lambda (λ) of the CFA. To test discriminant validity, the Fornell-Larcker criterion says that a factor’s AVE should be higher than its squared correlations with all other factors in the model. Thus, we discard discriminant validity problems [46] (Table 4).

Table 3: Measures, descriptive statistics and scales validity 

Notation: a Suppressed item, b mono-item, n.a. no applicable, SD: standard deviation α: Cronbach alpha, CR: composite reliability, AVE: average variance extracted.

Significance levels: **p<.01

Common method bias may exist when measures use the same method (self-report, scale type) and might yield inflated correlations between constructs. As recommended by [63], our assessment used the correlations between the Big-Five latent constructs, two self-efficacy constructs, intention to purchase, and the marker variable with a two-item measure of group norm (V44 and V45) to perform this analysis. The correlation between group norm and purchase intention was .47. Using [63] formula, the corrected values for the correlations between consciousness and group norm, we computed the Common Method Variance (CMV) adjusted correlations among the research constructs using the second lowest correlation (r=.04) as a proxy for marker variable. Table 4 shows the correlations between constructs below the diagonal and the adjusted correlations above the diagonal. We also determined the statistical significance of the adjusted correlations. All the correlations that were significant before adjustment also remained significant after adjustment. In this sense, marker-variable analysis indicated that common method bias is unlikely to affect the findings of this study.

Table 4: Results for Fornell-Larcker criterion to test discriminant validity and marker-variable analysis 

Significance levels: **p<.01, *p<.05

Note: values below the diagonal represent correlations between constructs, values on the diagonal (in bold) represent the square root of AVE and values above the diagonal represent the correlations between constructs adjusted for the marker variable (group norm).

The following step in the analysis was to test the proposed hypotheses using structural equation model (SEM) analysis (Figure 2).

Figure 2 reflects the proposed the results of the structural equation model estimation.

Significance levels: **p<.01, *p<.05 

All the hypotheses are supported at a confidence level of 95% except H1 and H5, which are rejected. Following prior research on self-efficacy [2], in order to ensure the validity of our results two gamers’ behavioral variables (frequency of gaming with apps and frequency of online purchasing) were included in the model as control variables. Attending [19] recommendation, control variables must hold theoretically meaningful relationships with predictors and criteria. In this sense, prior research suggests that usage and purchase should be employed as dependent variables and relate to the frequency of this purchasing and gaming behavior [37]. Specifically, [2] support a bidirectional association between frequency of playing video games, self-reported problem-solving skills and gamers’ behaviors (that may include online shopping related products and services). However, after controlling for these playing-related variables, no changes either in the significance or direction in the proposed relationships were detected. Table 5 presents the assessment of the overall model fit and the hypotheses supported.

Table 5: Results of SEM estimation of the proposed model 

Significance levels: **p<.01, *p<.05, ns non-significance

Notation: Beta coefficient (β); chi-square statistic (χ2), Root-Mean-Square-Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Goodness-of-Fit Index (GFI).

In addition, the Baron and Kenny [15] approach was followed in order to test the mediation effect of shopping self-efficacy in the model. We created a model including direct as well as indirect paths [80]. In other words, this model included all the paths in our research model (Figure 1) as well as an additional direct path from gaming self-efficacy (independent variable) to purchase intention (dependent variable). To test the mediating role of shopping self-efficacy, the mediation analysis results show that gaming self-efficacy remains significantly related to purchase intention after controlling for the mediating variable (shopping self-efficacy) [(ß=.252; p<.05)], suggesting a partial mediation effect of online shopping self-efficacy.

4.1 Managerial Implications

From a managerial perspective, understanding the phenomenon of self-efficacy transfer is crucial, as media such as game apps are beginning to play a major role in marketing activities. Companies are recognizing the potential of game apps as a means of marketing products and services. In agreement with [86] and the Theory of Innovation Diffusion, the popularity of video games makes them spread through social networks, motivating players through innovative and unique designs. Since, actual digital media users’ philosophy of open access implies no immediate payment in order to play video game [44], game developers ought to be aware of the alternative paths (such as micropayments or virtual assets transactions) to reach gamers’ purchase intention (in order to maintain the ecosystem developed for generating revenues).

Our results reveal how gaming might have practical implications for marketing managers. If firms teach players how to play, it could indirectly promote the online purchase of related products, which also supports the effectiveness of gamification strategies. In this sense, [53] propose the use of gamification strategies to offer a good video game, with added value, thus enhancing the playing experience and entailing benefits for both game developers and players alike. Keeping in mind that central to the game model is consumer choice, players will accept better non-imposed marketing strategies. In this sense, game developers may let master players (i.e., that show high gaming efficacy) to exchange information (tips, strategies or trick) for extra features with others, making easy microtransaction process between consumers (C2C commerce inside the game).

Managers should try to develop user-friendly and easy to surf sites so as to enhance convenience and allow users to become familiar with interfaces and show self-efficacy with Internet behaviors. In [43], the aspect of equivalent skill level as analogous to Rogers’ evaluation criteria of perceived complexity is studied, which predicts faster adoption when technology is easier to use. A practical guideline for gaming companies is to employ social networks to connect with their customers to ask for feedback about the perceived level of online tasks difficulty and self-efficacy perception (about gaming and purchasing).

This research denotes possibilities for interventions by consumer educators and managerial video game industry actors. For example, platforms that offer a sense of comfort and an appealing task-accomplishment orientation helps to promote self-efficacy among users [45], [94]. To that end, providing flexible learning environments, also among members of a group who play games [100], can also be useful to promote user self-efficacy and help them achieve their goals successfully. Regard this result, game firms might reward players with free items (such as new level trials in advance, outfits or skipping stages) in return for sharing their knowledge with other gamers, if they are coaching other players, they are promoting and recommending [11], [93] to consume certain games and, finally enlarging the life cycle of the games and related products.

Finally, consumers who register high emotional stability and extraversion and low agreeableness are the target audience for firms operating in the video game industry as these are the personality traits that most influence self-efficacy and, consequently, the opportunities to acquire new abilities and transfer them to online buying behavior of game-related products. As app developers employ various business models to generate income, such as pay to play games, in-app purchases during play, or in-app advertising [40], gaming companies might be able to incentive gamers with a free trial period, free virtual elements or ad-free period experience in exchange for the information regarding gamers’ personality and perceived self-efficacy to segment their market.

5.1 Limitations and Further Research

The results are valid for Spain and for the video game industry and cannot, therefore, be generalized unambiguously. The sample is composed of game app players, who display a certain level of gaming self-efficacy. It would thus be advisable to study the moderating role of playing experience and the online purchase experience. Furthermore, we use a short version of the Big-Five inventory. Future studies should use another measurement instrument such as the 60-item NEO five-factor inventory [33] and should go further about how could vary gamer’s consumption behavior when playing different games apps at the same time. Other related factors such as the ease of use and purchase within each game [37] could be considered as being also of interest for marketers.

It might also prove enlightening to study the negative aspects of playing and online purchasing, such as perceived risk, social isolation or addiction, which leads to the interesting issue of self-regulation [35], [51]. It would be prudent to compare real-world activities and virtual worlds so as to ascertain whether adopting certain activities such as game app playing is a substitute for the real-world counterpart, as [43] suggests doing. In the same line, it would be interesting to compare the behavior of gamers and non-gamers [51] as well as to consider other antecedents of self-efficacy, such as emotional intelligence [91]. More research could be conducted for specific games, as there can even be a self-efficacy specific to a certain game. Finally, different genres of game apps might spark a different degree of self-efficacy transfer and personality trait influence.