Mass collaboration and reading citizen journalism.
Shu, Wesley ; Lin ; Chia-Sheng, Hota 等
1. Introduction
The Internet now can be described as user-centric. Many platforms
allow users to interact, share, and cooperate. Citizen journalism has
merged under such scenario. Citizen journalism is a type of grassroots
reporting. With the Web 2.0, it shows more importance than before.
In many countries where journalism is not well developed, news
items are often untrusted. Newspapers are full of paparazzi reports and
gossips. Due to intense competition, media tend to focus on shocking
news instead of quality. Media need to take care of mass market and then
small groups' needs are usually ignored.
But the advent of Web 2.0 may change this. The term Web 2.0 is
associated with web applications that facilitate participatory
information sharing, interoperability, user-centered design, and
collaboration on the World Wide Web (Wikipedia, 2011). One product of
Web 2.0 is blogs. Following the concept, using blog has become all the
rage. The trend of using blog for recording and sharing has gradually
formed. Therefore, citizen journalism has been brought to a new level
(Rosen, 1999).
Before Web 2.0, media content was solely created by professionals.
Now, since different platforms offer users opportunity to make
contributions online, any users can write news items. In earlier stage
of Web 2.0, people simply used blogs to write their news reports. Later,
media companies collect these reports and become
'one-shop-for-all' news center. The typical example is
Huffington Post. Another type of citizen journalism, which is the focus
of our research, is an existing news company or a voluntary organization
creating a platform and allows everyone to contribute news contents into
that platform. The contributors are called citizen journalists. Examples
are Ohmynews in Korea, Ground Report in New York, Global Voices
sponsored by Harvard Law School, BBC News Online, AgoraVOX in France,
Nowpublic in Canada, Peopo, NEnews, and Newserr.com in Taiwan.
The success of citizen journalism relies on the journalists and the
readers. More citizen journalists create more content which may attract
more readers. When there are more readers and readers made comments or
form communities, there will be more readers and more citizen
journalists will contribute. This self-reinforcement makes citizen
journalism successful.
The reciprocity of authors and readers can be explained by the
concept of two-sided markets (Eisenmann, Parker, & Van Alstyne,
2006; Parker & Van Alstyne, 2005). A two-sided market has two groups
and each forms its own network. The two groups interact through a
platform. The utility of a participant in a network is determined by the
size of the other network. The SNS users and the SNS games are an
example. When Facebook collects users and becomes a giant social
network, developers will come to provide their games. It shows that as
long as one-side network is formed, the other will merge as long as the
platform works well. The evolution from blogs to citizen journalism can
be viewed as the evolution to two-sided markets where the platform plays
a crucial role. There are numerous blogs on the Internet but most blogs
are seldom visited. Readers usually do not care about what blogs they
read news items or articles from. This can be a problem for bother
writers and readers as writers do not know who their readers really are
and readers may need to spend time searching to find the right articles
they need. When a platform appears between the writers and the readers,
this problem can be solved and the writer-reader relationship can be
consolidated. A citizen news platform for citizen journalists thus is
similar to a social network site such as Facebook for game developers.
In such a two-sided market, the platform can get writers to contribute
articles and the issue is how to get readers involved. This is our
research question.
For example, the largest citizen news website OhMyNews.com issues
150 articles daily. There are 62,000 citizen journalists and 70 formal
employees (Ohmynews, 2011). These show that they have no problem to
establish the platform and form one of the participants of the two-sided
market--the journalists. But, they need to form reader network too but
they have no control of readers. One way to form the reader network is
to know the factors which cause the intention and the behavior of the
citizen news readers and then reinforce these factors. With more readers
involved, citizen news platforms will increase their values. Our paper
thus is to investigate the factors for readers to accept citizen
journalism created by mass collaboration.
Another reason this issue requires further investigation is the
profit decrease of traditional news media. The overall advertisement
income of news media in 2009 reduced 26% compared to the previous year
and the average loss of news industry was 43%. Since then the situation
does not improve much (State of News Media 2010, 2010). Some news
companies seek the solutions by providing more shocking news with undue
methods. The recent reports about Murdoch's hacking the phones of
private citizens are one example (Wikipedia, 2012). We believe a better
way to face the profit reduction in news industry is not to create more
paparazzi reports but to catch up the new wave of news production--mass
collaboration.
The data were collected in Taiwan. The first significance of
Taiwanese data is that Taiwan is the only Chinese society with absolute
freedom of publication. Although Taiwan cannot represent China, studying
Taiwan's citizen journalism today is studying the current status of
citizen journalism in Chinese society. Second, Taiwan's news
environment may foster citizen journalism. According to Reporters
without Borders, Taiwan's "Press Freedom Index" was
ranked 32 in 2007 and 36 in 2008(Press Freedom, 2007), but readers are
concerned with the quality of current news reports. There is even a
Wikipedia item to note Taiwan's media chaos and foundations were
established to prevent such chaos (Wikipedia, 2009). Citizen journalism
may be viewed as reaction to this phenomenon. In addition, since citizen
journalism is mass collaborative, it will increase citizens' social
participation.
This paper is divided into the following sections. Section 2
introduces the evolution of citizen journalism and its recent
development. Section 3 is the research framework and the research
hypotheses. Section 4 describes the survey and data analysis. Section 0
is the conclusion.
2. Definition of Citizen Journalism and its
Development
Citizen journalism can trace its origin to public or civic
journalism in the 80s in the US, after journalists themselves began to
question the predictability of their coverage of the 1988 U.S.
presidential election. Facing the vicious competition, newspapers were
full of biased reports and reports of scandals. People's
dissatisfaction was reflected the lowest voter turnout in history. To
save the crisis, Washington Post journalist David Broder and Journalism
scholar Jay Rosen proposed public or civic journalism (Cappella &
Jamieson, 1996; Dzur, 2002; Merritt, 1998; Rosen & Merritt, 1994;
Rosen, 1999).
Citizen journalism is also called participatory journalism,
open-source journalism, public journalism, or grassroots reporting. In
citizen journalism, news content is no longer provided by traditional
professionals but citizen reports. Anyone can be a citizen reporter
(Bowman & Willis, 2003; Lasica, 2003). The Internet gives citizen
journalism a new phase. First, traditional news readers decreased but
online news readers increased. In 2007, 37% readers are online readers,
increased 7% from two years ago (Johnson & Wiedenbeck, 2009). In
2010, there were 34% American got the news from the Internet, increased
5% from 2009. It is the only increase among TV, Radio, Newspaper, and
Internet (State of News Media 2011, 2011). Some media have thought about
changing. Christian Science Monitor has stopped the paper edition in
2008. New York Times is considering following (Huffington Post, 2010).
Due to the emergence of Web 2.0, citizen journalist websites are
increasing and many professionals joined citizen journalism. For
example, the number reached 1,500 in 2007 in the US (Johnson &
Wiedenbeck, 2009). Hellweg (2005) called this new type of citizen
journalism collaborative citizen journalism (CCJ) for its nature of mass
collaboration. He proposed that a new type of journalism is rising and
it can improve the transparency of news media. Through the collaborative
wisdom, this new type of news provides original news reports ignored by
traditional media, and compensates their distorted reports. Social
network websites then distribute these reports to the Internet users.
3. Research Framework and Hypotheses
To understand "the factors influencing readers accepting
mass-collaborative citizen journalism", we chose Unified Theory of
Acceptance and Use of Technology (UTAUT) (V. Venkatesh, Morris, Davis,
& Davis, 2003). The reasons are: First, citizen journalism is a kind
of human computer interface (HCI) and HCI is the research domain of
UTAUT (Davis, 1989). Second, UTUAT studies web-based systems and citizen
journalism is a web-based system. Third, UTAUT has been applied and
tested with high explanatory power ([R.sup.2])(AbuShanab, Pearson, &
Setterstrom, 2010; Al-Gahtani, Hubona, & Wang, 2007; Chan et al.,
2010; Chen, Wu, & Yang, 2008; Chiu & Wang, 2008; Im, Hong, &
Kang, 2011; Lu, Yu, & Liu, 2009; Yeow & Loo, 2009; Yuen, Yeow,
Lim, & Saylani, 2010). The purpose of UTAUT is to improve technology
acceptance and use which is the same as that of our study.
But UTAUT is not fully applicable to our study. First, when UTAUT
was developed, the Internet is not participatory. Users accepted the
content created by websites instead of creating their own content. The
Internet today is participatory. Users can contribute content. Thus, we
should ass "mass collaboration" into UTAUT. In addition,
similar to online reviews, user-created content may get more credibility
among the Internet users. This credibility can be enhanced due to the
fact that hyperlinks can allow other readers to get the content
creators' online personal profiles.
Second, since UTAUT is a unified model, there can be some
constructs not applicable to citizen journalism. According to Venkatesh,
parsimony is an important criterion to set up a research model (V.
Venkatesh & Morris, 2000). Thus, we may need to delete some
constructs. Our model is illustrated in Figure 1.
[FIGURE 1 OMITTED]
There are several features in this model:
1. We added mass collaboration as one construct. Mass-collaborative
citizen journalism is defined as "news content which is created by
professional or non-professional without pre-publication regulation and
edited by volunteers. The fundamental nature of citizen journalism is
news items are not monopolized by news media and everyone can
contribute. The reason of its popularity is to provide different news
channels and to save the overall news industry. According to Anthony et
al.(2007), the quality of mass-collaborative content is no less than the
content created by professionals.
2. The original moderators in UTAUT (gender, age, experience, and
voluntariness of use) were deleted in our model: First, our research is
not focused on these variables' influence. Second, the ANOVA showed
that these variables put no effects on any constructs (See Table 6,
Table 7 and Table 8).
3.
4.
5. Third, voluntariness is not applicable since all contributors to
citizen journalism are voluntary.
6. We added two moderators--role and degree of involvement:
a. Role: Since writers, which are also readers, and pure readers
are two roles on citizen journalist platforms, we want to see if
different roles lead to different consequences. We assume that writers
may be more active in citizen journalist activities and this may show
different impact on the constructs.
b. Degree of Involvement: Jackson et al.(1997) and Swanson
(Swanson, 1974) stated that user involvement may affect the use of
information system. Their argument may hold true since citizen
journalist websites are also information systems. Zaichkowsky (1985)
defined involvement as "a person's needs or values and a
general level of interest in or concern about an issue without reference
to a specific position."
During the time we did the survey, reading Citizen journalism was
new. To know its sustainability, we want to know if those already showed
their involvement has interest in continuing reading citizen journalsm.
Thus, we divided the respondents into two groups--higher involvement and
lower involvement and check if higher involvement may generate higher
interest.
7. The other UTAUT variables are applied to this study with some
modifications:
a. Performance expectancy: It is the expectation of the citizen
journalist readers that citizen journalism may provide utilities other
news media cannot. Blumler (1979) argued that readers will pursue the
media once the media contain helpful information or fit into their
preference.
b. Effort expectancy: It means readers believe the operation of
citizen journalism platform is easy. Different levels of effort readers
need may have different impact on their intention to use citizen
journalism (Smith, 1997).
c. Social influence: Readers' intention to use citizen
journalism is affected by their relatives, friends, or other people. The
case of the public journalism movement originated in 1988 affecting
readers' attitude toward traditional and new media is an example
(Brown, Broderick, & Lee, 2007; Cappella & Jamieson, 1996;
Merritt, 1998; Rosen & Merritt, 1994; Rosen & Merritt, 1994).
d. Facilitating condition: The reason people want to read citizen
journalism is the accessibility to citizen journalist platform the
Internet provides and the convenience the platform offers. These include
the news items a citizen can publicize to the citizen journalist
websites (Lasica, 2003).
8. Two endogenous variables are behavior intention and use
behavior. Their definitions are "the intention to use citizen
journalism based on the readers' subjective judgment" and
"the actual behavior the readers use the citizen journalism
platforms."
Table 1 shows the measures. All measures are modified from
Venkatesh (2003) with experts' opinions and consideration of the
practical situation in citizen journalism.
Based on the above discussions, we have the following hypotheses:
H1. Performance expectancy positively affect the intention to use
citizen journalist websites.
H2. Effort expectancy positively affects the intention to use
citizen journalist websites.
H3. Social influence positively affects the intention to use
citizen journalist websites.
H4. Mass collaboration positively affects the intention to use
citizen journalist websites.
H5. Facilitating condition positively affects the intention to use
citizen journalist websites.
H6. Intention to use citizen journalist websites positively affects
the behavior to use them.
H7. The influence of performance expectancy on intention to use
citizen journalist websites is moderated by roles, such that the effect
is stronger for writers.
H8. The influence of effort expectancy on intention to use citizen
journalist websites is moderated by roles, such that the effect is
stronger for writers.
H9. The influence of social influence on intention to use citizen
journalist websites is moderated by roles, such that the effect is
stronger for writers.
H10. The influence of mass collaboration on intention to use
citizen journalist websites is moderated by roles, such that the effect
is stronger for writers.
H11. The influence of facilitating conditions on intention to use
citizen journalist websites is moderated by roles, such that the effect
is stronger for writers.
H12. The influence of performance expectancy on intention to use
citizen journalist websites is moderated by degrees of involvement, such
that the effect is stronger for users with higher involvement.
H13. The influence of effort expectancy on intention to use citizen
journalist websites is moderated by degrees of involvement, such that
the effect is stronger for users with higher involvement.
H14. The influence of social influence on intention to use citizen
journalist websites is moderated by degrees of involvement, such that
the effect is stronger for users with higher involvement.
H15. The influence of mass collaboration on intention to use
citizen journalist websites is moderated by degrees of involvement, such
that the effect is stronger for users with higher involvement.
H16. The influence of facilitating conditions on intention to use
citizen journalist websites is moderated by degrees of involvement, such
that the effect is stronger for users with higher involvement.
4. Survey and data analysis
The measures in Table 2 then are put into the questionnaire. The
questionnaire has eight parts. The first part consists of items used to
test for demographic differences. The second to sixth parts are for
evaluating the factors affecting the attitude of reading citizen
journalism. The seventh and eighth parts are for testing the actual
attitude of reading citizen journalism. 7-point Likert scales are used
in Part 2 to part 7.
4.1. Pre-test
The first step in developing the pre-test was to invite 10 scholars
(two MIS professors, three Ph.D. candidates, and five professionals with
domain knowledge and extensive experience with citizen journalism) to
examine the above preliminary version of the questionnaire for internal
validity. All 10 judges agreed that the questionnaire "can measure
what it is supposed to measure" and that "all dimensions are
essential to the evaluation of citizen journalsm." Thus, face
validity and content validity were achieved.
Then the questionnaire was put on My3Q
(http://www.my3q.com/home2/274/sophyio/citizen.pht ml) and given to 42
people with experience in reading citizen journalism. 3respondents
submitted invalid questionnaires. The sampling period was May 1 through
May 6, 2010.
Cronbach's [alpha] was used to assess the reliability of the
scales composing the questionnaire. Guilford suggested that an a value
greater than 0.7 means that the reliability is adequate (Guilford,
1965). As shown in
Table 3, the [alpha] value for the facilitating conditions scale
was lower than 0.7; to remedy the situation, I discarded items FC03.
To determine if the scales had adequate validity and were suitable
for factor analysis, I applied the Kaiser-Meyer-Olkin (KMO) measure of
sampling adequacy (Kaiser, 1974). A KMO score greater than 0.69 shows
that the items have low partial correlations with the total scale of
which they are a part. The KMO value was 0.685.
We then conducted an exploratory factor analysis. Table 4 shows the
factor loadings after orthogonal rotation.
The table shows that effort expectancy and facilitating condition
should be combined. This shows that these two factors are the same
variable and thus we deleted FC.
4.2. Main test
The same questionnaire on the Internet was used for the main test.
Tan and Teo (2000) believed that online survey has some merits that
traditional surveys do not. The critical benefit is our study is focused
on online users. The sampling period was two weeks. 422 responses were
collected with 45invalid.
1. Demographic variables
Table 5 gives the demographic data.
I then conducted a series of one-way ANOVAs to test for
interactions between demographic variables and psychological variables.
The criterion for statistical significance was p <.05, two-tailed.
The results are presented in through Table 8. The overall results show
that the demographic factors did not affect the results.
2. Reliability
The procedure to test reliability is the same as in pre-test. The
Cronbach's a of the social influence scale was 0.609. After
removing SI01, it reached 0.732 as shown in Table 9.
3. Construct validity
To determine if the scales had adequate validity and were suitable
for factor analysis, we applied the Kaiser-Meyer-Olkin (KMO) measure of
sampling adequacy (Kaiser, 1974). A KMO score greater than 0.8 shows
that the items have low partial correlations with the total scale of
which they are a part. The KMO value was 0.952. We also applied
Bartlett's sphericity test, which was significant (p < 0.05).
Thus, the scales are factorable by both criteria.
The confirmatory factor analysis gave five psychological dimensions
corresponding to the actual structure of the questionnaire (see Table
10). This means that the constructs are valid.
We then conducted convergent and discriminant validity tests.
Convergent validity is the degree to which an operation is similar to
(converges on) other theoretically similar operations. According to
Fornell and Larcker (1981), convergent validity is achieved when all the
standardized factor loadings exceed 0.5 in the same dimension. The
factor loadings of PE04, PC03, and BI1 are lower than 0.5, so they are
removed from the main test.
Discriminant validity is the degree to which the operationalization
(construct) diverges from other operationalizations that it
theoretically should not be similar to. A construct is said to have good
discriminant validity when the factor loading of that construct and
another construct is low, usually lower than 0.5. All measures show good
discriminant validity.
4. Model fit
We next sought to determine whether our model is the best of the
available choices--the question of model fit. Absolute fit is a measure
of how well the model fits the sample data. It is measured by four
indices: [chi square]/df, GFI (Goodness of Fit Index), AGFI (Adjusted
Goodness of Fit Index), and RMSEA (Root Mean Square Error of
Approximation).
Incremental fit is a measure of how well the test model fits the
data compared to the baseline (null) model. It is measured by two
indices: IFI (Incremental Fit Index) and CFI (Comparative Fit Index).
Parsimonious fit measures if the model applies a minimum number of
scales and is free from overfitting. Two indices were used: PGFI
(Parsimony Goodness-of-Fit Index) and PNFI (Parsimony Normed Fit Index).
As shown in
Table 11, the results meet the minimum acceptable levels for all
the model fitting measures.
5. Hypothesis tests
Finally, we used maximum likelihood estimation to test the
hypotheses listed in Section 3. Figure 2 and Table 12 show the
structural equation model for the path analysis and the results of the
hypothesis test.
Figure 2 Path analysis for direct variables
[ILLUSTRATION OMITTED]
For the hypotheses about moderating variables, we also use maximum
likelihood estimation. Moderating variables are arranged as dummy
variables in the estimation. The results are listed Table 13.
5. Conclusion
Citizen journalist websites can be regarded as a two-sided market
composed by writers, readers, and the platforms. For the reason
mentioned in Section 1, we chose readers as our research topic. From
Figure 2 and Table 12, we can tell all dimensions except social
influence have positive impact on intention to read citizen journalism.
In addition to traditional dimensions, our study shows that mass
collaboration increase users' intention to read. Thus, news
companies need to put effort on developing mass collaboration and even
design new business models based on mass collaboration. For example,
companies can design a 'top contributor' system similar to the
one of Yahoo! Answers and allow user ratings and comments to encourage
citizen journalists and to create a self-selection mechanism.
Table 13 shows that the influence of effort expectancy and mass
collaboration on intention to read citizen journalism is positively
moderated by degrees of involvement (H13 and H15). Their influence is
also moderated by roles such that it is stronger for writers (H8 and
H10). It is not difficult to understand the moderating effects of roles
and degrees of involvement on effort expectancy. When a reader is also a
writer, she would have better system knowledge over the websites and
perceive it when the system is easy to use. It is easier for her
transforming such perception into the intention to use the system.
It's similar for degrees of involvement.
For mass collaboration, roles and degrees of involvement posit
similar effects: double roles and higher involvement allow users easier
to translate their perception of mass collaboration into the intention
to use. Double roles also make it easier to translate the perception of
citizen journalist sites' performance into the behavior intention
(H7). However, similar conclusion does not hold true for involvement
(H12). Roles and involvement do not show moderating effects on social
influence or facilitating condition (H9 and H14). Combined with H3, This
reveals that social influence does not affect users' intention to
read citizen journalism, this impact still does not exist even when
users take double roles or get more involved.
One conclusion is counter intuition: intention to read citizen
journalism does not necessarily lead to the actual use. This can be
explained by referencing to Table 5: 88.6% of the users read citizen
journalism less than three hours each week. This may show that citizen
journalism is not popular in Taiwan or readers do not realize they are
reading citizen journalism. Combining the high scores of that mass
collaboration, performance expectancy, and effort expectancy (see Table
6 through Table 8), we can tell that users have good opinions about
citizen journalism but they do not have much news resource from limited
citizen journalist websites. Thus, we should increase the number of
citizen journalist websites. It is also possible that readers do not
know the news they are reading are citizen journalism. In this case,
news media need to inform readers well because literature shows that
readers are disappointed by traditional media and our research shows
that they want news to be collaborative work.
From the viewpoint of management, our research is also important.
One feature of Web 2.0 is democratization of production tools. The
Internet became production line and consumers have become producers with
the production tools. For example, the Connect + Develop project by
P&G and the world share patents. Canyon Bike allows consumers design
their own bikes. Starbucks uses Internet to survey customers for new
flavors. Nestle develops a toolkit for customers to make new coffee
products. Lego develops new software to allow players to design new
Legos. These new business models which employ Web 2.0 have drawn
attention in business world, and news media also notice this trend. With
Web 2.0 models, they can cover many topics not covered by traditional
models, and create a platform for citizen journalists to write community
news. More importantly, applying Web 2.0 models may increase long-tail
effect (Erik Brynjolfsson, Yu Jeffrey Hu, Michael D.Smith, 2006):
Contrary to traditional business models, where companies who want to
satisfy mass market cannot satisfy local or elite market, the Internet
allow any kind of products listed online without space or time
restrictions and thus small groups can still find the right products
they need. Similarly, the current news media which follow traditional
models cannot provide news pertinent to specific groups. This can be
changed with citizen journalism. Thus, we believe citizen journalism is
not only a topic in journalism but also one in business.
Our research shows some limitations. Although we studied the
reasons to read citizen news, there are reasons for readers not to read
it. Knowledge is one of such reasons (Kaufhold, Valenzuela, & de
Zuniga, 2010) concluded that those with higher political knowledge tend
to consume professional journalism instead of user-generated journalism.
Does it imply that when users realize citizen journalism does not bring
political knowledge, they tend not to read it? If so, we may need to
include such negative force into our model. A push-pull-mooring
framework can be helpful to identify both positive and negative
antecedents (Cheng, Yang, & Lim, 2009; Fu, 2011; Zhang, Cheung, Lee,
& Chen, 2008). From online news to citizen news, the change is not
only the appearance. The content, authorship, writing style, coverage,
participation and so on are all different. Our next study will be
focused on this paradigm shift.
References:
AbuShanab, E., Pearson, J., & Setterstrom, A. (2010). Internet
banking and customers' acceptance in jordan: The unified
model's perspective. Communications of the Association for
Information Systems, 26, 67.
Al-Gahtani, S., Hubona, G., & Wang, J. (2007). Information
technology (IT) in saudi arabia: Culture and the acceptance and use of
IT. Information & Management, 44(8), 681.
Anthony, D., Smith, S. W., & Williamson, T. (2007). The quality
of open source production: zealots and good samaritans in the case of
wikipedia Retrieved from http://www.cs.dartmouth.edu/
reports/TR2007-606.pdf
Blumler, J. G. (1979). The role of theory in uses and
gratifications studies. Communication Research, 6(1), 9.
Bowman, S., & Willis, C. (2003). In Lasica J. D. (Ed.), We
media: How audiences are shaping the future of news and information The
Media Center at American Press Institute. Retrieved from
http://www.hypergene.net/wemedia/download/we_media.pdf
Brown, J., Broderick, A. J., & Lee, N. (2007). Word of mouth
communication within online communities: Conceptualizing the online
social network. Journal of Interactive Marketing, 21(3), 2-20.
Cappella, J. N., & Jamieson, K. H. (1996). News frames,
political cynicism, and media cynicism. The Annals of the American
Academy of Political and Social Science,, 71-84.
Chan, F., Thong, J., Venkatesh, V., Brown, S., Hu, P., & Tam,
K. (2010). Modeling citizen satisfaction with mandatory adoption of an
E-government technology. Journal of the Association for Information
Systems, 11(10), 519.
Chen, C., Wu, J., & Yang, S. (2008). Accelerating the use of
weblogs as an alternative method to deliver case-based learning.
International Journal on ELearning, 7(2), 331.
Cheng, Z., Yang, Y., & Lim, J. (2009). Cyber migration: An
empirical investigation on factors that affect users' switch
intentions in social networking sites. Proceedings of the 42nd Hawaii
International Conference on System Sciences, Hawaii.
Chiu, C., & Wang, E. (2008). Understanding web-based learning
continuance intention: The role of subjective task value. Information
& Management, 45(3), 194.
Davis, F. D. (1989). Perceived usefulness, perceived ease of use,
and user acceptance of information technology. MIS Quarterly,, 319-340.
Dzur, A. W. (2002). Public journalism and deliberative democracy.
Polity,
Eisenmann, T., Parker, G., & Van Alstyne, M. W. (2006).
Strategies for two-sided markets. Harvard Business Review, 84(10), 92.
Erik Brynjolfsson, Yu Jeffrey Hu, Michael D.Smith. (2006). From
niches to riches: Anatomy of the long tail. MIT Sloan Management Review,
47(4), 67.
Fornell, C., & Larcker, D. F. (1981). Evaluating structural
equation models with unobservable variables and measurement error.
Journal of Marketing Research, (18), 39-50.
Fu, J. (2011). Understanding career commitment of IT professionals:
Perspectives of push-pull-mooring framework and investment model.
International Journal of Information Management, 31(3), 279.
Guilford, J. P. (1965). Fundamental statistics in psychology and
education (4 ed ed.). N.Y.: McGraw-Hill.
Hellweg, E. (2005). Is 'CraigsNews' coming soon?.
Retrieved 04/03, 2011, from http://www. technol
ogyReview.com/communications/14459/?a=f
Huffington Post. (2010). Arthur sulzberger: 'we will stop
printing the new york times sometime in the future'. Retrieved
04/03, 2011, from http://www.huffingtonpost.com/2010/09/09/arthur-sulzberger-we-will n 710251.html
Im, I., Hong, S., & Kang, M. (2011). An international
comparison of technology adoption: Testing the UTAUT model. Information
& Management, 48(1), 1.
Jackson, C. M., Chow, S., & Leitch, R. A. (1997). Toward an
understanding of the behavioral intention to use an information system.
Decision Sciences, 28(2), 357.
Johnson, K., & Wiedenbeck, S. (2009). Enhancing perceived
credibility of citizen journalism web sites. Journalism and Mass
Communication Quarterly, 86(2), 332.
Kaiser, H. F. (1974). An index of factorial simplicity.
Psychometrika, 39, 31-36.
Kaufhold, K., Valenzuela, S., & de Zuniga, H. (2010). Citizen
journalism and democracy: How user-generated news use relates to
political knowledge and participation. Journalism and Mass Communication
Quarterly, 87(3/4), 515.
Lasica, J. D. (2003). Blogs and journalism need each other. Nieman
Reports, 57(3)
Lu, J., Yu, C., & Liu, C. (2009). Mobile data service
demographics in urban china. The Journal of Computer Information
Systems, 50(2), 117.
Merritt, D. (1998). Public journalism and public life: Why telling
the news is not enough Lawrence Erlbaum Associates.
Ohmynews. (2011). Retrieved 04/03, 2011, from
http://international.ohmynews.com/about/
Parker, G. G., & Van Alstyne, M. W. (2005). Two-sided network
effects: A theory of information product design. Management Science,
57(10), 1494.
Press Freedom. (2007). Eritrea ranked last for first time while G8
members, except russia, recover lost ground., 2007, from
http://en.rsf.org/spip.php? page=classement&id_rubrique=34
Rosen, J. (1999). The action of the idea: Public journalism in
built form. The Idea of Public Journalism,
Rosen, J., & Merritt, D. (1994). Public journalism: Theory and
practice Kettering Foundation.
Smith, A. (1997). Criteria for evaluation of internet information
resources. Retrieved November, 20, 1998.
State of News Media 2010. (2010). Overview introduction. Retrieved
04/03, 2011, from http://pewresearch.org/pubs/1523/state-of-the-news-media-2010
State of News Media 2011. (2011). Online: By the numbers. Retrieved
04/3, 2011, from http://stateofthemedia.org/2011/online-essay/data-page-7/
Swanson, E. B. (1974). Management information systems: Appreciation
and involvement. Management Science (Pre-1986), 27(2), 178.
Tan, M., & Teo, T. S. H. (2000). Factors influencing the
adoption of internet banking. Journal of Association for Information
Systems, 1(5), 1.
Venkatesh, V., & Morris, M. G. C. (2000). Why don't men
ever stop to ask for directions? gender, social influence, and their
role in technology acceptance and usage behavior. MIS Quarterly, 24,
115-137.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D.
(2003). User acceptance of information technology: Toward a unified
view. MIS Quarterly,, 425-478.
Wikipedia. (2009). Taiwan's media chaos. Retrieved 6/8, 2009,
from http://zh.wikipedia.org/wiki/% E8%87%BA%E7%81%A3%E5%AA%92%E9%AB%
94%E4%BA%82%E8%B1%A1
wikipedia. (2011). Wikipedia., 2011, from
http://en.wikipedia.org/wiki/Web_2.0
Wikipedia. (2012). News international phone hacking scandal.
Retrieved 01/25, 2012, from
http://en.wikipedia.org/wiki/News_International_phone_hacking_scandal
Yeow, P., & Loo, W. (2009). Acceptability of ATM and transit
applications embedded in multipurpose smart identity card: An
exploratory study in malaysia. International Journal of Electronic
Government Research, 5(2), 37.
Yuen, Y., Yeow, P., Lim, N., & Saylani, N. (2010). Internet
banking adoption: Comparing developed and developing countries. The
Journal of Computer Information Systems, 51(1), 52.
Zaichkowsky, J. L. (1985). Measuring the involvement construct.
Journal of Consumer Research, 12(3), pp. 341-352.
Zhang, K. Z. K., Cheung, C. M. K., Lee, M. K. O., & Chen, H.
(2008). Understanding the blog service switching in hong kong: an
empirical investigation. Proceedings of the 41st Hawaii International
Conference on System Sciences, Hawaii.
Wesley Shu, Lin, Hota Chia-Sheng
National Central University, Taiwan
Correspondence to:
Wesley Shu & Lin, Hota Chia-Sheng
Department: Information Management National Central University,
Taoyuan County, Taiwan
Email: hotalin@gmail.com
Table 1 Constructs and their measures
Construct Item Measures
performance expectancy PE01 Citizen journalistwebsites
help me obtain the news I
want.
PE02 Citizen journalist websites
have good number of news
items.
PE03 Reading citizen journalism is
efficient.
PE04 Reading citizen journalism is
effective.
PE05 Reading citizen journalism is
easy.
PE06 Information on citizen
journalist website is useful.
effort expectancy EE01 Citizen journalist websites
are easy to operate.
EE02 Using citizen journalist does
not take me much time.
EE03 The citizen journalist
websites are user friendly.
EE04 The functions at Citizen
journalist websites are easy
to use.
social influence SI01 My family or friends affect my
attitude to using citizen
journalist websites.
SI02 The society affects my
attitude to using citizen
journalist website.
SI03 I believe citizen journalism
is a trend.
mass collaboration MC01 People can share experience
and knowledge at citizen
journalism websites.
MC02 I believe the citizen
journalist websites are open
platforms for everyone to
share.
MC03 Readers can also be
journalists on citizen
journalist websites.
MC04 I can provide content to
citizen journalist websites.
facilitating condition FC01 I can use citizen journalist
websites even if I do not have
similar experience.
FC02 The technical support at
citizen journalist websites is
easy to obtain.
FC03 I am confused by the citizen
journalist website operations.
Behavior intention BI01 I like to use citizen
journalist websites.
BI02 I like to recommend citizen
journalist websites to other
people.
BI03 I will continue using citizen
journalist websites.
Table 3 Reliabilities of the pre-test scales
Scale Deleted items Cronbach's Cronbach's
[alpha] [alpha]
before after
deletion deletion
performance expectancy None 0.891 0.891
effort expectancy None 0.929 0.929
social influence None 0.730 0.730
mass collaboration None 0.909 0.909
facilitating condition FC03 0.699 0.727
Behavior intention None 0.882 0.882
Table 4. Pre-test factor loadings for the psychological scales
Factor loading
1 2 3 4 5
PE01 .136 .149 .209 .788 .119
PE02 .301 .410 .563 .727 .259
PE03 .120 .462 .403 .777 .165
PE04 .223 .318 .375 .638 -.136
PE05 .024 .203 .378 .733 .100
PE06 .004 .059 .100 .872 .092
EE01 .883 .179 .034 -.025 .193
EE02 .889 .051 .066 .135 .127
EE03 .892 .147 .015 .226 .009
EE04 .791 .341 .176 .044 .021
SI01 -.073 .031 .009 -.083 .648
SI02 .088 .370 .263 .011 .748
SI03 .071 .085 .204 .097 .866
FC01 .756 .372 -.171 -.090 .077
FC02 .652 .219 .129 .185 -.091
MC01 .121 .802 .298 .167 .224
MC02 .385 .810 -.003 .127 .110
MC03 .197 .783 .362 .185 .107
MC04 .317 .786 .087 .098 -.004
BI01 .198 .489 .671 .188 .290
BI02 -.068 .074 .756 .183 .231
BI03 .087 .257 .849 .231 .087
Table 5 Demographic variables
Variable Category N % Cumulative %
Most visited Peopo 225 59.7% 59.7%
CJ sites Wenews 56 14.9% 74.6%
Newserr 18 4.8% 79.3%
Other 78 20.7% 100.0%
Gender Female 166 44.0% 44.0%
Male 210 55.7% 100.0%
Age 15-18 11 2.9% 3.2%
19-23 159 42.2% 45.4%
24-28 164 43.5% 88.9%
29-35 34 9.0% 97.9%
36-41 2 1.5% 98.4%
41 or above 6 1.6% 100.0%
Hours per 2 hours or below 18 4.8% 4.8%
day viewing 2~4hours 86 22.8% 27.6%
Internet 4~6hours 95 25.2% 52.8%
6~8hours 85 22.5% 75.3%
8~10hours 60 15.9% 91.2%
10~12hours 22 5.8% 97.0%
12 hours or above 11 2.9% 100.0%
CJ 0.5 or below 161 42.7% 42.7%
experience 0.5-1 99 26.3% 69.0%
in years 1-1.5 22 5.8% 74.9%
1.5-2 43 11.4% 86.3%
2-2.5 28 7.4% 93.7%
2.5-3 2 0.5% 94.3%
3 or above 21 5.6% 100.0%
Hours per 0.5 hour or below 108 28.7% 28.7%
day viewing 0.5-1 hour 83 22% 50.8%
CJ 1-1.5 hours 62 16.5% 67.3%
1.5-2 hours 49 13% 80.3%
2-2.5 hours 43 11.4% 91.6%
2.5-3 hours 14 3.7% 95.5%
3 or above 17 4.5% 100.0%
Table 6 Scores on the psychological dimensions
as a function of gender
Dimension Mean F P
Male Female
performance expectancy 5.1701 5.1558 0.032 0.857
effort expectancy 5.2500 5.4100 0.070 0.063
social influence 4.6794 4.7470 0.528 0.468
mass collaboration 5.2571 5.2892 0.152 0.696
Behavior intention 5.1595 5.1958 0.129 0.719
Use behavior 2.3190 2.2018 1.453 0.229
Table 7 Scores on the psychological dimensions as a
function of age
Dimension Mean
15-18 19-23 24-28 29-35 36-41 [greater than
or equal to] 42
PE 5.16 5.14 5.60 5.05 5.06 4.93
EE 5.36 5.31 5.67 5.08 5.04 5.00
SI 4.60 4.74 5.29 4.79 4.65 4.50
MC 5.25 5.34 5.63 5.19 5.12 5.00
BI 5.08 5.33 5.70 5.01 4.84 4.50
Dimension F-Value P-Value
PE 1.592 0.148
EE 2.325 0.032
SI 2.077 0.055
MC 1.314 0.250
BI 3.121 0.055
Table 8 Scores on the psychological dimensions as a function of CJ
websites
Dimension Mean F-Value P-Value
Peopo Wenews Newserr
performance expectancy 5.1993 4.9748 4.3929 2.863 0.055
effort expectancy 5.3623 5.1250 4.4375 2.534 0.078
social influence 4.7236 4.6961 4.5000 0.089 0.994
mass collaboration 5.2914 5.1765 4.0833 3.029 0.061
facilitating condition 5.3278 5.2794 4.5000 2.171 0.057
Behavior intention 5.1811 5.2353 4.7500 0.984 0.427
Table 9 Cronbach's [alpha]
Dimension Cronbach's [alpha] N Questions
performance expectancy 0.890 6
effort expectancy 0.837 4
social influence 0.732 2
mass collaboration 0.829 4
Behavior intention 0.850 3
Table 10 Factor loadings for the psychological scales
Factor loading
1 2 3 4 5
EE01 .797 .145 .048 .050 .185
EE02 .686 .276 .152 .095 .326
EE03 .733 .201 .217 .208 .214
EE04 .699 .447 .076 .169 .252
PE1 .240 .687 .195 .299 .225
PE2 .278 .625 .110 -.019 .206
PE3 .167 .752 .137 .200 .140
PE5 .245 .769 .152 .207 .157
PE6 .273 .685 .249 .207 .086
MC1 .108 .444 .635 .108 .103
MC2 .469 .165 .665 .164 .037
MC4 .311 .219 .698 .167 .175
BI2 .158 .193 .171 .850 .231
BI3 .279 .442 .131 .601 .012
SI2 .239 .027 .301 .260 .682
SI3 .246 .099 .160 .294 .719
Table 11 Measurements of model fit
Statistic Value Threshold Result
Absolute fit
[chi square]/df 2.01 < 3 Good
GFI 0.89 > 0.8 Good
AGFI 0.86 > 0.8 Good
RMSEA 0.069 < 0.1 Good
Incremental fit
IFI 0.98 > 0.9 Good
CFI 0.98 > 0.9 Good
Parsimonious fit
PGFI 0.68 > 0.5 Good
PNFI 0.81 > 0.5 Good
Table 12 Results of hypothesis tests for direct variables
Path t-statistic p-value Result
H1 performance 19.104 0.0001 accepted
expectancy
[right arrow]
Behavior
intention
H2 effort 12.366 0.0001 accepted
expectancy
[right arrow]
Behavior
intention
H3 social -0.062 0.4753 Not accepted
influence
[right arrow]
Behavior
intention
H4 mass 14.455 0.0002 accepted
collaboration
[right arrow]
Behavior
intention
H6 Behavior 0.050 0.4801 Not accepted
intention
[right arrow]
Use behavior
*** p<0.001 ** p< 0.01 * p< 0.05
Table 13 Results of hypothesis tests for moderators
Path t-statistic p-value
Author (T) or Non-author (F)
H7 performance expectancy T 0.776 2.114 0.002
[right arrow] Behavior F 0.284
intention
H8 effort expectancy T 0.563 1.471 0.007
[right arrow] Behavior F 0.295
intention
H9 social influence T 0.638 0.081 0.502
[right arrow] Behavior F 0.621
intention
H10 mass collaboration T 0.659 2.317 0.012
[right arrow] Behavior F 0.388
intention
H11 facilitating condition T 0.725 0.521 0.321
[right arrow] Behavior F 0.702
intention
Higher involvement (H) or Lower involvement (L)
H12 performance expectancy H 0.727 0.186 0.168
[right arrow] Behavior L 0.685
intention
H13 effort expectancy H 0.766 2.243 0.015
[right arrow] Behavior L 0.509
intention
H14 social influence H 0.758 0.559 0.355
[right arrow] Behavior L 0.692
intention
H15 mass collaboration H 0.671 2.121 0.002
[right arrow] Behavior L 0.365
intention
H16 facilitating condition H 0.706 0.499 0.313
[right arrow] Behavior L 0.702
intention
Result
Author (T) or Non-author (F)
H7 Accepted
H8 Accepted
H9 Not accepted
H10 Accepted
H11 Not accepted
Higher involvement (H) or Lower involvement (L)
H12 Not accepted
H13 Accepted
H14 Not accepted
H15 Accepted
H16 Not accepted
*** p<0.001 ** p< 0.01 * p< 0.05
