Motivation factors influencing intention of mobile sports apps use by applying the unified theory of acceptance and use of technology(UTAUT)

Unified Theory of Acceptance and Use of Technology

In the communication literature, there are several theoretical approaches to explain the concepts of use intention of new innovation and technology. As mentioned, the current study adopted and modified the UTAUT in developing the current study. The UTAUT is a comprehensive synthesis regarding Information and Computer Technologies (ICTs) adoption drawn from the compounding theoretical factors of the Diffusion of Innovation Theory (DIT), the Social Cognitive Theory (SCT), the Theory of Planned Behavior (TPB), and the Technology Acceptance Model (TAM). With a dearth of study regarding mobile sports apps and its motivations factors, applying and modifying the UTAUT into mobile sports apps users’ motivation may provide theoretical and contextual understanding in the sport communication and management field.

The DIT is a psychological framework that identifies a sequential process influencing the adoption of innovation (i.e., idea. object, or product perceived as new by an individual) (Lyytinen & Damsgaard, 2001; Rogers, 1995). The DIT explains a process consisting of four states (i.e., dissemination, adoption, implementation, and maintenance) (Rogers, 1962; Rohrbach, Graham, & Hansen, 1993). The SCT suggests that a combination of behavioral, cognitive, and environmental factors influences human behavior (Bandura, 1997; Couros, 2010). The SCT formulates of social situation, self-efficacy, outcome expectations values, and self-regulation to predict one’s behavior (Winters, Petosa, & Charlton, 2003). Bandura’s (1994) defined the concept of self-efficacy that one’s courses of action required to attain designated types of performances. The TPB is a conceptual theory to explain human behavior through intentions, attitude toward the behavior, subjective norm, and perceived behavioral control which are preconditions to be met in the process (Ajzen, 1991, 2002; Kang, Lee, & Kwon, 2013). The TAM is to demonstrate how human perceives usefulness and usage intentions of a new technology in terms of social influence and cognitive instrumental processes (Davis, 1989; Venkatesh & Davis, 2000). Venkatesh (2000) examined two variables (i.e., perceived usefulness and perceived ease of usage) to determine one’s behavioral intention to use a technology. In fact, Kang et al. (2015) utilized TAM as a theoretical framework in understanding sport apps usages among college students.

In summary, the DIT posited that adoption or intention of new innovation depends of advantages and complexity (Rogers, 1962; Kang, 2014). On the other hand, the SCT and the TPB stated the human characteristics are dispositional or contextual factors that influence intention to adopt a new innovation (Ajzen, 1985; Bandura, 1997). From these literatures regarding adoption of innovation, the TAM was introduced to explain factors influencing adoption of ICTs and predicted that perceived easiness precedes perceived usefulness of ICTs (Kang, 2014).

The UTAUT, which is a comprehensive theory of prior technology acceptance research, expands and combines the ICTs theories. The theory, specifically, integrates the concepts of attitudes, social influence, and self-efficacy from the SCT and the TPB, easiness from the TAM, and compatibility from the DIT (Kang, 2014; Venkatesh, Thong, & Xu, 2012). The UTAUT, thus, uses four core factors (i.e., performance expectancy, effort expectancy, social influence, and facilitating conditions) and four moderating variables (i.e., age, gender, experience, and voluntariness of use) to explain behavioral intention or use of behavior of communication technology (Im, Hong, & Kang, 2011; Oshlyansky, Cairns, & Thimbleby, 2007; Venkatesh, Morris, Davis, & Davis, 2003; Venkatesh, Thong, & Xu, 2012). However, facilitating conditions which is mainly regarding of communication motivation factors are not included in this study because main reasons of using mobile sports apps are not communication purposes among sport sports apps users.

Performance expectancy is defined as the degree to how technology users believe that the use of technology provides functional advantages to perform certain activities (Rogers, 1995; Venkatesh et al., 2003). In other words, performance expectancy is the expected overall impact of technology’s functional advantage on job performance and outcome (Davis, 1993; Kang, 2014). In order to adopt the concept of performance expectancy for the current study, sport consumers use mobile sports apps because it enables to achieve their task performance (e.g., accurate and quick information) and find impacts and values of the apps. As performance expectancy predicts intention to use technology, which suggest the following hypothesis for the current study.

Effort expectancy refers to the degree of ease associated with people’s use of technology (Venkatesh et al., 2012). Technology users tend to adopt and use new technology which has an use of simplicity and maximum efficiency (Agarwal & Prasad, 1999; Venkatesh & Davis, 2000; Venkatesh & Morris, 2000). In other words, effort expectancy for the current study measures how easy sport consumers can use mobile sports apps. Kang (2104) stated that people more likely tend to use a new technology when effort on understanding of a new technology is not hard. Schaper and Pervan (2007) studied occupational therapists’ ICT acceptance and found effort expectancy was an influential factor of technology use. Martins, Oliveira, and Popovic (2014) also provided similar results that effort expectancy is a strong predictor of intention and use of technology for internet banking users. Additionally, previous studies adopting the UTAUT support that the effects of effort expectancy positively correlate and influence use intention of technology directly or indirectly through performance expectancy. The current study, thus, suggests the following hypotheses.

Significant others’ experience and opinion in technology adoption is one of the core key factors in the UTAUT. Social influence is the extent to which important others (e.g., family and friends) influence personal intention to use technology (Ajzen, 1985; Venkatesh et al., 2012). A volume of recent research also supports that significant others leverage on adoption of intention to use technology (Nikou & Bouwman, 2014; Teo & Noyes, 2014; Workman, 2014; Zhu & Chang, 2014). Therefore, social influence might be a predictor of intention to use mobile sports apps that suggest the following hypothesis.

The UTAUT, however, has been empirically validated by many previous studies, it is likely oriented toward functional, social, and extrinsic (e.g., utilitarian motivation) concepts in the adoption process of previous technology theories as Kang (2014) critically pointed. The intrinsic motivations, therefore, such as individual and psychological factors, should be included because people have different motivations of intention to use technology (Kang, 2014; Venkatesh et al., 2012). The motivation factors of entertainment, social utility, and communication are integrated in the UTAUT to accommodate mobile sports apps use context. For example, in the literatures of technology use motivation, many studies supported that entertainment, social utility, and communication are the major motivations to adapt and use of technology (Boyle et al., 2012; Joo & Sang 2013; Leung & Wei, 2000; Lin & Lu, 2011; Shim, You, Lee, & Go, 2015;Vrocharidou & Efthymiou, 2012). However, communication motivation factors are not included in the current study because mobile sports apps are not normally used for communication purposes among sport consumers.

Entertainment motivation refers to the extent to which people can obtain fun and escaping pressure from using a technology and plays an important role to accept and use of technology (Ko, Cho, & Roberts, 2005; Lee & Ma, 2012; Luo, Chea, & Chen, 2011). Social utility motivation refers to the extent to which people can take functional information and services (e.g., news, weather information, merchandising information) using technology (Alhabash, Chiang, & Huang, 2014; Luo & Remus, 2014; Kang, 2014).

The following hypotheses which are integrated with motivational, functional, and contextual factors to intensify the explanatory power of mobile sports apps use context are suggested:

The relationships among variables are displayed in a hypothesized model (see Figure 1).

Figure 1.

Hypothesized Mobile Sports Apps Use Model

Participants

A total of 320 paper questionnaires was distributed to college students with a variety of majors including sport management, tourism, journalism, informatics, business, hospitality, engineering, law enforcement, nursing, finance, and others at a large-sized university located in the Midwest. The subjects for the current study were college students who encompasses the greatest proportion of mobile technology devices users. For example, about 83% college students regularly use a smartphone and nearly four in ten students (38%) own a tablet computer in 2013 (Pearson, 2013). The questionnaire included an informed consent form, demographic backgrounds, and questions related to the modified UTAUT for mobile sports apps.

As a result of survey procedures and after eliminating incomplete questionnaires, a total of 267 responses (N=267) were garnered for main analysis. Participants consisted of 181 male students (n=181, 67.8%) and 86 female students (n=86, 32.2%). Approximately 48.3% of the total participants aged 17-20 (n=129; 48.3%), 21-25 (n=109; 40.8%), 26-29 (n=22; 8.2%), and older than 29 (n=7; 2.6%). Respondents were judged themselves they have very good mobile technology device skills (n=106; 39.7%), good skills (n=131; 49.1%), fair skills (n=26; 9.7%), and poor skills (n=4; 1.5%).

Instrumentation

A 35-item survey assessed performance expectancy, effort expectancy, social influence, entertainment motivation, social utility motivation, continuance intention, and demographic information (e.g., age, gender, and length of mobile sports apps usage). Each item of all variables was rated on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Four items of performance expectancy were modified from the UTAUT about the extent to which mobile sports apps help facilitate task performance. Four effort expectancy items regarding the degree of ease to use mobile sports apps were modified and four items of social influence were also examined. Specific motivations of mobile sports apps including entertainment and social utility also were modified from previous studies (Chang, Lee, & Kim, 2006; Park, 2010). Additionally, another four items of continuance intention which was measured the degree of mobile sports apps users continue to use in future (Agarwal & Prasad, 1999; Davis, 1989; Moore & Benbasat, 1991; Venkatesh et al., 2003; Venkatesh & Davis, 2000). (See Table 2).

Table 2

Means (M), Standard Deviations (SD), and Coefficients (λ) for Variables

(N=267)

Table 1

Demographics of Respondents

Data Analysis

In order to validate the measurement model of the current study, an omnibus confirmatory factor analyses (CFA) was applied using AMOS 22 with all constructs. In validating the relations of the observed variables and the underlying constructs, four indexes, the chi-square test, the Bentler Comparative Fit Index (CFI), and the Steiger-Lind Root Mean Square Error of Approximation (RMSEA) were utilized to examine the absolute and comparative fit of the model (Kline, 2010).

Next, descriptive statistics including means and standardized deviations as well as demographic information of the participants were calculated. Then, the inter-item reliabilities of each scale were confirmed by assessing Cronbach’s alpha internal consistency coefficients. Finally, the research hypotheses for the model were tested using structural equation modeling (SEM). The model testing followed maximum likelihood procedures. SEM assumes recursive relationships among latent variables and tests these relationships against the data collected in an integrative and simultaneous way. The analytical procedure assumes a causal structure between exogenous and endogenous variables and allows for the estimation of direct, indirect, and total effects (Kline, 1998). The same four indexes that were previously used for the CFA were also utilized to assess the structural model test.

Measurement Model Testing and Modification

The overall fit indices of the CFA including all constructs indicated somewhat poor fit with data [χ²(df) = 627.637 (194), χ²/df = 3.235, p < .001; CFI = .912; RMSEA = .092]. For performance expectancy, two additional residual correlations between items “I find mobile sports apps useful in my life” and “Using mobile sports apps increases my productivity” and items “Using mobile sports apps increases my chances of achieving things that are important to me” and “Using mobile sports apps helps me accomplish things more quickly” were made based on t two high modification index (MI) values. According to Byrne (2010), those items with high correlations (above r = .70) could cause correlation errors. For effort expectancy, two additional residual correlations between items “Learning how to use mobile sports apps is easy for me” and “My interaction with mobile sports apps is clear and understandable” and items “I find mobile sports apps easy to use” and “It is easy for me to become skillful at using mobile sports apps” were also made due to the high correlations with each other.

The model fits were reevaluated with the revised measurement model. The data [χ²(df) = 498.801(189), χ²/df = 2.639, p < .001; CFI = .937; RMSEA = .079] showed the acceptable fit. Convergent validity and discriminant validity were supported because average variance extracted (AVE) values of each factor exceeded .50 (Hair, Anderson, Tatham, & Black, 2006) and all estimated values of AVE were greater than the squared correlations between the factors (Fornell & Larcker, 1981). Inter-reliabilities of all constructs were also supported by Cronbach’s alpha coefficients of each construct exceeding the recommended benchmark of .70 (Nunnally & Bernstein, 1994). Table 2 provides the means, standard deviations, and factor loadings for all variables. The means for five influencing factors for continuance intention of sports apps ranged from a low for social influence (M = 3.92) to a high for effort expectancy (M = 5.89). Social utility motivation (M = 5.53) was the second most important factor for sports apps consumers, followed by entertainment motivation (M = 5.14) and performance expectancy (M = 4.28). However, social influence was below the scale midpoint (M = 4.00). Table 3 reports the correlations among the constructs and the Cronbach’s alpha coefficients of them.

Table 3

Correlations and Cronbach alpha (α)

Figure 2.

Final Research Model

*p < 0.05, **p < 0.01

Structural Model Testing

The results of the proposed structural model performance expectancy, effort expectancy, social influence, entertainment motivation, social utility motivation, continuance intention showed the acceptable fit indices [χ²(df) = 530.215 (191), χ²/df = 2.776, p < .001; CFI = .931; RMSEA = .082]. The results demonstrated performance expectancy (τ = *p < 0.05, **p < 0.01 *p < 0.05, **p < 0.01 .178; SE = .035; p < .01) and effort expectancy (τ = .257; SE = .073; p < .01) had positive impacts on intention of mobile sports apps use. In addition, effort expectancy (τ = .301; SE = .098; p < .01) had positive impact on performance expectancy. Thus, Hypotheses 1, 2 and 3 were supported. However, Hypothesis 4 (social influence → intention of mobile sports apps use) was not supported.

In terms of the relationships between two motivations and other variables, social utility motivation had positive impacts on all outcomes; performance expectancy (τ = .303; SE = .141; p < .01), effort expectancy (τ = .466; SE = .110; p < .01), social influence (τ = .247; SE = .146; p < .05), and continuance intention (τ = .339; SE = .109; p < .01), while entertainment motivations had had positive impact on continuance intention (τ = .216; SE = .075; p < .01). Thus, Hypotheses 5a, 5b, 6b, 7b, and 8b were supported, while Hypotheses 6a, 7a, and 8a were not supported.