Internet has elicited a new decision making process in the area of political marketing thanks to the emergence of new communication tools. This paper focuses on weblogs as pioneer players on Web 2.0 in order to ascertain their actual sphere of influence on individuals’ political behaviour. Focusing on the case of weblogs and assuming a wide and integrative theoretical approach, a hybrid model which merges marketing, technology and political science is proposed and tested using a sample of 39 weblogs and 890 blog users in an innovative attempt to shed light on these virtual-physical dynamics. The results are promising and open the door to the comprehension of a new, emerging framework where Internet tools are intended for a major, critical role in the political arena.
Internet ha favorecido un nuevo proceso de toma de decisiones en marketing político gracias al surgimiento de nuevas herramientas de comunicación. El presente artículo se centra en los blogs como agentes pioneros en el contexto de la web 2.0 con el objetivo de determinar su nivel de influencia real sobre el comportamiento político individual. Centrando el análisis en el caso de los blogs de naturaleza política y asumiendo un extenso e innovador marco teórico de partida, se propone un modelo híbrido en el que confluyen marketing, sistemas de información y ciencia política que es validado a partir de los datos de una muestra de 39 blogs y 890 lectores en un intento innovador de conocer más acerca de estas dinámicas reales y virtuales. Los resultados son prometedores y abren la puerta a la comprensión de un nuevo y emergente ámbito de trabajo en el que las herramientas de Internet cobran un papel más relevante dentro de la arena política.
Political activism is a well-known construct (Hoffman, Jones, & Young, 2013; Norris, 2005) whose effects on voting have already been addressed (Franklin, 1996; Vecchione et al., 2015). Nevertheless, research on how political activism is formed and its antecedents have scarcely been studied and, to a lesser extent, in an online setting. Since the term ‘weblog’ was first coined and defined, blogs have evolved rapidly, the barriers for starting a weblog have increasingly diminished and they have come to cover almost every topic of interest (Levy, Journell, He, & Towns, 2015; Moen-Larsen, 2014; Voicu, 2015). The political sphere is no exception and with a popular issue, a blog can attract attention and exert considerable influence on individuals (Hsu & Lin, 2008). Since 2003, when the first strategically-oriented political blog was created (Howard Dean's ‘Blog for America’), politicians and political organizations have massively incorporated the use of blogs and social media in their plans and programmes too (Adams & McCorkindale, 2013; Hong, 2013; Kruikemeier, 2014; Mergel, 2013; Yi, Oh, & Kim, 2013), whereby academic political marketing research seems to be falling behind.
Although a number of studies on the impact of technologies on political behaviour have been carried out recently (Criado & Rojas-Martín, 2013; Lilleker & Jackson, 2013; Panagiotopoulos, Bigdeli, & Sams, 2014; Park, 2013), the use of a political marketing approach is still scarce (Vesnic-Alujevic, 2012). This paper attempts to gain a deeper insight into this line of research by proposing an integrative model including innovative constructs such as political activism and subjective political knowledge where blogs occupy a central position. Hence, the objective here is to assess the influence of blogs on political activism by emphasizing on website factors, such as perceived usefulness, perceived trustfulness and perceived information quality.
Thus, a structural setting of factors in which blog usage and political activism are lastly connected has been proposed to foster a solid relationship between online and offline contexts and to increase the importance of a relationship marketing approach in the political field. Given that non-voting periods are less intense in political activism, focusing on these periods can exhibit straightforward relationships not biased by voting behaviours that might be triggered by immediacy and online and offline candidates’ campaigns. Voting behaviour, indeed, has been widely researched (Falkowski & Cwalina, 2012; Sheth, 1976), but little attention has been focused on political behaviour in the ordinary political cycle and even less on interactive technologies applied to political contexts.
The rest of the paper is organized as follows. First, a theoretical framework is developed and the derived hypotheses are proposed. A special emphasis will be given to the theoretical framework section, beyond the specific theoretical support of each hypothesis, aiming to anchor this new communication tool and the type of interactions into established conceptual frameworks. The second part describes the methodology used and then the discussion of the results is presented, which addresses the questions and outlines the managerial implications, limitations and directions for future research.
Theoretical framework and empirical modelTheoretical frameworkGenerally defined as websites where people write about recent events or topics of their own interest, blogs are technically analogous to common websites (Chesney & Su, 2010) and the core of the proposed model and a clear example of interactive communication tools. From a theoretical perspective this research is anchored in the following previous contributions: (i) considering blogs as an interactive nature, given that the quantity of information is the most important determinant of interactivity and that is, exactly, the common feature of all blog dynamics (Rafaeli, 1988; Song & Zinkhan, 2008). (ii) Since blogs allow various individuals –and even organizations- to engage in mediated communications which result in exchanges between them and blog authors and other readers over time, this can be embraced by the relationship marketing theoretical framework (Grönroos, 1994) and lastly fed by Vargo and Lusch's (2004) service-dominant logic. Fig. 1 delineates the conceptual relationships that will be justified below.
Relationship marketing has particularly been strengthened by the impact of new knowledge technologies (Sheth, 2002). However, in the field of political marketing, the application of this relational paradigm is new, as long-term relationships are created between voters and politicians in order to make public policies sustainable in times of crisis or uncertainty (Lees-Marshment, 2014). In turn, key topics of relational paradigms such as collaboration, trust and commitment or value addition seem to have been consistently transferred to the Internet scenario, where users constantly interact and blogs become an information source for a range of fields, such as political issues. This interaction orientation is becoming increasingly important in marketing according to Vargo and Lusch's (2004, 2008) service-dominant logic, in which the customer is always a co-creator of value. In our context, this means that blog users could get involved with the political blog in a wide range of interactive processes which might result in cognitive or behavioural experiences. These encounters between the site and the user are defined within the limits of the conceptual framework for co-creation of value (Payne, Storbacka, & Frow, 2008), whose guidelines are also assumed by this paper.
Since the political market is ideologically biased, many political blogs reinforce social cleavages and, as a result, groups of people become virtually linked and reinforced in virtual communities. This paper will not focus, however, on the aggregate blogosphere as a unit of study, instead, the analytical framework is based on pioneer consumer behaviour models (Engel, Kollat, & Blackwell, 1978; Howard & Sheth, 1969; Nicosia, 1963). Their key and wide-ranging features have been proved to be fully valid in virtual contexts (Richard, Chebat, Yang, & Putrevu, 2010) and certain related specific factors (e.g. perceived trustfulness, subjective knowledge) have also been applied to political marketing research. Furthermore, the Theory of Reasoned Action (Fishbein & Ajzen, 1975) has already been applied to voting behaviour (Hansen & Jensen, 2007) and seems especially suitable in contexts where decision changing barriers are simply irrelevant, as in the case of websites and blogs, spheres where users can quickly leave the page if content expectations are not satisfied.
Weblogs are also good examples of the drastic change in the way information is spread to the audience: political blog authors act as those leaders who filter information or give their opinion, so their influence becomes significant to their readers. As a result, as blogs emerge as a new and complex class of information system, factors such as information quality or perceived usefulness remain theoretically unavoidable. This emphasizes the allusion to DeLone and McLean's (1992) work as the core of a dimension directly engaged in a more technical and multidisciplinary description of our model which has conveniently been tested and extended over time (Petter, DeLone, & McLean, 2008). Thus, the proposed model is also fostered by the Technology Acceptance Model (TAM) (Davis, 1989), which constitutes one of the easiest, widespread and tested schemes to predict technological usage in a good number of applications, including websites. Indeed, in the case of this study, TAM is considered the most relevant theory which helps to explain why a blog is first considered a valid source of information and, eventually, fully adopted by an individual.
Given that certain key elements in information systems theory give feedback to TAM and in turn, TAM has also proved successful at explaining TRA in technological environments (Venkatesh, Morris, Davis, & Davis, 2003), the blog usage acceptance proposal by Hsu and Lin (2008) could finally depict a more specific blog-related context merging technology acceptance, social influence and knowledge sharing motivation. Specifically, there is direct contribution from these three theoretical approaches to the factors proposed in our model in the following manner: political involvement and political activism are used to establish a parallelism between behavioural intention and actual behaviour, a critical model sequence considered by TAM and TRA. Regarding our trigger factors-blog perceived information quality and blog perceived reputation-, both are developed by TRA in the sense of opinions of referent others, while Hsu and Lin's model also considered blog perceived reputation as one of its knowledge sharing factors. Subjective political knowledge could undoubtedly be linked to the subjective norm present in TRA. In addition, TAM and Hsu and Lin's blog usage acceptance model are major references to explain blog perceived usefulness and blog perceived trustfulness in the context of our proposal. This becomes feasible because these variables are delimited by two theoretical depictions: while TAM examines perceived ease of use and perceived usefulness, Hsu and Lin include perceived usefulness and ease of use as technological acceptance factors and trust as knowledge sharing factor. This entire theoretical scheme is shown in full in Fig. 1.
Empirical model, hypothesis development and proposed modelThe traditional understanding of information quality has definitely been altered by the emergence of a new, heterogeneous technological framework which has been used to trigger our model, depicted in Fig. 2. In our case, blog perceived information quality remains as the only trigger factor due to a multidimensional nature that refers to the semantic success of information when transmitting a searched meaning (Chen, Shang, & Li, 2014). This meaning is relevant as long as information quality has been linked to purchase evaluation assessment as a means of establishing a positive image of companies and strengthening long-term relationships with buyers (Hoffman & Novak, 1996). This logic is analogous for the political field: a political blog whose information quality may be perceived as high will have loyal readers and, eventually, will be able to develop a positive image.
Knight and Burn (2005) compiled up to twenty formative dimensions (e.g. usefulness, reputation, relevance, credibility or consistency) which can be categorized into informative content, informative format and physical context (Jeong & Lambert, 2001). The latter emphasizes a feasible shift towards a virtual context where usefulness, trustfulness or reputation gain key importance. Following Bailey and Pearson (1983) and considering user satisfaction as a key factor linking personal necessities and information system benefits, an assessment of information quality could be an accurate indicator of website usefulness (DeLone & McLean, 1992), defining this variable as the degree to which a person believes that using a particular system (a political blog in our case) would enhance his or her performance (Davis, 1989). This link has proved successful for blogs (Zehrer, Crofts, & Magnini, 2011), so we propose the following hypotheses:H1 Blog perceived information quality will positively affect blog perceived usefulness.
In offline and online buying behaviour, and also in political contexts where citizens are often interested in minimizing uncertainty, information search is regarded as a way of reducing risk in decision making (Zaichkowsky, 1985). This link might be more intense through the Internet (Spink, Bateman, & Jansen, 1999): since political blog users require quality information to be able to reduce their uncertainty on political issues and quality information search could also be featured as a user criterion relating to excellence and trustfulness (Taylor, 1986), the blogosphere appears to be a suitable platform for effective trust building, thanks to the exchange and sharing of these information opportunities. Social communities of mutual interests – such as political blogs – have become increasingly influential sources for their users due to the addition of value that long-term interaction creates in terms of credibility and reliability. Indeed, information quality has been identified as one of the most relevant variables to promote trustfulness in online interactions (Filieri, 2015; Yi et al., 2013), just as it occurs in political blogs. Therefore:H2 Blog perceived information quality will positively affect blog perceived trustfulness.
Reputation has been defined as what it is said or believed about the character or position of a person (McKnight, Cummings, & Chervany, 1998) and brands or products in different fields (Roberts & Dowling, 2002), including politics. It is a complex construct whose meaning has been explored from different approaches and also, specifically, in the online environment (Dholakia, 2005). Following Taylor's (1986) work, information quality and reputation have been consistently linked concepts in a wide variety of contexts, such as publishing (Janes & Rosenfeld, 1996), to which the political blogosphere displays close similarities. Consequently, quality and reputation show a strong relationship first highlighted by Shapiro (1982) regarding product contexts and later reinforced in technological environments, such as the Internet and its websites (Knight & Burn, 2005). We posit this relationship can also be applied to the online political environment:H3 Blog perceived information quality will positively affect blog perceived reputation.
There is an inferred similarity between corporate reputation and the political blogosphere. Based on Gotsi and Wilson's (2001) analysis, the perceived reputation of a political blog and corporate reputation (1) are both dynamic concepts which may fluctuate through the time; (2) are built and managed focusing on the long term, just as Dellarocas (2003) also pointed out for websites; (3) are closely related to the concept of corporate image, which utterly depends on symbolism and communication patterns, just as it happens with the political discourse; and (4) include specific rankings where both companies and blogs are compared to others. Since this construct could be applied to political blogs, in this virtual context reputation has also been linked to quality, perceived value and loyalty (Caruana & Ewing, 2010), for which the existence of trust becomes paramount. Indeed, the very definition of reputation (Doney & Cannon, 1997) connects this construct to trustfulness, a linkage solidly consolidated by subsequent studies (Jarvenpaa & Leidner, 1998; Yoon, 2002), too. Positive reputation has also been associated with performance, since the credibility of the messages of a website enhances the perception of trustfulness. Prior literature highlights the relationship between trust and reputation, but the two constructs show differences. In this vein, trust ultimately is a personal and subjective phenomenon based on various factors such as reputation (Jøsang, Ismail, & Boyd, 2007). Therefore, in our context it could be hypothesized that:H4 Blog perceived reputation will positively affect blog perceived trustfulness.
Subjective knowledge is defined as a subjective rather than a technological trait, related to self confidence in cognitive judgement (Brucks, 1985). This view is fully extendable to political marketing contexts through opinion leadership (O’Cass & Pecotich, 2005), implying social abilities to influence others. Although the relationship between subjective knowledge and involvement has been thoroughly studied, there seem to be no univocal results. While a substantive body of literature posits that involvement would affect subjective knowledge (Burton & Netemeyer, 1992; Liang, 2012; O’Cass, 2002; Zaichkowsky, 1985), other authors (Bloch, Black, & Lichtenstein, 1989; Park, Mothersbaugh, & Feick, 1994; Sohn & Leckenby, 2005) have found evidence for the opposite relationship: subjective knowledge drives involvement. This latter view is the one taken by this paper: the more political information a blog user seeks, the more knowledgeable about politics he becomes, and therefore an increasing involvement with this matter may arise. In this context, perceived risk emerges as the solid trigger (O’Cass & Pecotich, 2005) which makes information search a persistent task, a relevant issue when approaching contexts such as politics, where decisions are not strictly mercantile.
The perspective of the relationship between media usage and involvement also reinforces this reasoning, especially on Internet. Several works have highlighted a stronger involvement in individuals who used digital sources of information or participated in open discussions on political matters online (Rojas et al., 2005; Shah, McLeod, & Yoon, 2001). This implies that individuals exposed to political information on Internet do acquire subjective knowledge as a natural part of this process, which ends up with a deeper involvement with the issues being discussed. Therefore:H5 Subjective political knowledge will positively affect political involvement.
Perceived usefulness was prominently considered by Davis (1989) as the balance between a rise in professional performance and information system usage, a pragmatic approach whose effects have already been tested in virtual contexts (Pikkarainen, Pikkarainen, Karjaluoto, & Pahnila, 2004) such as the blogosphere (Hsu & Lin, 2008). Despite the fact that academic literature does not offer specific references where perceived usefulness drives subjective knowledge, a sizeable academic literature (Chang & Wang, 2008; Shih, 2004; Suh & Han, 2002) has established a solid relationship between certain attitudes towards a system and its very usefulness. If we widen our scope and focus on opinion leadership as one of the prominent attitudes related to the usage of a political blog, seminal references (Berelson & Steiner, 1964; Corey, 1971) will serve to depict the idea of leadership defined by (1) a higher involvement with political issues; (2) a deeper knowledge of this issues; and (3) a larger amount of information sources and political information acquired.
When focusing on the political blogosphere, opinion leadership could be easily linked to interaction (Flew, 2002) or collective communication (Drezner & Farrell, 2008), which also lead to a higher degree of social pressure to others as well as to WOM and advice in the assessment of a decision (Flynn, Goldsmith, & Eastman, 1996). Given that political opinion leadership lives in the heart of political blog usage, we could state that none of its traits could be feasible without the existence of subjective knowledge. Hence, becoming more knowledgeable in the online context would definitely depend on the usefulness of the systems an individual uses as his information sources. This is why our model proposes the following hypothesis:H6 Blog perceived usefulness will positively affect subjective political knowledge.
If we consider involvement based on a classic approach as the perceived personal relevance (Petty & Cacciopo, 1981), we should say the relationship between this construct and usefulness is not new at all. A good number of references (Boland, 1978; Swanson, 1974) established that the success of an information system – that is, its global degree of usefulness for users – could be increased by a higher involvement of these users with the development of the system, and more recent works (Abelein & Paech, 2015; Amoako-Gyampah, 2007; Bano & Zowghi, 2015) just contribute to strengthen these conclusions. Our field of study is not an exception: a political blog whose contents become useful will naturally lead to a more involved and eventually abundant community of readers.
Our model proposes this relationship works just the opposite way: initially, an individual seeks information for personal purposes – minimizing the perceived political risk, for instance – and it is the very usefulness of the source, a political blog in our case, which makes the difference and turns a functional need into a more emotional and intense issue such as an emerging, global and deeper involvement with political matters. The specific direction of this linkage has already been successfully tested in brand communities (Okazaki, Rubio, & Campo, 2012) and also in a digital context related to the blogosphere such as online communities (Dholakia, Bagozzi, & Pearo, 2004). Therefore:H7 Blog perceived usefulness will positively affect political involvement.
The online relationship between blog perceived usefulness and trustfulness – in both directions – has thoroughly been stated by academic literature (Gefen, Karahanna, & Straub, 2003; Koufaris & Hampton-Sosa, 2004) and represents a fair explanation for interaction logic in blog contexts. Since marketing confidence has been related to client experience with features of the seller such as expertise and/or friendliness (Doney & Cannon, 1997; Hawes, Mast, & Swan, 1989), an equivalence between the seller and the blogger might be admitted in the blogosphere. It could be argued, therefore, that the more useful websites are perceived, the more a user might trust their contents. If we extrapolate these reasoning to the context of the political blogosphere, we could set the following hypothesis:H8 Blog perceived usefulness will positively affect blog perceived trustfulness.
Trustfulness is a construct whose boundaries still seem wide and multidisciplinary (Urban, Amyx, & Lorenzón, 2009). In online contexts, trust includes confidence in a website, believability of its content and consumer perceptions of the expectations delivered by this site (Bart, Shankar, Sultan, & Urban, 2005). Besides, in these technological contexts, trustfulness can also be developed between an individual and a computer system (Flavián, Guinalíu, & Gurrea, 2006; Lee & Turban, 2001), so users need to gain experience and check their expectations through a process of repeated visits to the website, where the loop trust-action-learning-trust occupies a central role (Urban et al., 2009). When it comes to political information and opinion, the process seems to be analogous, as users only trust several weblogs which meet their requirements. Wakefield, Stocks, and Wilder (2004) adapted McKnight, Choudhury, and Kacmar (2002) trust building model to the case of websites and concluded that structural assurance – such as the WebTrust certificate, i.e. – and perceived site quality determined trust in this type of online sites.
Prior literature does not offer, however, direct evidence of the relationship between trustfulness and involvement, but this linkage can be solidly consolidated through commitment, providing that this factor is defined in the classical terms of maximum engagement with an issue (Moorman, Zaltman, & Desphande, 1992; Morgan & Hunt, 1994), which would naturally encompass involvement itself. A good number of investigations have proved trustfulness drives commitment (Morgan & Hunt, 1994; Park, Lee, Lee, & Truex, 2012), and other studies have reached identical conclusion for the tandem involvement-commitment (Bloemer & Odekerken-Schröder, 2007; Purnasari & Yuliando, 2015). Considering these references, we posit that political blogs which credit a decent level of trustfulness do foster user interaction with their contents or with other users and, therefore, this dynamic can lead to an increase in the involvement an individual develops with political matters trough the time. Thus:H9 Blog perceived trustfulness will positively affect political involvement.
Significant academic attempts to explore the construct of involvement have been made (Krugman, 1965; Zaichkowsky, 1985) but no agreement on its definition has been reached. Involvement basically refers both to consumer concern or uncertainty and, in general terms, to opinion leadership. Personal characteristics of users, lifestyle, perception of needs and situation have been pointed out as antecedents of involvement on the Internet (Wu, 2002), and these traits seem to be fully extendable to public affairs.
In the field of politics, involvement has been associated with general political and electoral involvement (Campbell, Converse, Miller, & Stokes, 1960), and using an analogy from O’Cass's (2004) definition, this paper considers involvement as the degree to which the citizens establish political issues as a relevant part of their lives, that which is, an important, meaningful and interesting topic. However, involvement is also extendable to the scope of (political) blog and readers (Hsu & Tsou, 2011). Since consumer concern is an antecedent of involvement, political blogs seem to be excellent showcase for reducing uncertainty and ultimately, to help to address consumer behaviour. In the Internet context, weblogs provide users who may be involved with certain issues with the opportunity to interact directly by sharing information and mixing with other users. The consequences of these virtual interactions could have real effects transferring from intention to actual behaviour.
According to Inglehart's (1977) sociological theory of post-materialism, Western democratic societies are abandoning individual values for a new set of values which emphasize on autonomy and self-expression as means of reaching social and collective goals. Political blog culture therefore constitutes a new way of increasing participation of individuals in their societies by creating new communication channels with other voters, and candidates and political parties. Political science scholars initially equated political participation with voting behaviour (Becker, Dalrymple, Brossard, Scheufele, & Gunther, 2010). Participation has been studied from different approaches and, according to Nie and Verba (1975), a total of four types of participative behaviour were identified: (1) voting; (2) participation in electoral campaigns; (3) participation in community activities and (4) contacting political representatives. Later, Dalton (2002) added a new dimension, political protest, which reflects the adaptive nature of this concept to social changes. In general terms, then, citizens, actions, influence, political agents and political issues constitute the analytical framework of political activism, a concept that could be defined in terms of all the political activities which embrace a social goal (Sandoval-Almazán & Gil-García, 2014), and whose potentiality is still a matter of discussion (Hoffman et al., 2013). In contemporary democracies, exchange between system agents (e.g. political institutions, government departments, political parties, mass media, lobbies…) and the individuals has increasingly been favoured by new technological communication tools, such as e-mail, social networking websites or political blogs. The more individuals use these tools, the more activist a society may become. This paper hypothesizes that political activism arises as a natural effect of ongoing interaction where new marketing tools occupy a central role. Therefore:H10 Political involvement will positively affect political activism through the Internet.
Previous published scales were used and adapted to measure the variables of the research model. Blog perceived information quality (F1) was measured using a twelve-item scale adapted from the works of Doll and Torkzadeh (1988), Doll, Xia, and Torkzadeh (1994) and Hackbarth (2001). Blog perceived reputation (F2) was measured via a five-item scale as a result of the accumulative work of Anderson and Weitz (1992), Doney and Cannon (1997), Jarvenpaa and Leidner (1998) and Jarvenpaa, Tractinsky, Saarinen, and Vitale (1999). These last three contributions were also employed to measure blog perceived trustfulness (F5). Subjective political knowledge (F3) and political involvement (F6) were measured using the scales proposed by O’Cass (2002) for political contexts. Blog perceived usefulness (F4) was measured using the reviewed scale by Segars and Grover (1993), and adapted to the online environment by Hackbarth (2001). Finally, political activism (F7) was measured by adapting the original scale included in the European Science Foundation research (European Social Survey, 2002) into the subjects of political issues and later adapted by Norris (2005) from a twenty-two-item work into a sixteen-item list. Except for this last variable, which is dichotomic, all factors were measured with 1-to-7 point Likert scales. The scales and their items are displayed in Table 1.
Measurement model: reliability and convergent validity.
| Mean | Standard deviation | Standardized loading | Weight | t value (Bootstrapping) | |
|---|---|---|---|---|---|
| F1 – Blog perceived information quality (α=0.89) Composite reliability=0.91/AVE=0.54 | 5.50 | 1.38 | |||
| 1. This blog provides the precise information I need | 5.26 | 1.37 | 0.76** | 38.58 | |
| 2. The informational content of this blog meets my needs | 5.31 | 1.38 | 0.77** | 39.44 | |
| 3. This weblog provides web pages that seem to be just about exactly what I want | 5.48 | 1.41 | 0.71** | 31.19 | |
| 4. This weblog provides sufficient information for obtaining a political opinion | 4.91 | 1.60 | 0.78** | 52.22 | |
| 8. The information on this blog is presented in a useful format | 5.84 | 1.19 | 0.77** | 38.00 | |
| 9. The information on this weblog is clear and understandable | 5.97 | 1.12 | 0.73** | 27.98 | |
| 10. I am happy with the layout of this blog | 5.20 | 1.40 | 0.63** | 21.51 | |
| 11. I get the information I need from this blog to be able to form a political opinion | 4.80 | 1.60 | 0.78** | 49.10 | |
| 12. This blog provides up-to-date information | 5.89 | 1.16 | 0.70** | 27.09 | |
| F2 – Blog perceived reputation (α=0.87) Composite reliability=0.91/AVE=0.66 | 5.51 | 1.34 | |||
| 1. This weblog is well known | 5.41 | 1.27 | 0.76** | 31.23 | |
| 2. This blog has a good reputation in the online environment | 5.65 | 1.28 | 0.91** | 101.37 | |
| 3. This weblog has a good reputation for its political views | 4.92 | 1.55 | 0.71** | 29.08 | |
| 4. This blog has a good reputation for being honest | 5.75 | 1.36 | 0.88** | 98.14 | |
| 5. This blog is concerned about his users | 5.80 | 1.23 | 0.79** | 39.27 | |
| F3 – Subjective political knowledge (α=0.89) Composite reliability=0.92/AVE=0.75 | 5.07 | 1.39 | |||
| 1. I know a lot about politics | 5.10 | 1.28 | 0.90** | 124.55 | |
| 2. I would classify myself as an expert in political issues | 4.14 | 1.64 | 0.86** | 87.22 | |
| 3. Compared to most people. I know more about politics | 5.37 | 1.45 | 0.85** | 65.28 | |
| 4. I am knowledgeable about political issues | 5.66 | 1.17 | 0.86** | 81.79 | |
| F4 – Blog perceived usefulness (α=0.95) Composite reliability=0.96/AVE=0.84 | 4.91 | 1.59 | |||
| 1. Using this blog makes it easier to form my own opinion on political issues | 5.05 | 1.52 | 0.91** | 111.29 | |
| 2. Using this blog helps me form my own opinion on political issues more quickly than other sources | 4.74 | 1.63 | 0.89** | 90.19 | |
| 3. Using this blog gives me greater confidence to form my own opinion on political issues | 4.85 | 1.63 | 0.92** | 103.57 | |
| 4. This blog is an effective channel for forming my own opinion on political issues | 4.91 | 1.57 | 0.93** | 101.64 | |
| 5. This blog is useful for forming my own opinion on political issues | 5.91 | 0.83 | 0.94** | 151.20 | |
| F5 – Blog perceived trustfulness (α=0.90) Composite reliability=0.92/AVE=0.66 | 5.54 | 1.42 | |||
| 1. This weblog is trustworthy | 5.88 | 1.34 | 0.88** | 92.10 | |
| 2. This blog keeps a consistent editorial line | 6.19 | 1.05 | 0.78** | 31.47 | |
| 4. I find it unnecessary to be cautious when visiting this blog | 5.48 | 1.60 | 0.78** | 37.72 | |
| 5. I can trust this blog will be consistent with its editorial line | 5.92 | 1.18 | 0.76** | 26.84 | |
| 7. This blog exhibits high integrity | 5.91 | 1.31 | 0.89** | 110.32 | |
| 8. I can trust this blog to know what is right regarding political issues | 4.87 | 1.68 | 0.77** | 44.55 | |
| F6 – Political involvement (α=0.94) Composite reliability=0.95/AVE=0.76 | 5.81 | 1.22 | |||
| 1. Politics means a lot to me | 6.17 | 1.06 | 0.88** | 74.33 | |
| 2. Political issues are significant for me | 5.81 | 1.21 | 0.90** | 99.85 | |
| 3. Political issues are an important part of my life | 5.22 | 1.52 | 0.86** | 70.85 | |
| 4. Politics is personally important to me | 5.50 | 1.38 | 0.86** | 75.26 | |
| 5. I am interested in political issues | 6.16 | 1.03 | 0.89** | 83.58 | |
| 6. I am involved in politics | 6.00 | 1.13 | 0.84** | 47.66 | |
| F7 – Political activism (Formative construct) | % Yes | % No | |||
| 1. I have contacted a politician, the Government or a local government official | 53.1 | 45.3 | 0.20** | 2.69 | |
| 2. I have participated in a lawful public demonstration | 73.0 | 26.2 | 0.07 | 0.89 | |
| 3. I have boycotted the purchase of certain products | 33.7 | 66.0 | 0.13 | 1.52 | |
| 4. I have worked for an association or organization | 68.9 | 30.9 | 0.15 | 1.92 | |
| 5. I have participated in illegal protest activities | 64.2 | 33.7 | 0.19* | 2.53 | |
| 6. I have been a member of a sports club | 46.2 | 52.6 | 0.04 | 0.64 | |
| 7. I have been a member of an organization for cultural or hobby activities | 21.8 | 75.8 | 0.08 | 1.25 | |
| 8. I have been a member of a religious or church organization | 21.5 | 77.6 | 0.03 | 0.50 | |
| 9. I have been a member of an organization for humanitarian aid, human rights, minorities or immigrants | 74.5 | 24.8 | 0.06 | 0.70 | |
| 10. I have been a member of a consumer rights organization | 43.9 | 55.7 | 0.30** | 3.21 | |
| 11. I have donated money to a political organization or group | 32.9 | 66.4 | −0.07 | 1.07 | |
| 12. I have worn or displayed a political badge or sticker | 63.4 | 36.2 | −0.14 | 1.87 | |
| 13. I have signed a public petition | 45.1 | 54.0 | 0.19** | 2.65 | |
| 14. I have collaborated with a political party or group | 18.1 | 81.0 | 0.01 | 0.11 | |
| 15. I have deliberately bought certain products for political, ethical or environmental reasons | 24.7 | 74.7 | 0.03 | 0.48 | |
| 16. I have been a member of a trade union | 33.9 | 65.3 | −0.08 | 1.22 | |
| 17. I have been a member of a business or professional organization | 42.3 | 57.1 | −0.11 | 1.72 | |
| 18. I have been a member of a science, education or teachers and parents organization | 20.0 | 79.4 | −0.05 | 0.70 | |
| 19. I have been a member of an environmental protection, peace or animal rights organization | 28.7 | 70.4 | −0.04 | 0.69 | |
| 20. I have been a member of an organization for the young, the women or the retired or elderly people | 27.7 | 71.6 | 0.03 | 0.51 | |
| 21. I have been a member of a political party | 23.9 | 73.1 | 0.33** | 3.27 | |
*p<.05
**p<.01.
Data used to test the model hypotheses were collected in an online survey. Preliminary research was carried out to objectively identify the most relevant political blogs in the Spanish context and to avoid diversity in terms of voting and political traditions. Spain was chosen for proximity and also because of the increasing importance of this phenomenon in the Spanish context: according to the last available Bitacoras.com report on the Hispanic Blogosphere, there were 1.4 million active blogs and over 1 million bloggers in Spain. 10.16% of these blogs were categorized as news and current affairs (Bitacoras.com, 2011).
Weblogs were selected based on popularity: a list of leading blogs was drawn up from the top-100 political weblogs taken from the rankings of Alianzo.com and Wikio.es, two of the major social aggregators for this national context. Thus, a total of 200 sites were closely analyzed, scrutinizing their contents and avoiding duplications. No pages were discriminated against for their ideology or any other content reason, as this research only required active, Spanish language blogs providing a valid e-mail address or contact form. Based on these criteria, and after the prior screening, a number of 78 weblogs were finally selected.
Every blogger was personally contacted and asked to collaborate. An explanatory briefing on the research was sent to every participatory blog and collaboration – a specific post in which the blogger encouraged their readers to fulfil a linked questionnaire – was requested. 51.3% of the selected blogs answered the message, so a second communication was sent including the text that should be posted, the link to the questionnaire and several suggestions about the publishing date and other details. Finally, a total of 39 political blogs took part in the survey.
A questionnaire was designed using an e-survey platform. Reading any of the collaborative blogs was the only requirement for users to take part in the research. Given the nature of the data collection, all the weblogs were monitored daily and various non-bias comments were also made by the researcher on the sites to increase participation or simply to clarify study details.
SampleAfter data refinement, a total of 890 valid responses from 39 different political blogs were gathered. Basically, the sample consisted of young individuals (80.6% aged under 45), an uneven gender split (71.3% men) and high educational levels (nearly 75% with a university degree), a general profile that does not differ too much from the results shown by the last Spanish AIMC1 Internet report. Additionally, individuals showed high political involvement (92.6% voted in the last General Elections) and similar voting options to the general national population (30.1% PSOE -left-wing-, 30.3% PP -right-wing-). Further details are shown in Table 2.
Basic characteristics of research sample (N=890).
| Gender | Male | 71.3 | Work status | Employee | 54.8 |
| Female | 28.7 | Self-employed | 15.8 | ||
| Total | 100.0 | Student | 14.0 | ||
| Age (years) | <18 | 1.0 | Housewife/househusband | 2.8 | |
| 18–30 | 31.2 | Retired or pensioner | 2.9 | ||
| 31–45 | 48.4 | Unemployed | 6.7 | ||
| 46–64 | 18.2 | Other cases | 3.0 | ||
| >64 | 1.2 | Total | 100.0 | ||
| Total | 100.0 | Ideological positioning | Far-left | 4.8 | |
| Education | No qualifications | 0.1 | Left (+) | 22.1 | |
| Primary | 2.2 | Left (−) | 16.9 | ||
| Secondary or high school | 22.7 | Centre-left (−) | 10.6 | ||
| 3 years degree | 23.0 | Centre-left (+) | 10.3 | ||
| 5 years degree | 35.4 | Centre-right (+) | 7.7 | ||
| Graduate studies | 16.6 | Centre-right (−) | 10.1 | ||
| Total | 100.0 | Right (−) | 10.7 | ||
| Province of residence | Madrid | 29.0 | Right (+) | 6.0 | |
| Barcelona | 8.9 | Far-right | 0.8 | ||
| Valencia | 7.7 | Total | 100.0 | ||
| Rest of Spain | 54.4 | Political party voted in the last General Elections | PSOE | 30.1 | |
| Total | 100.0 | PP | 30.3 | ||
| Weblog from which the questionnaire was linked | Malaprensa | 23.1 | IU | 14.1 | |
| Contando estrelas | 14.0 | UPD | 7.7 | ||
| Wonkapistas | 7.9 | Rest | 4.6 | ||
| Caballero ZP | 5.1 | Blank vote | 6.9 | ||
| Quien Mucho Abarca | 5.0 | Did not vote | 6.3 | ||
| Reflexiones progresistas | 4.9 | Total | 100.0 |
The empirical model was estimated using Partial Least Squares (PLS), a technique which is suitable for newly developed research models and especially useful when formative constructs are present (Hair, Hult, Ringle, & Sarstedt, 2013; Wold, 1979), as in this case. Standard indicators of reliability and convergent and discriminant validity were used to validate results. All factors exhibited Cronbach's Alpha over 0.70 (Churchill, 1979) and composite reliability exceeding 0.7 (Fornell and Larcker, 1981). All items loadings were significant (bootstrap estimation with 5000 samples of the original sample size). Results are shown in Tables 2 and 3.
Measurement properties: discriminant validity.
| 1 | 2 | 3 | 4 | 5 | 6 | |
|---|---|---|---|---|---|---|
| 1. Political involvement | 0.87 | 0.68 | 0.20 | 0.17 | 0.16 | 0.26 |
| 2. Subjective political knowledge | 0.62 | 0.87 | 0.15 | 0.13 | 0.11 | 0.23 |
| 3. Blog perceived information quality | 0.18 | 0.13 | 0.74 | 0.78 | 0.82 | 0.78 |
| 4. Blog perceived reputation | 0.15 | 0.12 | 0.70 | 0.81 | 0.88 | 0.56 |
| 5. Blog perceived trustfulness | 0.14 | 0.10 | 0.73 | 0.80 | 0.81 | 0.59 |
| 6. Blog perceived usefulness | 0.25 | 0.22 | 0.72 | 0.51 | 0.54 | 0.92 |
Note: Values on the diagonal are the square root of the Average Variance Extracted (AVE) of each factor. Values below the diagonal are correlations between factors and values above the diagonal are the HTMT ratios.
Discriminant validity was assessed using commonly used (Barrera, Navarro, & Peris, 2015) Fornell and Larcker's (1981) criterion and the recently proposed Henseler, Ringle, and Sarstedt (2015) criterion (Table 3). The first criterion shows discriminant validity tensions between quality and trustfulness, but meeting the criterion. Supporting this conclusion, the HTMT ratio are always lower than .90 proving this tension not to be critical according to Gold, Malhotra, and Segars (2001) and Teo, Srivastava, and Jiang (2008) benchmark. HTMT criterion has demonstrated a superior performance in detecting discriminant validity threats.
Finally, in order to assess the predictive relevance of the model, Stone-Geisser's Q2 statistics were also estimated via blindfolding. According to Stone (1974), the criterion is adequate if Q2 is positive. Results are displayed in Fig. 3.
Results
The estimation of the model validated eight out of the ten proposed research hypotheses, as Fig. 3 shows. Blog perceived information quality seems to be a key variable for examining blog structure, since it directly influences other components of these websites, such as usefulness (H1), trustfulness (H2) and reputation (H3), the three core dimensions of a political blog as described in the proposed model. These results are consistent with Jeong and Lambert's (2001) findings. In short, the better perceived the information in a blog is, the better the whole blog becomes when generating further functional and behavioural effects.
Additionally, usefulness can be defined as the central comprehensive variable that links the online core environment of a weblog – this is, the elements that give the weblog its self-entity – to the social and personal variables which are expected to be observed only in offline contexts. Thus, blog perceived usefulness influences subjective political knowledge (H6) and political involvement (H7). Blog perceived reputation, in turn, influences blog perceived trustfulness (H4). These results reveal the existence of an effective set of interrelationships between affective factors such as information quality, usefulness, reputation and trustfulness, which play an important role in blog culture and political effects, which emerge as mere consequences of blog interaction.
Finally, the model test successfully proves the political connection, considering the influence of subjective political knowledge on political involvement (H5) and also the influence of this latter factor on political activism (H10), an association which depicts a real effect from online interaction to offline behaviour regarding political issues where subjective knowledge is a major trigger. In contrast, the model did not validate the link between blog perceived trustfulness and political involvement (H9). This lack of association is enhanced by the non-existent influence of blog perceived usefulness on trustfulness (H8), a factor which emerges almost as an isolated variable in the model whose boundaries may require justification and further research.
DiscussionThis paper examined the role political blogs play in encouraging their users’ political behaviour by proposing a model that merges marketing and technological variables from a wide theoretical approach. Since blog culture is becoming increasingly important for a good number of topics and blogs have arisen as more influential sources of information than traditional media (Burns, 2008; Johnson & Kaye, 2004), there needs to be greater understanding of the virtual dynamics of these emerging and interactive tools and how they influence behaviours in conventional, and offline environments, mostly when living in ongoing contexts of permanent campaign (Blumenthal, 1980; Dulio & Towner, 2008). Political blogs can exert the same influential role, if not more, than blogs on other topics (i.e. fashion, consumer goods), but also provide an additional cognitive bias which connects online environments with the offline, real world. This connection may be one of the most interesting findings of this research. Thus, our model shows there is a clear connection between blog usage and political involvement and activism. In general terms, political information quality in blogs increases the reputation of the sites and their audience and, consequently, reinforces their trustfulness and usefulness. This set of relationships is consistent with previous research on marketing field (Park & Lee, 2009) but is new in political studies, and helps to identify several key factors related to blog effectiveness which suggest that, regardless of an individual's political knowledge, the more useful a blog is perceived, the more influence it will be able to exert on political involvement.
This latter statement may be applied not only to frequent blog readers, but also to those who occasionally search for political information on the Internet. Thus, institutional blogs linking to relevant topics or to other blogs could also constitute a source of political involvement for groups of interest. It remains a thrilling challenge for political marketers to determine optimal strategies which can definitely cope with traditional ways of political communication, considering the growing importance of interactive tools in the Web 2.0 era (Mossberger, Tolbert, & Stansbury, 2003; Zheng & Zheng, 2014) and the large amount of resources politicians and political parties use to spread their messages or simply to increase their degree of noticeability. Indeed, the two-way communication that blogs have enhanced offers a huge potential: communication within networks is a significant predictor of web-based forms of political engagement (Valenzuela, Kim, & Gil de Zúñiga, 2012). Political marketing through blogs can bring politicians closer to citizens, as one-way communication is replaced by direct, feedback-friendly messages and even personal videos which could elicit a more reliable and deeper relationship between candidates and citizens by sharing information, knowledge and opinions over time. Candidates are considered political leaders and this function is closely related to the purpose of being communicators (Jackson, 2011). Therefore, weblogs are a key opportunity for changing politicians’ traditional image and for involving new groups of people who might be increasingly interested in certain political issues and demand real, specific solutions from their representatives.
From a theoretical perspective, the factor that best explains the effect on political involvement in our model is subjective knowledge, which arises as a central variable. This close relationship had been highlighted by previous marketing works (O’Cass, 2002; Zaichkowsky, 1985), but not in the same direction as this research has established, so this could be thought as an additional, theoretical asset of this paper. Hence, as blog users increase their political knowledge, so does their involvement with political issues. Considering the model, this relevant association becomes feasible when blogs are perceived as useful, which depends on content quality, a statement which also fits the existing research (DeLone & McLean, 1992). Consequently, individuals who have been properly informed or engaged by political weblog contents not only increase their theoretical involvement, but also are able to transfer these premises to an offline environment. This conclusion may be taken as one of the major practical contributions of the present investigation. Collaborating with political groups, signing political petitions, belonging to a consumer rights organization or contacting political representatives symbolize several of the most common activist behaviours triggered by blogs. For political agents, identifying e-fluentials whose blogs could mediate and support certain political issues by boosting shared interests and creating long-term relationships amongst users might become a useful marketing strategy. Voters could then also be reached by an alternative and trustworthy channel which is seen as a more objective information source and might be suitable for the new increasingly untrusting audiences and could even appeal to transfer its influence to offline contexts.
On balance, this study opens the door to a new approach to online frameworks such as social media where political marketing still has much progress to make. The fact that the research is focused not on the voting period, but on the ordinary period of the political cycle, enhances the adoption of a relationship marketing approach applied to new technologies and yields a good number of implications. The main theoretical contribution, however, relies on setting a model which integrates inputs from different approaches in marketing (relational paradigm, TRA, models of consumer behaviour…) where TAM plays a prominent role. This is a coherent and innovative guidance which enables the development of other multidisciplinary studies in the area of marketing. At the same time, the proposed blog interaction model constitutes a truly effective channel for a more participatory democracy where interaction and marketing equity are enhanced, especially for social groups who have been underrepresented by traditional mechanisms (Quintelier & Vissers, 2008). This effect, previously highlighted by Norris's (2001) mobilization thesis, emerges as a great opportunity for the minorities trying to communicate specific political ideas, a possibility which has been clearly narrowed down by traditional political marketing and which requires further research.
Limitations and future lines of researchConsidering the specific online profile of this study, several limitations have to be noted. First, broad generalizations about citizens’ political behaviour are unsuitable, due to the fact that this research has only focused on an emergent but still a small part of the voters, those who usually read political blogs and are keenly aware of its interactive possibilities.
Additionally, it might be argued that these people would not be reading political blogs if they were not previously involved with political issues, an inbred cycle that would compromise the expansion of the theoretical model itself. Besides, cyberbalkanization (Putnam, 2000) of blogs – consisting of the radical isolation of groups of weblogs with diverging interests – cannot be omitted either, as it might constitute a serious barrier for the universalization of blog interaction. These limitations, however, do not seem to have a significant influence on the results of this study, as it is intended to shed light on the blog dynamics that finally foster political activism from an extensive perspective.
The very complexity of blog interaction enables new lines of research that might strengthen the model proposed here or even expand it to new scenarios. For example, as blog usefulness has been identified as a key factor in this study, defining what this variable consists of might remain a challenge for future research. The same logic could be applied to information quality, given the continuous technological improvements in blogs and their integration into the emerging and more comprehensive environment of social networks.
Further study of blog perceived trustfulness also constitutes an interesting contribution to the theoretical framework, as its connection to involvement has not been properly explained in spite of being a long-term and relevant variable in other website contexts. Perhaps, the opposite relation where involvement drives trustfulness could also be explored. Future studies could also attempt to integrate consolidated variables into political marketing such as social presence, online opinion leadership or eWOM. Methodologically, netnography (Kozinets, 2002) might be considered a suitable and interesting technique for analysing the dynamics in these online communities given its simplicity, quickness and low cost. Finally, validating this model in different blog contexts might represent a fair extension of the study to more conventional marketing environments where information about product or services also constitutes a relevant part of the decision process.
Conflict of interestNone declared.
