W

(\l

 
       
             
   
 

1

M

H

1
1

LHHHIUH‘MH

W

E

W

M 1‘

_x_s
01-h
(DCON

 

int UBRARY
Michigan State
University

 

 

 

This is to certify that the
thesis entitled

THE INFLUENCE OF PRIOR ATTITUDE TOWARD AN
ISSUE ON INDIVIDUALS’ PERCEPTION BIAS AND
PERCEIVED CREDIBILITY ABOUT ONLINE HEALTH
INFORMATION

presented by

JEHOON JEON

has been accepted towards fulfillment
of the requirements for the

MA. degree in ADVERTISING

 

 

    

x" __.-"

.-"

f/

 

 

Major‘F-fiTéssofis-Srgnfim
12 AUG 09
Date

MSU is an Affinnafive Action/Equal Opportunity Employer

 

.-.n-v- -fi-o--I-I-I-I-o-o-I-O-

'o-n---I-n--—----o-u-.--u-a--------on.

PLACE IN RETURN BOX to remove this checkout from your record. .
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5I08 K:/Prolecc&Pres/ClRC/DateDue.indd

 

 

 

 

THE INFLUENCE OF PRIOR ISSUE ATTITUDES
ON PERCEPTION BIAS AND PERCEIVED MESSAGE CREDIBILITY:
ONLINE HEALTH INFORMATION ABOUT SMOKING BANS
By

Jehoon Jeon

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
of the degree of
MASTER OF ARTS

Advertising

2009

ABSTRACT
THE INFLUENCE OF PRIOR ISSUE ATTITUDES
ON PERCEPTION BIAS AND PERCEIVED MESSAGE CREDIBILITY:
ONLINE HEALTH INFORMATION ABOUT SMOKING BANS
By

Jehoon Jeon

Due to the potential problem that individuals with a prior attitude toward an issue may
. not see online health information as it is originally intended, this thesis investigated how
information users differently perceive online health information according to their prior
attitude toward the issue of smoking bans in public places. An online survey with 321
students at Michigan State University was conducted using an online discussion board
website with posts from both smoking—ban—supporting and smoking—ban—opposing sides.
Analysis of the data found that individuals with a prior attitude perceive online
health information in different ways. The results exemplified confirmation bias in the
case of perception bias toward the smoking-ban—supporting posts and toward the content
as a whole. As regards perception bias toward the smoking—ban—opposing posts, the data
provided evidence of relative hostile media perception. In addition, the smoking—ban-
. supporting group perceived the smoking—ban—supporting posts to be more credible than

the smoking-ban-opposing group did and vice versa. However, the level of perceived

credibility of the content as a whole did not significantly differ among participants.
This study contributed important ideas to our knowledge about perception bias
and perceived credibility of online health information. Moreover, the findings might be

useful to researchers when they aim to create persuasive online health information.

TABLE OF CONTENTS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LIST OF TABLES ........................... - -_ ................ vii

' LIST OF FIGURES - ....................................... viii
CHAPTER 1

INTRODUCTION - - - ......... 1

1.1. Use of online health information and benefits ...1

1.2. Prior studies in online health information and limitations ................ ----2

1.3. Problem definition ................. 4

1.3.1. Active information users and individuals’ perception bias 5

1.3.2. Nature of online information and divisive issues ........... 7

1.4. Purpose of the study ......... 8

1.5. Implications ......... - - . ......................................... 11
CHAPTER 2

LITERATURE REVIEW AND HYPOTHESES ........................................ 13

2.1. Divisive health issue: Smoking bans in public places ...................................................... 13

2.2. Prior issue attitudes 16

2.3. Perception bias- .................... 17

2.3.1. Confirmation bias .............................................. 18

2.3.2. Hostile media perception ..................................................... 20

2.4. Perceived credibility ............................... 23

 

iii

CHAPTER 3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

METHOD -- - - 27
3.1. Design - .............................................................................. 27
3.2. Participants _ ............................................... 27
3.3. Protection of human subjects.....- -- .......................................................... 29
3.4. Stimulus materials ........................................................ 30
3.5. Procedure - -- .................................................... 31
3.6. Measurement ....................... 32

3.6.1. Prior issue attitudes... ................................................. 32

3.6.2. Perception bias ............................................................................................. 33

3.6.3. Perceived credibility ............................................................ .34

3.7. Analytic strategy ....................................................... 35
CHAPTER 4

RESULTS ...... ............................................. 37
4.1. Grouping participants ................................ 37
4.2. Descriptive statistics ................................. 38
4.3. Test of the hypotheses ......................................................... 42

4.3.1. Confirmation bias - -- ...................... 42

4.3.2. Hostile media perception ............................................................................................... 46

4.3.3. Perceived credibility . -. ................ -48
CHAPTER 5

DISCUSSION - ......................................... 53
5.1. Prior issue attitudes and perception bias ................................................. 53
5.2. Prior issue attitudes and perceived credibility ..................................... 56
5.3. Limitations I .................. .57

 

 

APPENDICES .............................

REFERENCES .....

 

LIST‘OF TABLES

TABLE 1 Frequency Distribution onge, Race, and Gender of Participants (n=321) ..... 29
TABLE 2 Means and Standard Deviations of Prior Issue Attitudes .................................. 38
TABLE 3 ANOVA Results for Prior Issue Attitudes by Groups ....................................... 38
TABLE 4 Means and Standard Deviations of Perception Bias ......................................... 39
TABLE 5 Means and Standard Deviations of Perceived Credibility ................................ 41
TABLE 6 ANOVA: Perception Bias of the Content across Smoking—ban—supporting
Group, Neutral Group, and Smoking—ban—opposing group .............................................. 45
TABLE 7 Partial Correlation Results: Prior Issue Attitudes and Perception Bias ............. 46

TABLE 8 ANOVA: Perceived Credibility of the Content across Smoking—ban—supporting
Group, Neutral Group, and Smoking—ban—opposing group .............................................. 50

TABLE 9 Partial Correlation Results: Prior Issue Attitudes and Perceived Credibility....52

vi

LIST OF FIGURES

FIGURE 1 Prior Issue Attitudes and Perception Bias ....................................................... 48

FIGURE 2 Prior Issue Attitudes and Perceived Credibility .............................................. 52

vii

INTRODUCTION

Use of online health information and benefits

With the rapid explosion of the Internet, the public came to use online health
information to learn about health issues. According to Eastin (2001), the Internet has
become the first source of health information for many people. Online health information
is defined as any online information that enables individuals to have a better
understanding of health and make health—related decisions. It usually covers general tips
about staying healthy as well as professional information on such topics as medical
treatment for specific diseases or health conditions. Compared to the health information
that is delivered by traditional media, online health information provides an efficient way
to guide the public on diverse health issues. In sum, recent online health information
makes it possible for people to have greater access to more kinds of content related to
health issues.

Following this growing use of online health information, a number of research
studies have suggested how frequently people use online health information and how
satisfied they are with the information (Lewis, 1996; Glowniak, 1995). These studies also
maintained that online health information has great potential to improve public health by
spreading health promotion, preventing disease, and supporting clinical care for
individuals and the public (Sonnenberg, 1997; Strauss, 1997; Larkin, 1999). For example,
Pew Internet & American Life Project’s “online health search 2006” showed that most
Internet users start at a search engine when they look for health information. The report

estimated that 80% of the survey respondents, about 113 million adults, have searched for

online health information, and 94% of these found it easy to search for that information.
“Online health search 2006” indicated that people generally perceive online health
information as useful and credible, so they search for information in order to make

decisions about how to treat an illness (Fox, 2006).

Prior studies in online health information and limitations

While a number of research studies repeatedly prove that online health
information has effectively provided the public with tailored health messages (Abrams et
al., 1996), some researchers have started to doubt whether online health information
actually benefits users (McClung, 1998; Impicciatore et al., 1999; Berland et al., 2001;
Eng et al., 1998; Viswanath, 2006). Especially in light of the ultimate impact of the
Intemet’s efficacy and reach, which have been considered as salient benefits to online
health information, the above researchers have demonstrated that online health
information has a potentially harmful dark side as follows.

First of all, sources of online health information that are not credible or lack
expertise often mislead general information users, and this fact often relates to
' misdiagnosis and mistreatment (Cline & Haynes, 2001). For example, McClung’s study
(1998) about children’s diarrhea found that only 20%, 12 out of 60 websites, qualified for
the American Academy of Pediatrics (AAP) recommendation for treatment. In other
words, ordinary people are easily exposed to incorrect health information and follow the
misleading recommendations unless they have professional knowledge regarding the
treatment of childhood diarrhea.

Moreover, a study about childhood fever also found that almost 90%, 37 out of

41 websites, do not provide complete and accurate information (Impicciatore et al., 1999).
The researchers systematically searched online health information through two search
engines, Yahoo and Excite. Although the search engines generated the most relevant and
popular search results, the information that was obtained did not provide useful tips on
feverish children for parents. Thus, the researchers concluded that only a few websites
provided complete and accurate online health information, even though fever in children
is a common healthcare problem.

In addition, the Federal Trade Commission evaluated the appropriateness and
quality of online health advertisements about six major diseases: arthritis, cancer,
diabetes, heart disease, HIV/AIDS, and multiple sclerosis. They have organized “surf
days,” a time period of federal surveillance campaign, to identify commercial websites
that may mislead online users. After only a few hours of searching, they found more than
400 websites containing potentially false or deceptive advertising claims about treatment,
cure, or prevention, which reveals the potential harm of inappropriate online health
information.

Secondly, in addition to the fact that online health information might be
inaccurate, it is also too difficult for general users to understand as well as inaccessible to
users with certain barriers, so that some individuals are not able to use it efficiently.
According to the studies about problems in online health information, contrary to the
general belief that it helped all individuals get diverse instruction and guidance, the
illiteracy of users and their inability to adopt the new technology should be enhanced
before online health information can be used appropriately and effectively.

For example, Berland and his colleagues (2001) claimed that most online health

information is designed for people who are educated to more than an 8th grade level.
Considering that the average reading ability for US. adults is 8th grade, they worried that
many health—related websites contain information that is too difficult for lay people
without higher educational experiences to understand. Eng et al. (1998) have also pointed
out that online health information has barriers to access for common users as well as
those who were underserved by the traditional healthcare system. A recent study by
Viswanath (2006) also confirmed that online health information does not improve health
care service for all individuals because the Internet is still used by specific groups of

people, which are higher—income, more—educated, younger, and employed groups.

Problem definition

Many researchers and organizations have been exploring alternative ways of
making high-quality online health information more user friendly. Also, individuals are
being educated a variety of ways to figure out credible information from a trustworthy
source from the Web and to understand the limitations of online health information.
However, even though these efforts have continuously increased the quality of online
health information and its comprehension by users, there is still a huge gap between the
actual benefits of online health information and the way those benefits are perceived by
users as discussed above. In other words, most individuals perceive online health
information to be usefirl, credible, and satisfactory although it actually does not contain
sufficient health information nor is it easily understandable.

This disparity between individuals’ perception about online health information

and its actual benefits are increasing due to the following issues regarding the nature of

information users and online information itself. First of all, individuals have perception
bias when they interpret information. Secondly, online information often refers to the

controversial issues and represents the firmly held attitudes of diverse users.

1) Active information users and individuals’ perception bias

While individuals’ use of mass media and the gratifications from those media
have long been interests of communication studies, recent mass media research studies
have focused on the concept of active audience. Contrary to traditional media use, current
studies explain what people do with media rather than what media do to people. For
example, Bauer (1973) demonstrated the changing concept of the audience from “an
aggregate of passive individuals” to “an interactive social system.” Besides, Katz,
Blumler, and Gurevitch (1974) stated that active audiences are able to control their own
media consumption. In sum, individuals are aware of their needs and can thus evaluate
diverse media channels and content before they select specific media that provides the
most gratification they seek (Katz, Blumer, & Gurevitch, 1974; Palmgreen, Wenner, &
Rosengren, 1985; Rubin, 2002).

In particular, researchers suggested that media users play an active role in
choosing and using media — both traditional and new media — including newspaper, radio,
television, and the Internet (KO, Cho, & Roberts, 2005). After Kuehn (1994) suggested
that the uses and gratification perspective can serve as an appropriate theoretical
framework to explain users’ psychological and behavioral tendencies in media studies,
subsequent researchers also indicated that the uses and gratification approach applied to

the study of computer—mediated communication, such as Internet use (Eighmey, 1997;

Lin, 1996; Morris & Ogan, 1996; Newhagan & Rafaeli, 1996).

While an active audience has been considered to react differently to the same
messages depending on their characteristics, cognitive decision making literature also
demonstrated that individuals have an opportunity to seek new information and reconcile
it with their existing beliefs (Kuhn, 2001). In his study, lay people interpreted the
information according to their pre—existing beliefs in systematic ways, so called
perception biases. For example, people try to search through new information for
evidence that supports their pre—existing beliefs (Chapman & Elstein, 2003) or to ignore
contradictory opinions in it (Chinn & Brewer, 1993). In addition, individualsjudge the
. new information to be less credible because they perceive it as unfavorable to their own
point of view (Choi, Watt, & Lynch, 2006).

Because diverse communication takes place across cyberspace, individuals’ prior
knowledge of content has been considered as a significant factor impacting the way
people evaluate a new message critically (Eagly & Chaiken, 1993). For example, in the
study of Keselman and his colleagues (2008), individuals’ prior understanding of heart
disease was related to their interpretation about the new messages. Lay people who have
imprecise knowledge of heart disease had searched for information on irrelevant sites to
confirm their incorrect initial belief. In sum, individuals’ pre—existing interests, values
and prejudices are reinforced rather than challenged, and many online discussion
participants simply seek out online information only from where their interests lie
(Dahlberg, 2001).

Considering that individuals are able to perceive misleading information to be

appropriate as they please, and the information possibly provides the most gratification

they seek, these perception biases of online health information users might be a critical
, problem. In sum, lay people who are unaware of the limitations of health information just
feel satisfied with what they obtained, although it is not the information most relevant to

their needs (Lorence & Greenberg, 2006).

2) Nature of online information and divisive issues

Moreover, the above possibility of individuals’ perception bias is enlarged
because online health information often consists of controversial health—related issues
and represents the firmly held attitudes of each side. Lee and Leets (2002) suggested that
people with an extreme opinion have been increasingly using the Internet to express their
ideas, spread their beliefs, and recruit new members. For example, individuals with an
extreme opinion, such as a point view on one side of a divisive issue, seek to persuade the
general public of their viewpoints and post their opinions on those websites. Thus,
regarding general health—related information, when lay people search for online health
information through the web, they are easily exposed to diverse extreme opinions, which
might contain differing extreme messages from both sides of the argument.

What is more, those extreme opinions posted on websites are supported with
scientific evidence from trustworthy sources. Wyatt (1997) suggested that the content of
, the websites may not be correct even if the original information sources were reliable, so
that the determination of information from a health website is good or bad has become a
critical issue. Jadad and Gagliardi (1998) also point out determining the accuracy of the
information could be more challenging than the search for the information itself. In this

situation, online health information users easily confused about making a decision as to

 

which information is credible even though credibility is assured by scientific evidences.
In recent Internet use, people with common interests have gathered virtually to
share experiences, ask questions, or provide emotional support and self help advice.
Individuals freely communicate with each other by posting their opinions on web—based
communication channels, such as online discussion boards. For example, Yahoo!
(www.yahoo.c,om) listed almost 300,000 online community groups in the health and
wellness section as of April 2009. Considering that individuals are increasingly forming
and facilitating social networks to share their own opinions through websites, it could be
a problem if websites represent firmly held extreme opinions on specific controversial
issues. In particular, in the case of divisive health—related issues containing arguments
from both sides, lay people may be exposed to online information where certain
individuals advocate for one side of an issue with some salient points, but so do their

opponents in other online information.

Purpose of the study

As previous research studies have shown, there is a discrepancy between user
perception about online health information and the actual benefits which user obtains
through it, thus individuals with a prior attitude toward a specific topic might perceive
online health information as they please. That is, when people read online health
information, their perception bias based on their prior issue attitudes is likely to
predispose them not to see online health information as the message is originally intended.
For example, Keselman et al. (2008) suggested that an incorrect or imprecise previous

attitude leads individuals to search for information from irrelevant sites, often seeking out

data to confirm their incorrect beliefs. Moreover, individuals overlook essential
information while they mistakenly believe misleading information to be valuable. At the
same time, they may quickly bypass information that seems too professional for their
prior knowledge level because an active audience can search out extremely diverse online
health information on the Web. In sum, individuals possibly perceive certain online health
information to be more or less credible according to their prior health beliefs.

In light of the potential problems above related to the differential between users’
perception about online health information and its actual value, information users may
selectively perceive specific information to strengthen or weaken their prior issue
attitudes. The fact that websites ofien contain extreme opinions also aggravates the
potential perception bias. Therefore, how individuals with a prior attitude toward a
specific topic actually perceive online health information needs to be explained.

Despite the critical harm caused by users’ biased perception about online health
information, little is known about how people differently perceive health information
from websites. Moreover, despite substantial evidence of a huge gap between the actual
- benefits of online health information and how those benefits are perceived by users, the
way in which people with different prior attitudes on a specific issue perceive online
information differently remains relatively unexplained. Considering the fact that the
Internet has become the first source of health information for many people (McMullan,
2006) and that the public can freely search for and learn about health and medical issues
posted by other users (Bylund et al., 2007), understanding the public's perception toward
divisive messages in online health information will lead to appropriately informing and

educating the public about the health—related issues in the perspective of message senders.

Moreover, information users might understand the limitations of online health
information and be able to use a common sense and good judgment when they evaluate
the quality or accuracy of the information. Thus, this study investigates how online health
information users perceive that information by focusing on individuals’ prior issue

attitudes:

RQ: Do online health information users perceive the same message differently
according to their prior issue attitudes? If so, how individuals’ prior issue attitudes are

related to perception bias and perceived credibility of online health information?

To help comprehend the research question, this study investigates individuals’

. prior issue attitudes, perception bias, and perceived credibility with respect to an online
discussion board about the issue of “Smoking ban in public places, such as restaurants
and bars.” Since smoking is largely known as the leading cause of preventable death, the
topic regarding smoking ban might be helpful to promote a better health for public and to
positively impact their health conditions. Besides, this topic has been an issue with a
moderately high profile in the mass media, a durable shelf life, and identifiable groups of
individuals with strong attitudes on one side or the other. In order to provide an online
media environment to the participants, an online survey is conducted. According to the
extent of their prior issue attitudes, the subjects are divided into three groups: smoking—

ban-supporting group, neutral group, and smoking—ban-opposing group. Then, one—way

ANOVA is performed to explain how individuals’ perception bias and perceived

credibility vary according to their prior issue attitudes. In addition, after controlling for

ID

the gender and smoking status, the data is analyzed with partial correlation to find out the

relationship of prior issue attitudes to perception bias as well as to perceived credibility.

Implications

This study contributes to recent research trends related to active audience
because it focuses on how individuals’ perception on the same message varies according
to their prior attitude toward the topic. Also, with the increasing use of Internet and online
health information, knowledge about how individuals perceive online health information
will be directly related to understanding the effective and practical methods of health
communication. From the perspective of message senders, understanding how
information users perceive the messages is vital for communicating more clearly and
creating more persuasive messages. Additionally, the more we understand how
individuals’ perception bias caused by their prior issue attitudes or issue involvement
influences perceived credibility about the information they obtain, the better we can
create more user friendly and personalized online health information. Although numerous
studies have investigated how we can create credible health information, only a few of
them have examined how users perceive the same message differently according to their
prior attitude strength toward the topic. Thus, this study will be a great contribution to
understanding how we can develop current online health information, not only to be more
credible but also to be more persuasive to information users according to their prior issue
attitudes.

The following chapter introduces the issue of smoking bans in public places as a

divisive health issue. Subsequently, concepts and previous literature regarding perception

ll

bias - such as confirmation bias and hostile media perception - and perceived

credibility are discussed to provide a theoretical framework for the study and to generate

hypotheses.

12

LITERATURE REVIEW AND HYPOTHESES

Divisive health issue: smoking bans in public places

For some issues, a large number of people may have strong previous viewpoints.
For other less salient topics, a fewer number of people may be interested. Considering the
main purposes of this study are to examine how individuals’ prior issue attitudes are
related to perception bias, and individuals’ perceived credibility of online health
information, this study advances the controversial topic related to “Smoking ban in public
places, such as restaurants and bars.” To test perception bias and perceived credibility
toward online health information, this study searched for an issue with a moderately high
profile in the mass media, a durable shelf life, and identifiable groups of individuals with
strong attitudes on one side or the other. Among diverse divisive issues, “smoking ban in
public place” might be a good example of such a divisive health-related argument where
individuals might feel differently when they judge new online health information.

Moreover, smoking bans in public places has been one of the hottest issues in the
state of Michigan (York, 2009). For instance, the Campaign for Smokefree Air (CSA) is
currently working to pass legislation in 2009 to make all public places in Michigan
smokefree, including restaurants, bars, and casinos. The State Senate and the House of
Representatives are also in negotiations about a smoking ban for the state of Michigan.
According to Barbeau et al. (2007), while both sides are having a difficult time coming to
an agreement, public places smoking policies have suggested as a divisive issue among
members who smoke and those who do not. In particular, when the Michigan Senate

voted to end smoking in bars and restaurants, the voices on either side of the debate grew

13

 

more intense.

In addition, the message that smoking harms the public health has been widely
known and disseminated by the anti—smoking campaign for decades. While the most
people have known for some time that smoking can cause lung cancer, as well as many
other risks and illnesses, the exposure of non—smokers to the second—hand smoke has
been highlighted since the late 19805 by the US National Research Council (1986).
Evidence for how smoking is bad for those around smoker has been found that the cancer
risk from non—smoking spouses and children living with smokers increased almost
double in occupational settings (IARC, 1986; Jones, 1999; CEPA, 1999). For example,
The American Medical Association has reported that more than 40,000 people in the
United States annually die from the effects of second-hand smoke. They also found that
second—hand smoke increases the non—smokers’ risk of death from heart disease and
diverse types of cancer by 25%.

In this situation, people who are in favor of a state—wide smoking ban argue that
second—hand smoking critically harms public health. Second—hand smoking is defined as
the involuntary and passive tobacco inhalation by non—smokers. As a combination of
sidestream smoke and mainstream smoke, it includes smoke from the end of a lighted
cigarette, pipe, or cigar, but also exhaled smoke from smokers (National Toxicology
Program, 2005). Because smoke lingers in the atmosphere for a couple of hours even
after a cigarette has been extinguished, non—smokers are involuntarily exposed to the
same nicotine and other toxic chemicals as smokers. Therefore, the more non—smokers
_ breathe in second—hand smoke, the higher possibility is for them to be exposed to the

danger of diverse diseases, including cancer, respiratory infections and asthma (Ries et al.,

14

2004). According to the 2000 National Household Interview Survey (NHIS), about 25%
of the US. population is exposed to second—hand smoke. Compared to the 1992's data
which shows 35.6% of individuals were exposed to second—hand smoking, it is obvious
that the exposure of second—hand smoking has decreased. However, second—hand
smoking still remains a major public health threat as it is the third leading cause of death
in the U.S., while it is preventable (Wilson & Thomson, 2002).

On the other hand, people who are against a state—wide smoking ban doubt that
second—hand smoke is truly a health hazard. First, very few of the research studies have
been investigated secondary smoke as a major topic, and those research studies have not
come up with the statistical significance necessary to be acceptable as anything more than
chance or research bias. For example, according to a Congressional Research Service
Report (1995), the statistical evidence does not appear to support a conclusion that there
are substantial health effects from passive smoking, which means second—hand smoking
showed no measurable harm at all. Besides, the massive WHO IARC study (1998)
concluded that the risk to a non-smoker living and working with smokers for an extended
length of time (40—50 years) was still statistically insignificant. Therefore, people against
a smoking ban argue that second-hand smoking is not as harmful as it is made out to be.
Second, there is an issue regarding an individual's right to make a decision to go into
either a smoking place or non—smoking place. Thus, a smoking ban at the national or
state level could be an infringement of the individual's right to smoke. Third, a smoking
ban can affect individual's property right because loss of economic benefits for business
owners, such as restaurants and bars, is explicitly expected if smokers are not able to use

those places while they smoke.

15

As discussed above, the smoking—ban—supporting side of the divide, which
represents people in favor of a state—wide smoking ban, says second—hand smoking is a
critical threat to public health and personal responsibility. However, the other side, the

smoking-ban—opposing group, sees the issue as an intrusion on personal rights and

property rights. They also doubt that the statistics used in second-hand smoking research
studies were significant and that second—th smoking is actually a serious harm to
public health. Since advocates on both sides of the issue produce arguments through the
form of online health information to insist on their opinion and to persuade lay people,
both sides argue their point of view by showing data from professional sources and
trustworthy organizations. As a result, general information users are exposed to the
opinions of both sides, and likely to feel confused in making a decision as to which side
is more believable. Moreover, online information users are likely to perceive the same
messages differently according to their prior issue attitudes. Therefore, this study settled

’ on the controversial issue over the smoking bans in public places, such as bars,

restaurants, and workplaces.

Prior issue attitudes

The concept of attitude has been broadly defined as a construct that implies an
individual ’3 positive or negative view of the attitude object, such as a person, place, thing,
issue, or event. In several studies, prior issue attitudes were considered as a main cause
which induces people to interpret incoming stimuli as congruent with the perceiver’s
preconceptions (Lord, Ross, & Lepper, 1979) or hostile against their side (Vallone, Ross,

& Lepper, 1985). Although other terms, such as issue-involvement or partisanship, used

16

in the previous studies, Perloff (1989) interpreted prior attitude as stemming from ego-

involvement of participants with the topic and groups of highly involved, so called

partisans. In addition, GinerSorolla and Chaiken (1994) pointed out that prior beliefs

could act as ajudgmental heuristic and influence perception and memory in a way that
emphasizes the contrary nature of the world. The current study conceptualized the prior

issue attitudes to include all those concepts of prior beliefs, issue-involvements,

partisanships, preconceptions, or expectancies. Considering prior issue attitudes as what
target audience participants have in mind when evaluating the influence of information,
individuals’ prior issue attitudes might cause a perception bias and a different perceived

credibility when subjects estimate obtained information.

Perception bias

Perception bias has been considered to negatively influence individuals when
they attempt to make unprejudiced decisions or interpret information objectively
(Friedlander & Phillips, 1984; Silverman, 1992). Among diverse types of perception bias
— such as confirmation bias, hostile media perception, overconfidence bias, race bias,
gender bias, familiarity bias, and anchoring - this study has focused on two
representative perception bias in media studies: confirmation bias and hostile media
perception. Since these two phenomena affect individuals’ media exposure and
perceptions of media bias in the contradictory way, this study investigated how
individuals differently perceive the media content by addressing these two perception
biases.

As noted earlier, while perception biases of lay people make them interpret the

17

information by their pre—existing beliefs in systematic ways, confirmation bias explains
the way people try to search for evidence supporting their pre—existing beliefs from new
information. On the other hand, hostile media perception demonstrates that individuals
perceive new information to be unfavorable to their prior beliefs. In particular, when it
comes to the online health information, which is often controversial and contains diverse
firmly held opinions of both sides, the current study assumes that these two perception
biases could appropriately demonstrate the research question as to how individuals’
perceptions about online health information vary according to their prior attitude toward

an issue.

1) Confirmation bias

Research studies on motivated information processing have suggested that pre—
existing attitudes and preferences affect the selection, perception, acceptance, and recall
of messages in diverse contexts and situations (Baumeister & Newman, 1994; Biek,
Wood, & Chaiken, 1996; Chaffee, Saphir, Graf, Sandvig, & Hahn, 2001; Ditto & Lopez,
1992; Kunda, 1990; Lodge & Taber, 2000; Snyder & Campbell, 1980). As Nickerson
(1998) defines it, confirmation bias connotes the seeking or interpreting of information in
ways that support pre—existing attitudes, beliefs, expectations, or a hypothesis. In other
words, confirmation bias is a tendency to search for or interpret new information to
confirm one's preconceptions and to avoid information and interpretations which
contradict prior beliefs. Thus, people sometimes seek out, notice, recall, and process
information to confirm their own prior issue attitudes. At the same time, people

sometimes ignore, fail to remember, forget, and undervalue information which

18

contradicts their opinion.

Wason (1960) first demonstrated confirmation bias in a study in which subjects
were asked to find out a rule from a series of three—number strings. Participants started
with the number string ‘2-4—6’, and asked to find out what the rule is. When subjects
check their assumptions by announcing some examples of different number string, they
were informed whether their answers were correct or not. Although the rule was just any
three numbers increasing in magnitude, most participants were highly confident that the
rule was numbers increasing by two and they did not offer disconfirming number strings,
such as increasing by one. Through this experiment, he demonstrated those who believe
that they had already discovered a rule were more likely to suggest confirming examples
and ignored disconfirming ones. Mahoney and DeMonbreun (197 8) also replicated his
finding that even scientists showed a tendency to confirm rather than disconfirrn their
pre—existing belief as non scientists do. Additionally, according to Snyder and Swann
(1978), people tended to seek confirming information even when the source of their
belief was not credible. Later studies also confirmed that people lend greater weight to
information that supports their prior beliefs than to information that counters them
(Shaklee & F ischhoff, 1982; Skov & Sherman, 1986).

As Hogarth (1980) has stated, confirmation bias is a systematic information
processing shortcut and possibly results in judgment errors. Individuals attempt to
confirm their pre—existing belief by truncating the cognition process to select known
information when they are exposed to both already known information and opposing
information (Shaklee & Fischhoff, 1982). Therefore, in regards to the individuals dealing

with diverse online information, they are likely to perceive it based on'their pre-existing

l9

attitude or preference. That is, based on previous studies of confirmation bias, online
health information users with pro—existing attitudes and preferences might ignore
opposing information. At the same time, they might confirm their prior issue attitudes by
interpreting information in ways that support prior issue attitudes.
Therefore, drawing on the previous research studies on confirmation bias, this study
advances the following hypotheses:
HI .° Online health information users with a prior issue attitudes are likely to
perceive the content as congruent with their own point of view, that is to say,
a) The smoking—ban—supporting group is likely to perceive the content as
congruent with their own point of view, which is in favor of a smoking ban.

b) The smoking-ban -opposing group is likely to perceive the content as congruent

with their own point of view, which is opposed to a smoking ban.

2) Hostile media perception
Demonstrating the attitude-influenced psychological mechanism of audience,
hostile media perception studies have explained what actually happens in people’s minds
that leads them to sense media content as biased and hostile to their viewpoint on an issue.
Pointing out a wide range of possible cognitive and perceptual processes, hostile media
perception suggests that an audience with strong attitude toward an issue tends to
perceive media content as biased against their own point of view, even if the content is

well-balanced.

Vallone, Ross, and Lepper (1985) first suggested that hostile media perception

explains the tendency for a highly involved audience to judge mass media coverage as

20

unfavorable to their own point of view. When they showed balanced news coverage of
the 1982 Beirut massacre to pro—Arab and pro—Israeli undergraduates, both groups of
participants judged the content as biased in favor of the opposite side and as produced by
a person of opposing opinion. Further, both sides estimated that the news coverage
contained a lower percentage of favorable references to their own point of view.

While a number of researchers have shown that hostile media perception appears
across diverse types of controversial issues (Christen, Kannaovakun & Gunther, 2002;

’ Dalton, Beck, & Huckfeldt, 1998; Matheson & Dursun, 2001), Gunther and Schmitt
(2004) added empirical evidence that is particularly relevant to media content. That is,
individuals with a strong prior issue attitudes tend to perceive media coverage as unjustly
slanted against their opinion even if the message of the media is impartial. In their
experiment, participants were provided with both a newspaper article and a student essay,
which contained the same balanced content. However, participants perceived the
newspaper article as biased while they perceived the student essay as impartial.

In regards to the role of the individual’s strength of opinion, Perloff (1989)
insisted that individual’s high involvement with a topic is a main cause of hostile media
perception, and Gunther and Liebhart (2006) also maintained that the individual’s ego—
involvement with the topic is positively related to the perceived bias in the news content.
That is, those who have a strong attitude or opinion toward an issue perceive mass media
content to be partial against their own opinions although media coverage does not include
any presence of bias. Among those individuals with high involvement with a topic, this
tendency to make biased evaluations of media content was also confirmed even though

the coverage was clearly unbalanced (Gunther et al. 2001). Gunther and his colleagues

21

described this phenomenon as relative hostile media perception, by which people on both
sides may see the media content to actually be positioned in the same directions, but each
side sees the slant as relatively more disagreeable to or less in agreement with their own
position.

In hostile media perception studies, researchers also have highlighted the concept

of active audience in regards to partisans (Schmitt et al., 2004). They define partisans as

9

 

“highly involved individuals who hold strong and deeply felt opinions on a specific issue.’
Traditionally, partisan groups and their members have actively expressed their viewpoint
by picketing, making threats, and demonstrating. Recent active audience with an extreme
opinion has been increasingly using the Internet to express their ideas by searching for

the information and creating online communities and discussion forums on the Web. As
Hagel and Armstrong (1996) stated, the fastest growing phenomenon on the Web is the
posting of evaluations, opinions, or experiences using diverse online communication
tools, such as Web forums, blogs, instant messaging tools, and social networking sites.

As noted earlier, individuals are exposed to diverse opinions of the issue online,
and people with an extreme opinion are increasingly using the online media to express
their ideas. In regards to the previous research studies in hostile media perception, online
health information users with strong prior attitude on one side of the issue might perceive
new information to be biased against their prior issue attitudes. Thus, this study advances
the following hypotheses:

H2: Online health information users with a strong prior issue attitudes are likely to

perceive the content as favorable to the opinions of the opposite side, that is to

say,

22

a) The smoking—ban—supporting group is likely to perceive the content as
favorable to the smoking—ban—opposing point of view.

b) The smoking-ban -opposing group is likely to perceive the content as favorable

to the smoking—ban—supporting point of view.

Perceived credibility

Perceived credibility can be defined as believability, the extent to which
individuals perceive the information represent their own belief (Tseng & Fogg, 1999).
For example, Zanna and Del Vecchio (1973) demonstrated that credibility is enhanced
when individuals perceive the position taken by television news to match their own prior
issue attitudes. They suggested that television news is perceived as less credible when the
audience does not hear what it would like to hear. A subsequent study by Johnson and
Kaye (2002) also supported the conclusion that in regards to online media, the similarity
of individuals’ prior attitude to the media is likely to predict its perceived credibility. In
sum, the perception about the similarity between individuals’ prior issue attitudes and the
obtained information are positively related to their perceived credibility of the content.

As perceived credibility result from evaluating multiple dimensions
simultaneously, prior credibility studies branch into three overlapping categories: source
credibility, media credibility, and message credibility (Sundar, 1999). First, source
credibility explains how characteristics of a source influence the way individuals perceive
a message to alter pre—existing attitudes to certain topics (Metzger et al. 2003). According
to Hovland and Weiss (1951), sources with expertise and trustworthiness produce a more

positive attitude and induce more behavioral compliance than less credible sources do.

23

 

That is, source credibility significantly influences an individual’s attitude and behavior
responses and thus acts as a key element for message effectiveness in regards to
individuals perception.

Although source plays a significant role in determining persuasive effectiveness,
recent credibility studies regarding online health information have demonstrated that
individuals pay little attention to the source. According to Bates and his colleagues (2006),
individuals evaluated the quality of online health information without focusing on source
credibility. Whether the message was attributed to a highly credible national organization
or to a general website, the difference in source did not have a significant effect on user
evaluations of the quality of the information. Moreover, Eysenbaoh and Kohler (2002)
also demonstrated that individuals do not check the “about us” section of websites where
they could evaluate the credibility of a source. In sum, online health information users
make little use of source credibility when they assess the credibility of online health
information.

Secondly, media credibility investigates the characteristics of the channel by
which messages are delivered (Slater & Rouner, 1996). Typically, these studies have
explored which media — such as newspaper, radio, or television — are better perceived as a
believable source of information and therefore more efficiently deliver a persuasive
message than the others. For example, Johnson and Kaye (1998) found that online media
was more likely to be perceived as having greater credibility than newspapers when
Internet use first began to spread. Schweiger (2000) countered that although the
credibility of the Web was fairly high among the general public, newspapers were rated

ahead of it. Additionally, Kiousis (2001) explored perceptions of news credibility for

24

 

television, newspapers, and online news and argued that newspaper had the highest
credibility, followed by online news and television news, respectively.

Finally, message credibility concerns the perceived credibility of a message itself
(Kiousis 2003). Instead of measuring the credibility of the Internet as a whole or
examining the credibility of sources, Sundar and Nass (2001) studied message credibility
by measuring peoples’ perceived credibility of online news stories. In sum, studies about
message credibility were not focused on how the channel or the source is related to the
perceived credibility. Instead, message credibility focuses solely on the message itself.

In spite of the categorization of credibility discussed above, recent studies
suggested that individuals evaluate diverse overlapping criteria when perceiving the
credibility of information. For example, Eysenbaoh and Kohler (2002) stated that
information users know they need to evaluate the credibility of a website by looking for
the source, a professional design, scientific evidence, language, and ease of use, as well
as content itself. Fogg and his colleagues (2002) suggested that site Sponsor, contact
information, currency of information, customer service response time, and site structure
were the most influential in perceptions of the credibility of information.

Therefore, instead of using any of above three specific credibility categories,
source credibility, media credibility, and message credibility, recent online health
information studies have examined overall perceived credibility which operationalized
the credibility indicating whether information and information sources are believable
(Buoy, 2003). Considering perceived credibility is generally agreed to the results from
characteristics of source, media and message, perceived credibility examined the degree

to which individuals see the content as conveying trustworthy, believable, accurate, or

25

expert information. As a result, mass communication scholars started to consider
credibility by measuring diverse factors, such as accuracy, fairness, lack of bias,
completeness, depth, and trustworthiness (Flanagin & Metzger, 2001; Johnson & Kaye,
2002).

As individuals have different prior issue attitudes before they are exposed to
media content, they generally perceive differently for the media and indicate different
perceived credibility according to their perception bias. Therefore, in regards to the
previous research studies in perceived credibility, this study advances the following
. hypotheses:

H3: Online health information users’ prior issue attitudes are related to the
perceived credibility about the content, that is to say,

a) The smoking—ban—supporting group is likely to perceive the opinions of the
smoking—ban—supporting side to be more credible than the smoking-ban-
opposing group does.

b) The smoking-ban -opposing group is likely to perceive the opinions of the
smoking—ban—opposing side to be more credible than the smoking—ban—

supporting group does.

26

METHOD

Design

This study was designed to test how individuals’ perception bias and perceived
credibility about online health information vary according to their prior issue attitudes. In
order to examine this research question, the current study employed an online survey
which measured participants’ prior issue attitudes, perception bias, and perceived
credibility on the issue of “Smoking ban in public places.” This online survey required

about 15-20 minutes to complete.

Prior issue attitudes were calculated based on participants’ responses to three

questions and was manipulated to classify three different groups (smoking—ban—
supporting group VS neutral group VS smoking-ban-opposing group). After being

exposed to the created online discussion board website, participants also asked questions

about the perceptions regarding the stimulus - perception bias and perceived credibility.

Participants

To empirically test the relationships proposed by the research question and the
three hypotheses, a survey was conducted online from April 19 to April 26, 2009
regarding the issue of “Smoking ban in public places.” The data for this study were
collected fiom undergraduate and graduate students attending three different introductory
communication classes at Michigan State University. The instructors of those classes
assigned extra credit points for each completed survey to encourage students to take the

survey.

27

 

As shown in Table 1, the four demographic measures were age, grade, race, and
gender. Respondents were asked the year of birth to minimize confusion which comes
from the different types of age system. Whereas the prior survey asked about individual’s
highest level of education, modified survey included year in school. Thus, they were
asked to select current grade among “Freshman, (Undergraduate, First grade),”
“Sophomore (Undergraduate, Second grade),” “Junior (Undergraduate, Third grade),”
“Senior (Undergraduate, Fourth grade),” and “Graduate student.” Participants were asked
their gender and chose their race from the following list: “Caucasian,” “African
American,” “Hispanic or Latino,” “Asian,” “Multi—racial,” “Others.” Participants who
chose “Others” were asked to specify their race.

A total of 321 participants participated in this study (67.0% female and 33.0%
male; mean age = 21.17; 72.9% Caucasian, 8.1% African American, 0.6% Hispanic, 14.6%
Asian, and 4.4% other). Although student samples had a limitation in representing the
general population, the sample criteria appear to be acceptable because student samples
were more reflective of current intemet users who are used to diverse ways of selecting
credible online health information from a trustworthy source (LaRose et al., 2001).
Moreover, using previous studies as a guide (Giner-Sorolla and Chaiken, 1994), student
samples would be acceptable in the current study to measure how individuals’ attitude
affect their perception regarding the content since student samples consist of individuals
with variety level of prior issue attitudes including neutral subjects. In their study, student
samples were used to suggest that prior beliefs may lead people to prejudge information
and explored perception toward media content. Finally, this relatively homogeneous

group relates to more internally reliable analysis.

28

Table 1

Frequency Distribution of A ge, Race, and Gender of Participants (n=321)

 

Mticipants1n=3212

 

 

 

 

 

 

 

Demographics F %
Gender
Male 106 33
Female 215 67 A
Age
19 yrs 7 2.2
20 yrs 106 33
21 yrs 131 40.8 i
22 yrs ‘ 40 12.5
23 yrs 14 4.4
24 yrs 9 2.8
25 yrs 6 1.9
27 yrs 3 0.9
28 yrs 3 0.9
30 yrs 2 0.6
Grade
Freshman 0 0
Sophomore 7 2.2
Junior 158 49.2
Senior 128 39.9
Graduate 28 8.7
Race
Caucasian 234 72.9
African American 26 8.1
Hispanic 2 0.6
Asian 47 14.6
Other 12 3.8
Total 321

 

Protection of human subjects

Consideration of protection of human subjects was addressed by applying to the
Michigan State University’s Institutional Review Board before conducting human subject
research. The ethical principles and standards outlined by the American Psychological
Association (2009) for human subject research were adhered to. Subjects who agreed to

29

participate in the research were asked to do so, and they were able to quit the survey
when they become uncomfortable or disoriented during the survey. Participants’ answers

and results were also treated with strict anonymity.

Stimulus materials

Using previOus studies as a guide (Gunther & Schmitt, 2004), an online
discussion board website was constructed as online health information. To increase
I external validity, real issue and real arguments of both smoking—ban-supporting group
and smoking—ban—opposing group were used in this study. This stimulus was assembled
from the public’s posts of several recent news articles as well as diverse posts fi'om online
discussion forums about the topic. Created website entitled “How do you feel about a
Smoking bans in public places, such as restaurants and bars?” in the template of an online
discussion board.

The constructed content consisted of two parts. At the beginning, an introductory
section that briefly explained the divisive surrounding of the issue of smoking ban in
public place, such as restaurants and bars. Then, posts for and against the smoking ban
fiom several actual online arguments were arranged. In order to prevent bias which might
be caused by the order of arguments, two different types of stimuli were created. Type 1
starts from the post of the smoking—ban—supporting group and then comes with opinions

of the smoking-ban—opposing group in turns. For type 2, the post of the smoking-ban-

opposing group comes first, and opposite postings followed in rotation. In an effort to
produce a balanced content, this study took care to make a minimum modification for all

those six posts to be equal in length and similar in style. That is, all posts were quoted

30

 

from existing online posts, and minor modifications made only for language flow and
syntax demand.

Containing actual opinions from both smoking—ban—supporting and smoking—
ban—opposing posts, attempts to create a believable online discussion board appeared to
be successful, as no participants found the stimulus fictitious until they were debriefed. In
I addition, the created website was revised until all of the judges agreed it was balanced as
containing opinions of both sides equally. Finally, it was pre—tested with 28 non—partisan
judges who held a balanced position on the issue of smoking bans in public places. The
check of measure of perception bias revealed a mean of .11 (SD = .79), close to the
neutral midpoint on a 7 point scale where —3 means “extremely biased in favor Of

smoking ban” and where 3 means “extremely biased against smoking ban.”

Procedure

The survey proceeded in the order of introduction, consent form, random
assignment (participants would not notice this stage), the first part of survey (questions
on demographic information, smoking status, and prior attitudes toward the issue of
“Smoking ban in public places”), stimulus exposure, the second part of survey (questions
on perception bias and perceived credibility), and debriefing information.

Participants first read and signed a consent form that stated the survey as
questions of evaluating the online information related to the issue of smoking bans in
public places. Participants were also informed that their participation would result in
extra credit. In the first part of survey, questionnaire examined the position and degree of

attitude toward the issue of smoking bans in public places. To verify the participants'

31

’ position and strength of attitude, an online survey including filter questions asked
questions related to how much participants support or oppose smoking bans in public

places. Later, based on the first part of survey, participants were categorized to be three

groups, smoking—ban—supporting group, neutral group, and smoking—ban-opposing
group.

After finishing the first part of survey, the issue of smoking ban in public place
was introduced and participants were asked to read the stimulus, an online discussion
board website. Since the content contains both positions of opinions about the issue, it
was considered to be balanced. As discussed above, in order to prevent bias which might
be caused by the construction of the content, two different types of stimuli were used for
this study. Those stimuli were previously randomized before participants received a
survey questionnaire.

The second part of survey measured perception bias and perceived credibility of
online health information. All participants were asked how they perceived the online
health information to be positioned on the issue. The responses about each six posts as
well as the content as a whole were evaluated. As this study also interested in participants’

perceived credibility, additional questions asked about believability, accuracy, fairness,

and depth of information of the stimulus.

Measurement
Prior issue attitudes

To figure out where participants position themselves and to classify them into

three different groups (smoking-ban—supporting group, neutral group, and smoking—ban-

32

opposing group), three questions were asked to measure individual’s prior attitude toward
. the issue. Following a study of Gunther and his colleagues (2001) regarding the prior
attitude, which is the measurement of favorableness or unfavorableness Of an individual’s
evaluation of a given object, a series of modified attitude questions about the issue of
smoking bans in public places were asked, 1) extent of position, 2) attitude toward

smoking—ban—supporting group, and 3) attitude toward smoking-ban-opposing group.

First of all, participant’s position on the issue was assessed by asking “When you
consider the issue of ‘Smoking ban in public places’, where do you position yourself on
this issue?” Response options ranged from —3 (Extremely in favor of smoking ban) to 3
(Extremely against smoking ban) by a 7—point scale. To incorporate group identification
in this question about the extent of position, additional questions about attitudes toward

members of smoking—ban—supporting group and smoking-ban-opposing group were

asked on a 5 point scale. The scale ranged fi'om 1 (not at all support) to 5 (extremely
support). Afier appropriate recoding, these two questions about group identification were

combined to calculate an attitude toward the group. Then, the average of the two items -

extent of position and attitude toward groups - was calculated. Exploratory Factor

Analysis (EFA) with Principal Component Analysis (PCA) extraction method clearly
showed one factor with 73.34 percent of total variance explained using the eigenvalue
criteria of 1. Cronbach’s alpha reliability also indicated high internal consistency (a

= .830). Thus, above three items were used as indicators of prior issue attitudes.

Perception bias

Afler reading an online discussion board website, respondents were asked

33

questions about perception bias. A measurement for perception bias was adapted from a
previous study (Chia et al., 2007) and modified to determine how they perceive the
stimulus. Participants were asked to answer the question “How do you identify ‘the post’
to be positioned on the issue?” for each of the six individual posts -— three smoking—ban—
supporting and three smoking—ban—opposing posts. In addition, a replication question,
“How do you identify ‘the content, as a whole’ to be positioned on the issue?” was also
asked to measure overall perception bias toward the whole content. All of these questions
were followed by a 7—point scale anchored by —3 (Extremely in favor of smoking ban) to
3 (Extremely against smoking ban).

Since the stimulus was composed of six separate posts of both smoking—ban—
supporting and smoking—ban—opposing sides, analyses about both perception bias toward
smoking—ban—supporting posts and perception bias toward smoking—ban—opposing posts
were performed. For perception bias toward smoking—ban—supporting posts, the EPA
with PCA extraction method based on varimax rotation showed one factor with 59.09
percent of total variance explained. Also, for perception bias toward smoking—ban—
opposing posts, a factor with 63.53 percent of total variance explained. Cronbach’s alpha
also indicated strong internal consistency for perception bias toward smoking—ban—
supporting posts (a = .648) and perception bias toward smoking—ban—opposing posts (a

= .713).

Perceived credibility
Using previous studies as a guide (Gaziano & McGrath, 1986; Johnson & Kaye,

2002; Meyer, 1988; Newhagen & Nass, 1989), this study measured credibility as a multi—

34

 

dimensional construct of believability, accuracy, fairness, and depth of information. These
researchers have used these four measurements to operationalize the perceived credibility.
Therefore, participants were asked how believable, accurate, fair, and in depth they
judged the content as a whole to be. The EFA with PCA extraction method clearly
showed one factor with 73.15 percent of total variance explained using the eigenvalue
criteria of 1. Because these four items had high reliability as predicted (a = .868), they
were also averaged into a variable of perceived credibility.

In addition, since the stimulus was composed of six individual posts of both
smoking-ban—supporting and smoking—ban—opposing opinions equally (three posts each),
analysis about the perceived credibility of each side would explain more about
individuals’ perception of the content. In order to investigate individuals’ perceived
credibility toward posts both for and against a smoking ban, questions asked how
believable, accurate, fair, and in depth they judged each post to be by using a five—point
scale, ranging from 1 (not at all) to 5 (extremely). The EFA with PCA extraction method
clearly indicated one factor solution with 75.11 percent of variance explained for
perceived credibility toward smoking-ban—supporting posts. Also, one factor with 75.32
percent of variance explained for perceived credibility toward smoking—ban—opposing
posts. As those each 12 items had also good reliability (or = .663; a = .706) for internal
consistency, the measures of each post were averaged into two variables, a perceived
credibility toward smoking—ban—supporting posts and perceived credibility toward

smoking—ban—opposing posts.

Analytic strategy

35

This study analyzed the collected data using the statistical software program
Statistical Package for the Social Sciences (SPSS 14.0). One—way ANOVA was
performed to explain how above groups differed with respect to the perception bias and
perceived credibility. Post—hoc tests were also conducted to find out which group’s mean
was significantly different from each other. Then, to find out the relationship between
prior issue attitudes and perception toward the content, partial correlation was used after

controlling for the gender and smoking status.

36

 

RESULTS

Grouping participants
According to the extent of the individual’s prior issue attitudes, the classification

of subjects as smoking—ban—supporting group, neutral group, and smoking-ban—opposing

group was performed first. After participants indicated their prior issue attitudes by
responding to three questions, the current study examined three groups of individuals
with different prior issue attitudes according to the calculated measurement of prior issue
attitudes. That is, participants with prior issue attitudes below 0 on the scale were
considered to be in the smoking—ban—supporting group, and above 0 to belong to the

smoking-ban-opposing group. Individuals with 0 prior issue attitudes were considered to

be in the neutral group. In the end, dividing the sample at this point resulted in 234
participants (73%) classified as smoking—ban—supporting group, 30 (9%) as neutral group,
and 57 (18%) as smoking-ban-opposing group.

In order to determine the effectiveness of this classification and to test if these
three different groups showed a significantly differential prior issue attitudes, a one—way
ANOVA was conducted. As expected, the result revealed a statistically significant
difference among the three groups regarding their prior attitude toward the issue, F (2,3 19)
= 385. 19, p < .01, 112: .71. In addition, post hoc comparison using Scheffe’s procedure at
p < .05 showed that the three groups were significantly different from each other:
smoking—ban-supporting group (M = -2.29, SD = 1.06), neutral group (M = 0, SD = 0),

and smoking—ban—opposing group (M = 1.66, SD = 1.04).

37

 

Table 2

Means and Standard Deviations of Prior Issue Attitudes

 

 

 

Factor Variables Min. Max. Group Means Standard Deviation
Smoking—ban—supporting group —3.50 —0.50 —2.29 1.06
Neutral group .00 .00 .00 .00
Smoking—ban—opposing group .50 3.50 1.66 1.04
TOTAL (N = 321) —3.50 3.50 -1.38 1.85

 

 

I"Positive number indicates smokinghban-opposing prior issue attitudes and negative
number indicates smoking-ban-supporting prior issue attitudes.

Table 3

ANO VA Results for Prior Issue Attitudes by Groups

 

 

Sum of Mean
Squares df Square F Si 3. [12
Between
pum- Groups 777.20 2 388.60 385.18 .00 .71.
Issue . .
attitudes WM“ 320 82 318 101
Groups ' "
Total 1098.02 320

 

Descriptive statistics

The purpose of this study was to demonstrate how online information users’
perceptions about online health information vary acbording to their prior attitude toward
an issue. Therefore, in order to effectively analyze the data, statistical analysis for three
variables - prior issue attitudes, perception bias, and perceived credibility - was
performed before testing the hypotheses. Since the content contained three posts each
from both smoking—ban—supporting and smoking—ban—opposing sides, individuals’

38

 

 

perception bias and perceived Credibility were measured in regards to both smoking—ban—
supporting and smoking—ban—opposing posts as well as the content as a whole.
Descriptive statistics of prior issue attitudes are shown above in Table 2; perception bias
toward each post, smoking—ban-supporting posts, smoking—ban—opposing posts, and the
content as a whole in Table 4; perceived credibility toward each post, smoking—ban-

supporting posts, smoking—ban—opposing posts, and the content as a whole in Table 5.

Table 4

Means and Standard Deviations of Perception Bias

 

 

 

 

 

 

 

 

 

Factor Variables Min. Max. Group Means Standard Deviation

(lst Post)
Smoking-ban—supporting group —3.00 3.00 —2.07 1.59
Neutral group —3.00 3.00 —1.17 1.80
Smoking-ban—opposing group —3.00 3.00 —0.74 2.35
TOTAL (N = 321) —3.00 3.00 —l.75 1.84

(2nd Post)
Smoking—ban—supporting group —3.00 3.00 1.56 1.51
Neutral group —3.00 3.00 0.80 1.52
Smoking—ban-Opposing group —3.00 3.00 0.75 1.65
TOTAL (N = 321) —3.00 3.00 1.35 1.57

(3rd Post)
Smoking—ban—supporting group —3.00 3.00 —1.06 1.75
Neutral group —3.00 3.00 —0.37 1.43
Smoking—ban—opposing group —3.00 3.00 —0.14 1.83
TOTAL W = 321) —3.00 3.00 —0.83 1.77

(4th Post)
Smoking—ban—supporting group —3.00 3.00 1.57 1.49
Neutral group —2.00 3.00 1.27 1.36
Smoking—ban—opposing group —3.00 3.00 1.00 1.72
TOTAL W = 321) —3.00 3.00 1.44 1.53

 

 

39

Table 4 (cont’d)

 

 

 

 

 

 

 

 

 

 

 

Factor Variables Min. Max. Group Means Standard Deviation
(5th Post)
Smoking—ban—supporting group —3.00 3.00 —1.91 1.50
Neutral group —3.00 2.00 -1.30 1.47
Smoking—ban—opposing group -3.00 3.00 —0.65 1.97
TOTAL (N = 321) —3.00 3.00 -1.63 1.66
(6th Post)
Smoking-ban—supporting group —3.00 3.00 2.08 1.48
Neutral group —3.00 3.00 1.30 1.73
Smoking—ban—opposing group -3.00 3.00 0.70 2.23
TOTAL (N = 321) —3.00 3.00 1.76 1.74
(smoking—ban—supporting posts)
Smoking—ban—supporting group -3.00 2.67 —1.68 1.24
Neutral group —3.00 1.00 -0.94 1.19
Smoking—ban—opposing group —3.00 3.00 —0.51 1.75
TOTAL (N = 321) -—3.00 3.00 —1.40 ' 1.41
(smoking—ban—opposing posts)
Smoking—ban—supporting group -2.67 3.00 1.74 1.19
Neutral group -1.33 3.00 1.12 1.28
Smoking-ban—opposing group —2.33 3.00 0.82 1.64
TOTAL (N = 321) —2.67 3.00 1.52 1.34
(the content as a whole)
Smoking-ban—supporting group —3.00 3.00 —0.20 1.11
Neutral group -2.00 2.00 0.10 0.80
Smoking—ban—opposing group —2.00 3.00 0.35 1.22
TOTAL (N = 321) —3.00 3.00 -0.07 1.12

 

 

 

* lst, 3rd, 5th Posts were Smoking-ban—supporting messages and 2nd, 4th, 6th Posts
were Smoking-ban-opposing messages.

*Positive number indicates smoking-ban-opposing perception bias and negative
number indicates smoking-ban-supporting perception bias.

40

Table 5

Means and Standard Deviations of Perceived Credibility

 

 

 

 

 

 

 

 

 

 

 

 

 

Factor Variables Min. Max. Group Means Standard Deviation

(1 st Post)
Smoking-ban—supporting group 1.00 5.00 3.30 0.77
. Neutral group 1.00 4.00 2.93 0.60
Smoking—ban—opposing group 1.00 5.00 2.72 0.87
TOTAL (N = 321) 1.00 5.00 3.16 0.81

(2nd Post)
Smoking—ban—supporting group 1.00 4.50 2.34 0.81
Neutral group 1.25 3.75 2.84 0.62
Smoking—ban—opposing group 1.50 5.00 3.10 0.77
TOTAL 0V = 321) 1.00 5.00 2.52 0.84

(3rd Post)
Smoking—ban—supporting group 1.00 5.00 2.83 0.89
Neutral group 1.00 4.75 2.86 0.71
Smoking-ban—opposing group 1.00 4.00 2.43 0.93
TOTAL (N = 321) 1.00 5.00 2.76 0.89

(4th Post)
Smoking—ban—supporting group 1.00 5.00 2.93 0.83
Neutral group 1.75 4.25 3.20 0.63
Smoking—ba‘n—opposing group 1.00 5.00 3.12 0.82
TOTAL (N = 321) 1.00 5.00 2.99 0.81

(5th Post)
Smoking-ban—supporting group 1.00 5.00 3.40 0.80
Neutral group 2.25 4.00 3.22 0.53
Smoking—ban—opposing group 1.00 4.25 2.71 0.86
TOTAL (N = 321) 1.00 5.00 3.26 0.83

(6th Post)
Smoking—ban—supporting group 1.00 5.00 2.50 0.94
Neutral group 1.00 4.25 2.73 0.77
Smoking—ban—opposing group 1.00 5.00 2.92 0.90
TOTAL (N = 321) 1.00 5.00 2.60 0.93

 

 

41

Table 5 (cont’d)

 

 

 

 

 

 

 

Factor Variables Min. Max. Group Means Standard Deviation
(smoking—ban—supporting posts)
Smoking-ban-supporting group 1.75 5.00 3.18 0.60
Neutral group 2.00 4.00 3.00 0.43
Smoking—ban—opposing group 1.00 4.25 2.62 0.76
TOTAL (N = 321) 1.00 5.00 3.06 0.65
(smoking—ban—opposing posts)
Smoking-ban—supporting group 1.00 4.25 2.59 0.67
Neutral group 1.83 3.75 2.92 0.51
Smoking—ban—opposing group 1.67 5.00 3.05 0.68
TOTAL (N = 321) 1.00 5.00 2.70 0.68
(the content as a whole)
Smoking—ban-supporting group 1.50 4.50 2.99 0.67
Neutral group 2.25 4.00 3.06 0.46
Smoking—ban—opposing group 1.00 5.00 2.83 0.67
TOTAL (N = 321) 1.00 5.00 2.97 0.66

 

 

" lst, 3rd, 5th Posts were Smoking-ban-5upporting messages and 2nd, 4th, 6th Posts

were Smoking—ban-opposing messages.
* The scale ranges from 1 (not at all credible) to 5 (extremely credible)

Test of the hypotheses
1) Confirmation bias
Hypothesis I predicted that online health information users with a prior issue
attitudes are likely to perceive the content as congruent with their own point of view. That
is to say, a) the smoking—ban—supporting group is likely to perceive the content as
congruent with their own point of view, which is in favor of a smoking ban b) the
smoking—ban—opposing group is likely to perceive the content as congruent with their

I own point of view, which is opposed to a smoking ban.

42

 

The data appear to partially support hypothesis 1. To test this hypothesis, a one
way ANOVA as well as a partial correlation were conducted to test whether or not the
perception bias varied among the three groups and to check whether any theoretical
relationship between prior issue attitudes and perception bias exists, respectively. As
shown in Table 6, in each perception bias toward individual posts and the content as a
whole varied significantly among the three groups. To be more precise, the three groups
differed significantly from each other in perception bias toward the content as a whole,
F (2,3 19) = 6. 14, p < .05, n2 = .04; toward the smoking—ban—supporting posts, F (2,319) =
19.55, p < .05, n2 = .12; toward the smoking—ban—opposing posts, F(2,319) = 13.12, p
< 05,112 = .08.

In each post hoc comparison, the significant differences occurred as follows:

First of all, in the case of perception bias toward the content as a whole, post hoc
comparison using Scheffe’s procedure at p < .05 showed that the smoking—ban-
supporting group (M = -.20, SD = 1.10) perceived the balanced content to be more
favorable to the opinion of the smoking—ban—supporting side. At the same time, the
smoking—ban—opposing group (M = .35, SD = 1.22) perceived the balanced content to be
more favorable to the smoking—ban—opposing side. The Neutral group (M = .10, SD =.80)
did not differ from either the smoking—ban—supporting or smoking—ban—opposing group.
Therefore, both hypothesis l-a) and l-b) were supported in regards to perception bias
toward the content as a whole and the result was consistent with previous findings in

confirmation bias - participants perceived the online health information to be congruent

with their prior issue attitudes.

Secondly, in the case of perception bias toward smoking—ban—supporting posts,

43

 

post hoc comparison using Scheffe’s procedure at p < .05 showed that the smoking-ban—
supporting group (M = —1.68, SD = 1.24) perceived the smoking—ban—supporting posts to
be more favorable to the opinion of the smoking—ban—supporting side. At the same time,
the neutral group (M = -.94, SD = 1.19) and the smoking—ban—opposing group (M = -
.51, SD = 1.75) perceived the smoking—ban—supporting posts to be less favorable to the
smoking—ban—supporting side. Since all the participants perceived the content to be
positioned in favor of the smoking—ban—supporting side, hypothesis l-a) was supported.
However, hypothesis l—b) was not supported with perception bias toward smoking—ban—
supporting posts.

Finally, in the case of perception bias toward the smoking—ban—opposing posts
only, post hoc comparison using Scheffe’s procedure at p < .05 showed that the smoking—
ban-supporting group (M = 1.74, SD = 1.19) perceived the smoking—ban—opposing posts
to be more favorable to the opinion of the smoking—ban—opposing side. At the same time,
the neutral group (M = 1.12, SD = 1.28) and the smoking-ban-opposing group (M =.82,
SD = 1.64) perceived the smoking—ban—opposing posts to be relatively less favorable to

the smoking—ban—opposing side. Therefore, although hypothesis l-b) was supported,
hypothesis l-a) was not supported with perception bias toward smoking-ban—supporting

posts because all the participants perceived the content to be positioned in favor of the

smoking—ban—opposing side.

 

Table 6

and Smoking-ban—opposing group

ANO VA: Perception Bias of the Content across Smoking-ban—supporting group, Neutral group,

 

 

 

 

 

 

 

 

 

 

Sum of Mean
Squares df Square F Si & 112
Perception Bias Between Groups 93.08 2 46.54 14.93 .00 .09
(1 st Post) Within Groups 990.98 318 3.12
Total 1084.06 320
Perception Bias Between Groups 39.59 2 19.80 8.38 .00 .05
(23t Post) Within Groups 751.02 318 2.36
Total 790.62 320
Perception Bias Between Groups 45.91 2 22.95 7.60 .00 .05
(3st Post) Within GrOUps 961.01 318 3.02
Total 1006.92 320
Perception Bias Between Groups 15.79 2 7.90 3.41 .03 .02
(4st Post) Within Groups 737.27 318 2.32
Total 753.07 320
Perception Bias Between Groups 76.49 2 38.24 15.19 .00 .10
(Sst Post) Within Groups 800.40 318 2.52
Total 876.88 320
Perception Bias Between Groups 93.68 2 46.84 16.99 .00 .11
(6st Post) Within Groups 876.85 318 2.76
Total 970.53 320
Perception Bias Between Groups 69.96 2 34.98 19.55 .00 .12
smokin -—ban—
gupportigng posts) Within Groups “8'88 318 1’79
Total 638.84 320
Perception Bias Between Groups 43.59 2 21.80 13.12 .00 .08
smokin ban-
lrpposing-posts) Within Groups 528’“ 318 1'66
Total 571.96 320
Perception Bias Between Groups 14.96 2 7.48 6.14 .00 .04
the content as a
ivhore) Within Groups 387'“ 318 1'22
Total 402.21 320

 

 

In addition, the measures of correlation between prior issue attitudes and
perception bias were tested after controlling for the gender and smoking status. The

results showed that prior issue attitudes were significantly related to perception bias (See

45

 

Table 7). Specifically, prior issue attitudes among all participants were significantly
related to perception bias toward the content as a whole as follows: r(319) = .198, p =.Ol.
Although content as a whole was composed both of smoking—ban—supporting and

. opposing posts, the smoking—ban—supporting group perceived the content as congruent
with their own point of view, which is in favor of a smoking ban. At the same time, the
smoking—ban—opposing group perceived the content as congruent with their own point of

view, which is opposed to a smoking ban.

Table 7

Partial Correlation Results: Prior Issue Attitudes and Perception Bias

 

 

 

Perception Bias toward: Prior Issue Attitudes
lst Post .407"
2nd Post —.306**
3rd Post .296"
4th Post —.195"‘*
5th Post .356"
6th Post —.367**
Smoking—ban—supporting posts .439**
Smoking—ban—opposing posts —.3S4**
The content as a whole .198“r

 

 

**. Correlation is significant at the .01 level (2—tailed)

2) Hostile media perception
Hypothesis 2 predicted that online health information users with a prior attitude
toward the issue are likely to perceive the content as favorable to the opinions of the
opposite side, that is to say, a) the smoking—ban—supporting group is likely to perceive
the content as favorable to the smoking—ban—opposing point of view b) the smoking-
ban-opposing group is likely to perceive the content as favorable to the smoking—ban—

46

 

 

supporting point of view.

Based on the results discussed above, hypothesis 2 did not appear to be supported
because both Sides perceived the smoking-ban—supporting and smoking—ban—opposing
posts to actually be positioned in those directions, not in favor of the opposite side. That
is to say, in the case of the smoking—ban—supporting posts, all of the participants
recognized the content to support a smoking ban. Moreover, participants saw the
smoking—ban—opposing posts as against a smoking ban. In addition, in the case of the
content as a whole, participants tend to perceive the balanced content to be congruent
with their prior issue attitudes. In other words, both smoking—ban—supporting and
smoking—ban—opposing groups perceived the content as a whole to be favorable toward
their own point of view, not as favorable toward the opinions of the opposite side.

However, perception bias toward the smoking—ban—opposing posts provided
evidence of the relative hostile media perception in that the smoking—ban—supporting
group (M = 1.74, SD = 1.19) perceived the message as significantly more biased in favor
of the smoking-ban—opposing side than did the neutral group (M = 1.12, SD = 1.28) or
the smoking—ban-opposing group (M =.82, SD = 1.64). In addition, the data revealed that
there were negative correlations in the case of the smoking—ban—opposing posts, r(319) =
—.354, p =.01. Even though all of the participants saw the smoking—ban—opposing posts
as positioned in the smoking—ban—opposing direction, the smoking—ban—supporting
group perceived the smoking—ban—opposing posts to be relatively more biased toward the

smoking—ban—opposing side than did the smoking-ban-opposing group.

47

 

Figure 1.

Prior Issue Attitudes and Perception Bias

 

Prior Attitude and Perception Bias

 

 

 

 

 

 

2.00
I
1.50 ——-—-‘x
‘ ~
1 00 ‘I. s —0- - Smoking-ban-
' ~ - supporting
0.50 I M”

 

 

 

tion bias

- l- Smoking-ban-
0.00 / opposing POStS
“-0.50 3‘
+The content as

-1.00 3’ a whole
-I .50 r

 

 

Perce

 

 

 

 

 

 

 

 

 

-2.00
Smoking-ban- Neutral group Smoking-ban-
supporting opposing group
group
Groups

 

 

 

" Positive scores on perception bias indicate participants perceived the message to be
more smoking-ban-opposing side. At the same time, negative scores on perception

bias indicate participants perceived the message to be more smoking-ban-supporting
side.

3) Perceived credibility
Hypothesis 3 predicted that online health information users’ prior issue attitudes
are related to the perceived credibility of the content, that is to say, a) the smoking—ban—
supporting group is likely to perceive the opinions of the smoking—ban—supporting side to
be more credible than the smoking—ban—opposing group does b) the smoking-ban—
opposing group is likely to perceive the opinions of the smoking—ban—opposing side to be

more credible than smoking—ban—supporting group does.

48

 

In order to test these hypotheses, a one—way ANOVA was conducted to test
whether or not the perceived credibility varied among the three groups toward smoking-
ban—supporting posts. The results suggested that there was a significant difference among
the three groups regarding the perceived credibility of smoking—ban—supporting posts,
F(2,319) = 18.5 1, p < .05, n2 = .10. Post hoc comparison using Scheffe’s procedure at p
< .05 showed that the smoking—ban—supporting group (M = 3.18, SD =.60) and the
neutral group (M = 3.00, SD =.43) showed higher perceived credibility toward the
smoking—ban—supporting posts than the smoking—ban—opposing group does (M = 2.62,
SD =.76). The smoking—ban—supporting group and the neutral group did not significantly
differ from each other. In other words, the smoking—ban—supporting group and the, neutral
group showed higher perceived credibility than does the smoking—ban—opposing group
toward the smoking—ban—supporting posts.

Secondly, a one—way ANOVA was performed to test perceived credibility toward
smoking—ban—opposing posts among the three groups. The result suggested that there
was a significant difference among these groups regarding the perceived credibility of
posts of the smoking—ban—supporting group, F (2,319) = 12.80, p < .05, 112 = .07. Post hoc
comparison using Scheffe’s procedure at p < .05 showed that the smoking—ban—opposing
group (M = 3.05, SD =.68) and the neutral group (M = 2.92, SD =.51) showed higher
perceived credibility toward smoking—ban—opposing posts than the smoking—ban—
supporting group does (M = 2.59, SD =.67). The smoking-ban-opposing group and
neutral group did not significantly differ from each other. In sum, the smoking—ban—
opposing group and neutral group showed significantly higher perceived credibility than

does smoking—ban—supporting group toward the smoking—ban—opposing posts.

49

 

 

Additionally, in order to test how perceived credibility toward the content as a

‘ whole containing opinions of both sides varied among the three different groups, a one—
way ANOVA was conducted. The result suggested that there was no significant difference
among the three groups regarding the perceived credibility toward the content as a whole,
F(2,319) = 1.64, p < .05, 112 = .01. Participants judged the content to be at approximately
the same levels of perceived credibility. Additional post hoc comparison using Scheffe’s
procedure at p < .05 also showed that the smoking—ban—supporting group (M = 2.99, SD
=.67), neutral group (M = 3.06, SD =.46), and smoking—ban—opposing group (M = 2.83,

SD =.67) did not significantly differ fi'om each other.

Table 8

ANO VA: Perceived Credibility of the Content across Smoking—ban—supporting Group, Neutral
Group, and Smoking—ban—opposing group

 

 

 

 

 

 

 

Sum of Mean
Squares df Square F Sig. J12

Perceived credibility Between Groups 16.90 2 8.45 14.05 .00 .08
(lst Post) Within Groups 191.21 318 .60

Total 208.11 320
Perceived credibility Between Groups 29.56 2 14.78 23.91 .00 .13
(2st Post) Within Groups 196.54 318 .62

Total 226.10 320
Perceived credibility Between Groups 7.79 2 3.90 4.99 .01 .03
(3st Post) Within Groups 248.33 318 .78

Total 256.12 320
Perceived credibility Between Groups 3.18 2 1.59 2.43 .01 .02
(4st Post) Within Groups 207.95 318 .65

Total 211.13 320
Perceived credibility Between Groups 21.48 2 10.74 17.30 .00 .10
(Sst Post) Within Groups 197.42 318 .62

Total 218.90 320
Perceived credibility Between Groups 8.68 2 4.34 5.18 .01 .03
(6st Post) Within Groups 266.31 318 .84

Total 274.99 320

 

50

 

Table 8 (cont’d)

 

Perceived credibility Between Groups 14.19 2 7.09 18.51 .00 .10

 

 

(Smoking-ban-
supporting posts) Within Groups ”1'85 318 '38
Total 136.04 320
Perceived credibility Between Groups 11.14 2 5.57 12.80 .00 .07
(Smoking—ban—
opposing posts) Within Groups ”8'43 318 '44
Total 149.57 320
Perceived credibility Between Groups 1.41 2 0.70 1.64 .19 .01
(the content as a
whole) Within Groups “5'98 318 ’43
Total 137.39 320

 

Then, measures of correlation between prior issue attitudes and perceived
credibility toward the smoking—ban—supporting posts were tested after controlling for the

- gender and smoking status. Result showed that prior issue attitudes were significantly

related to perceived credibility as follows: r(3 19)= -.3 14, p =.01. Therefore, the

smoking—ban—supporting group perceived smoking—ban—supporting posts to be more
credible than the smoking—ban—opposing group does. Additionally, measures of partial
correlation between prior issue attitudes and perceived credibility toward the smoking-
ban—opposing posts were tested after controlling for the gender and smoking status.
Result showed that prior issue attitudes were significantly related to perceived credibility
as follows: r(3l9) = .322, p =.01. Thus, the smoking—ban—opposing group is likely to
perceive the opinions of the smoking—ban—opposing side to be more credible than the
smoking—ban—supporting group does. Finally, measures of partial correlation between
prior issue attitudes and perceived credibility toward the content as a whole were tested
after controlling for the gender and smoking status, and result indicated that prior issue

attitudes were not significantly related to perceived credibility, r(3 19) = -.074, p =0].

51

Table 9

Partial Correlation Results: Prior Issue Attitudes and Perceived Credibility

 

 

 

 

Perceived Credibility toward: Prior Issue Attitudes

lst Post —.272**

2nd Post .373"

3rd Post —.179**

4th Post .161"

5th Post —.288**

6th Post .229”

Smoking—ban—supporting posts —.314**
Smoking—ban-opposing posts .322“
The content as a whole —.074

 

**. Correlation is significant at the .01 level (2—tailed)

Figure 2

Prior Issue Attitudes and Perceived Credibility

 

 

Prior Attitude and Perceived Credibility

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.40
3.20 -O - Smoking-ban-
supporting
; 3.00 POStS
ES .. - - - -
E 2 80 I Smoking ban
opposrng
9 sts
'§ 2.60 W
'g -t-The content as
o 2.40 a whole
a.
2.20
2.00 ‘ ‘
Smoking-ban- Neutral group Smoking-ban-
supporting opposing
group group
Groups

 

52

 

DISCUSSION

While there is a discrepancy between actual benefit of online health information
and the way those benefits are perceived by information users, individuals with a prior
issue attitudes might perceive online health information as they please, and perception
bias based on their prior issue attitudes might predispose them not to see online health
information as it was originally intended. Therefore, the current study questioned how
individuals differently perceive online health information according to their prior attitude
toward the issue. An online survey was performed with content containing opinions on

both sides of the issue of “Smoking ban in public places.”

Prior issue attitudes and perception bias

Findings of this study indicated that individuals with a prior issue attitudes
systematically perceived health information in the online media differently. In particular,
the results suggested a significant gap in perception bias among the three groups not only
in regards to the individual posts but also to the content as a whole. Although both the
smoking—ban—supporting and smoking—ban—opposing group identically perceived the
smoking—ban—supporting posts to be positioned in favor of the smoking—ban—supporting
side and the smoking—ban—opposing posts to be positioned in favor of the smoking—ban—
opposing side, there was a significant difference in the extent of perception bias among
the three groups.

First of all, in the case of perception bias toward the smoking—ban—supporting

posts, the results demonstrated prior issue attitudes and perception bias were positively

53

 

 

correlated with each other. Therefore, the smoking—ban—supporting group perceived the
smoking—ban-supporting posts to be more favorable to their opinion while the neutral
group and the smoking—ban—opposing group perceived the smoking—ban—supporting
posts to be relatively less favorable to the smoking—ban—supporting side. Although all the
participants perceived the content to be positioned in favor of the smoking—ban—
supporting side, a positive correlation between prior issue attitudes and perception bias
relatively indicated the presence of confirmation bias.

The findings in this case of perception bias toward the smoking-ban-supporting

posts allow us to take a broader perspective on individuals’ perception of online media.
While past research studies suggested that mass media condition may generate hostile
media perception, the results of this study represent the idea that the new media
. environment — such as online discussion board — might influence differently on the
individuals’ perception bias toward a specific content. That is, it is an intriguing issue that
requires further research why online media environment, such as online discussion board
in the current study, generate a confirmation bias instead of hostile media perception.
Secondly, for perception bias toward the smoking—ban—opposing posts, the
results demonstrated that prior issue attitudes and perception bias were negatively
correlated with each other. That is, the smoking—ban—supporting group perceived the
smoking—ban—opposing posts as significantly more biased in favor of the smoking—ban—

opposing Side than did the neutral group or smoking-ban-opposing group, although all of

the participants saw the smoking—ban—opposing posts as positioned in the smoking—ban-
opposing direction. This negative correlation between prior issue attitudes and perception

bias indicated a relative hostile media perception: the smoking—ban—supporting group

54

perceived the smoking—ban—opposing posts to be relatively more biased in favor of the
smoking—ban—opposing side than did the smoking—ban—opposing group.
Then, what makes this result of relative hostile media perception only in the case

of the smoking-ban-opposing posts? Considering that the current study found an
evidence of confirmation bias from the other cases, in the case of the smoking-ban-

supporting posts and the content as a whole, the characteristic of audience might be

related to this unique result from the perception bias toward the smoking-ban-opposing
posts. As discussed earlier, only 18% of the entire participants were smoking-ban-

opposing group and they might believe their own point of view to be the minority opinion
due to the current anti-smoking atmosphere in the US. Since it has been considered that
individuals perceive hostile media perception and they assume others do not share the

same opinion on the issue, future research is required whether individuals in smoking-
ban-opposing side believe their own point of view to be relatively minor opinions and are

likely to overestimate the commonness of their or other individuals’ opinions.

Thirdly, as regards perception bias toward the content as a whole, the results
demonstrated that prior issue attitudse and perception bias were positively correlated with
each other and that individuals looked at the content in very different ways, even though
they were exposed to exactly the same message. As predicted, the result was consistent
with previous studies regarding confirmation bias: participants perceived the online
health information to be congruent with their prior issue attitudes although the content
was actually well-balanced and included opinions of both sides equally. However, the
results were inconsistent with previous findings in hostile media perception studies as

participants perceived the balanced message to be positioned in favor of their prior

55

 

 

attitudes toward the issue. Although some individual cases in this study supported the
predictable hostile media perception, the tendency of individuals’ perception bias toward
the balanced content differed from previous findings.

It is important to note that the results for the content as a whole agree with the
previous findings of confirmation bias studies, not with hostile media perception studies.
Considering that hostile media perception studies suggested that individuals perceived
balanced content to be biased toward the opposite opinions in the case of mass media,
whether online media especially generate confirmation bias instead of hostile media
perception is an issue that requires future research. In other words, the results of the
current study suggest that online media might mediate confirmation bias although the
previous study of Gunther and Schmitt (2004) suggested that mass media condition
. generate hostile media perception. In regards to online media, which has become the first
media source for recent information seekers (Eastin, 2001), subsequent study is needed to

explain why this result fi'om online media differs from previous findings.

Prior issue attitudes and perceived credibility

Findings of this study also indicated that individuals’ perception bias relatively
causes information users to perceive more credibility fiom information which supports
their points of view. As predicted by previous credibility studies, the smoking—ban—
supporting group perceived the smoking—ban—supporting posts to be more credible than
' the smoking-ban—opposing group did and vice versa.
However, when it comes to the perceived credibility of the content as a whole,

there were no significant differences according to participants’ prior issue attitudes. Since

56

 

both smoking-ban—supporting and smoking—ban—opposing sides were equally covered in
the content, it seems that individuals felt more credibility from the opinions supporting
their own, and less credibility from the opposing opinions. Therefore, the level of
perceived credibility of the content as a whole did not Significantly differ between the

smoking—ban—supporting group and smoking-ban-opposing group. It is notable that this

result indicates that individuals perceive more credibility in the specific part of the online
information that represents their prior issue attitudes, which finally might result in

perception bias toward the content.

Limitations

Although the result of the study explained how individuals’ perception toward
online health information varied according to their prior issue attitudes, it is important to
note that there were several limitations to the research. First, a sample of participants
which could represent the general population would be ideal. Since the sample for this
study was the students in Michigan State University, who are relatively homogeneous in
age, and educational background, future research may benefit fi'om expanding its
participant pool to include a broader range of participants, such as people of different
generations. In particular, the older generations still tend to be wary of online media and
their inability to adopt the new technology might cause different results.

Secondly, this study used divisive and contentious issue of smoking bans in
public places to simultaneously examine the effects of confirmation bias and hostile
media perception. Since online health information is defined as any online information

that enables people to have a better understanding of health issues and make health-

57

 

 

related decisions, it is obvious that the issue of smoking bans in public places — as online
health information - could warn lay people of the dangers of smoking and provide

general tips and information. However, compared to the purely informational topics, such
as tips about staying healthy or professional information on medical treatment for specific
diseases, the issue of smoking bans in public places could be more related to the political
or normative topics, which might produce different results. Therefore, one of the next
steps in a future study might be using purely informative topics and looking into the
results whether the findings of the future study would be consistent with the current study.
Finally, online discussion boards are merely one type of recent Internet use sand
they do not represent all the uses of online media. Therefore, additional study might be
needed whether the findings of this study would be consistent with the diverse types of

online media.

With the increasing use of the Internet and online health information, knowledge
about how individuals perceive online health information will be directly related to
understanding the effective and practical methods of health communication in the firture.
Therefore, the data theoretically contributed important ideas to our knowledge about
perception bias and perceived credibility of online health information, especially about
the perception of individuals with a prior attitude toward controversial issues. Practically,
the findings in this study might be useful to researchers and organizations when they aim
to create persuasive online health information. As discussed above, individuals perceived
the online health information differently according to their prior attitude toward the issue.

In particular, once individuals shape a firmly held attitude toward the issue, it could be

58

 

- extremely difficult to persuade them to the opposite way of thinking because individuals
perceive more credibility when they look information reflecting their prior issue attitudes
whether they are correct or not. Therefore, health communicators should start educating
the public about the health-related issues as early as possible - before individuals have a
firmly held attitude on an issue. For example, the effect of early education for individuals
without prior attitudes on smoking’s harmful effects would be more influential and

pervasive than for those who already have a positive or negative attitude toward the issue.

59

 

 

Appendix A - CONSENT FORM

Title of the Study: Perceived Credibility of Online Information

You are invited to the research study, “Perceived Credibility of Online Information.” We
; ask that you to read this form and ask any questions you may have before agreeing to
participate in this study.

Background Information:

You are being asked to participate in this study on this single occasion for extra credit in
your class. This survey is about evaluating your perceived credibility of selected online
information messages. We are examining what kinds of message you select from the
search engine results page, how you evaluate credibility from the selected online
information and whether those selected messages affect your prior attitudes or beliefs.
The survey will take about 20 minutes to complete. You are not expected to participate in
any treatments that would incur the risk of physical or mental injury during your

- participation in this study. Please note that you need to be Older than 18 years old to be
eligible to participate in this study. If you do not wish to participate in this study or if you
are younger than 18 years old, please see your instructor immediately. Your instructor
will give you an alternative task you can complete.

Benefits and Risk:

You are not expected to participate in any treatment that would incur the risk of physical
or mental injury during your participation in this study. While this study is not expected
to yield any immediate benefit to the individual participants, the information you provide
will be used for designing more credible online messages and advancing social scientific
knowledge. If you become uncomfortable or disoriented during the survey, verbally
notify the instructor as soon as possible.

Anonymity:

Your answers and results from this study will be treated with strict anonymity. Your
privacy will be protected to the maximum extent as allowable by law. No one will know
what you write. You can feel free to answer each question based on your actual
knowledge and behavior. The questions that ask about your background will be used only
to describe the types of students completing this survey. The information will not be used
to find out your name. Your name will not be associated with the answers you provide to
questions in any report of research findings. Your name is only requested so that we may
inform your instructor of your participation solely for the purpose of awarding credits.

‘ Voluntary Nature of the Study:

You are being asked to freely participate in this study. Completing the survey is
voluntary. Therefore, you may choose not to participate, and may withdraw from the

60

 

study at any time without penalty. Moreover, you can withdraw or refuse to answer any
particular question without penalty.

Contacts and Questions:

You have the right to contact the investigators if you have any objections to or concerns
with any aspect of this study. The contact information is as follows: Dr. Hye—J in Paek
(517—432—8377/paekh@msu.edu), or Jehoon Jeon (734—846—0757/jeonjeho@msu.edu).

If you have any questions or concerns regarding your rights as a study participant, or are
dissatisfied at any time with any aspect of this study, you may contact — anonymously, if
you wish Michigan State University Human Subject Protection Programs at Michigan
State University, by phone: (517) 355—2180, fax: (517) 432-4503, email: irb@msu.edu,
or regular mail: 202 Olds Hall, East Lansing, MI 48824.

'- Your signature below indicates your voluntary agreement to participate in this study.

 

' Print your name (e.g., Last Name, First Name)

 

Your signature and date

 

' Class Name (e.g., ADV XXX)

61

 

 

Appendix B — QUESTIONNAIRE

SECTION 1 - The following questions ask about your background.

1. Year of birth?

 

2. Year in school?
1) Freshman (Undergraduate, lst grade)
2) Sophomore (Undergraduate, 2nd grade)
3) Junior (Undergraduate, 3rd grade)
4) Senior (Undergraduate, 4th grade)
5) Graduate student

3. Which one of these groups BEST describes you?
1) Caucasian
2) African American
3) Hispanic or Latino
4) Asian
5) Multi—racial
6) Others (please specify)

4. Your gender is:
1) Male
2) Female

SECTION 2 - The following questions ask about your smoking experience.

1. In general, how would you describe your health?
1) Excellent
2) Very good
3) Good
4) Fair
5) Poor

2. During the past 30 days (one month), on how many days did you smoke cigarettes?
1) 0 days
2) 1 or 2 days
3) 3 to 5 days
4) 6 to 9 days

62

 

5) 10 to 19 days
6) 20 to 29 days
7) All 30 days

3. During the past 30 days (one month), on the days you smoked, how many cigarettes
did you usually smoke?

1) I did not smoke cigarettes during the past 30 days (one month)

2) Less than 1 cigarette per day

3) 1 cigarette per day

4) 2 to 5 cigarettes per day

5) 6 to 10 cigarettes per day

6) 11 to 20 cigarettes per day

7) More than 20 cigarettes (1 pack) per day

SECTION 3 - The Following questions ask about your attitude about the
“Smoking Ban in public place.”

 

Now, please consider people in the following groups: the “Smoking-ban—
supporting group” and the “Smoking—ban—opposing group.”

“Smoking—ban—supporting group” argues that second—hand smoke can be a
potential harm for non smokers due to its increasing risk of heart disease,
cancer, emphysema, and many other diseases.

On the contrary, “Smoking—ban—opposing group” suggests that the state—wide
smoking ban has been threatening individuals’ rights and property rights with
misleading statistic results of smoking harm.

 

 

 

1. Which best describes your level of interest in the issue of a state—wide smoking ban
in public places?

1) None 2) Little 3) Moderate 4) Strong 5) Very strong
2. When you consider the issue of “Smoking Ban in Public Places”, how do you

identify yourself as belonging to the “Smoking—ban—supporting Group” versus
“Smoking—ban—opposing Group”? (Please mark on the number)

 

—3 —2 —1 0 1 2 3
smoking— - Smoking—
an— .
supporting Neutral ban—opposciflg
group 81' P

63

 

3. Where on this line, do you think the “Smoking—ban—supporting Group” positions
themselves? (Please mark on the number)

 

l 2 3 4 5
Not at all Slightly Somewhat Moderately Extremely
Favorable Favorable Favorable Favorable Favorable
towards towards towards towards towards
Smoking— Smoking— Smoking— Smoking— Smoking-
ban— ban— ban- ban— ban—
supporting supporting supporting supporting supporting
group group group group group

4. Where on this line, do you think the “Smoking—ban—opposing Group” positions
themselves? (Please mark on the number)

 

1 2 3 4 5
Not at all Slightly Somewhat Moderately Extremely
Favorable Favorable Favorable Favorable Favorable
towards towards towards towards towards
Smoking— Smoking— Smoking— Smoking— Smoking—
ban— ban- ban— ban—opposing ban—
opposing opposing opposing group opposing
group group group group

64

 

 

 

SECTION 4 - You are about to look into “the online discussion board” about the
“Smoking Ban in public place.” You will find arguments between the “Smoking—

ban—supporting group” and the “S ' ' U ban n ' a group.” Please read next

 

discussion board to answer following questions.

 

restaurants and bars"? l mm“ 9.9.?!"

How do you feel about a “Smoking Ban in Public Places, such as

 

Topic: Smgking 3ng in Public Places - Should smoking be banned in
public places?

This discussion Board is for people who wants to argue on the issue, “Smoking Ban in
Public Places."

A “Smoking ban in public places" prohibits tobacco smoking in restaurants, bars,
workplaces and other public spaces to protect people from the effects of second—hand smoke.
While Mid1igan has long pursued the policy of letting the market rule on whether bars and
restaurants would allow smoking, recently, there have been many debates over whether there .
should be a state—wide ban or not.

This issue is VERY oontrovertial and there are continuous debates between two groups of
people, “Smoking—ban—supporting” groups & “Smoking—ban—opposing” groups. Should
Smoklng be banned in public places??? What is your opinion about the issue in Michigan?

 

 

 

Displaying posts 1 — 6 out of 93,135 by 18,627 people.

 

GuGEnGétlilIthIGAN) wrote Yesterday *
Smoking is dangerous not only for you but also for people near you who did not
CHOOSE to smoke!! Smoking causes many diseases and harms to people nearby,
who just breathe in the smoke. Smokers choose to smoke, but people nearby DO
NOT choose to smoke. A complete ban on smoking in public' Is needed to protect

 

 

  
  

.; Hey DUde, if you do net want to smOke passively, it’s just easy. You can go to
‘ ,;;; places where smoking is prohibited. You guys are adults and you can decide
where to go. People CHOOSE either smoking places or non—smoking places. Thus,
N0 need for the state—wide ban! If enough people want to go to non-smoking
barsz business owners will set up non-smoking bars
1'11ng1) (MICHIGAN) wrote yesterday
LOL!!! In many places, there are NO non—smoking bars or restaurants!!! NONEII
Unless people decide not to go out with friends and stay home all the time, they
cannot avoid passive smoking. Also, workers in smoky work—places NEVER freely
d'Ioose this If no other jobs are available. Workers are exposed to other people’s
smoke but th couldn’t make choice out of their own will to avoid it.
(MICHIGAN) Wrote yesterday
What about jobs like mining or the armed forces?? Individuals decide those jobs in
smoky work places rather than not having a job at all. A state—wide ban on
smoking in public places would put many bars, pubs and clubs out of business.

  
  
   

   

 

 

 

Smokers would not Q to these places. My uncle owns a large bar in MN. After the

65

l

 

 

statewide ban in MN, his business has slashed by 50% due to the loss of smoking
customers. State—wide ban kills small businesses and all those workers. Pubs
provide jobs for people with few skills. A complete ban is not necessary —
ventilation fans can remove smoke.

 

-. déhnuzsttlimlfliil wrote yesterday
5 Every time I go out for dinner with my family, I saw foggy smoke all over the

l restaurant. I do not want to smell that crap. Smokers do have right to harm

 

 

 

" themselves but they don’t have right to harm others. It is more important to
protect people’s health than to protect businesses. Pubs and clubs should adapt,
for example, by trying to earn more money from selling food. After a ban in New
South Wales (Australia, there was only 9% of small businesses reported a drop in
sales. Some small business owners actually earn much more money by attracting
more and more non-smokersll

 

 

; La9t03d(MICHIGAN)wrobeyesterday . ,
That is totally bull shitll If you live in the city, you inhale the equivalent of 2
cigarettes a day in carcinogen from the SMOG, NOT from the SMOKEll There’s
"’ more carbon monoxide than u get from a cigarette - are people all dying from
lung cancer? ‘NO’ Therefore, state-wide smoking ban never make sense at all! If
you ban smoking, why don’t you ban all the baffics — causing smog??

 

66

 

 

 

SECTION 5 - The following questions ask about your perception about the 1st post.

 

   
 
  
 

 

GNGLOG (MICHIGAN) wrote yesterday . .
Smoking is dangerous not only for you but also for people near you who did not
CHOOSE to smokell Smoklng causes many diseases and harms to people nearby,
who just breathe in the smoke. Smokers choose to smoke, but people nearby DO
NOT choose to smoke. A complete ban on smoking in public is needed to protect
le's health from assive smoking

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
lst Post’ are positioned on this issue?

 

—3 —2 —1 0 1 2 3
Extremely Neutral Extremely
In favor of ,

. (Equally In favor of
Smokrng— .
ban— represent Smoking—

. both ban-opposing
supporting sides) group
group

2. Considering ‘believability’, in your opinion, ‘the 1st Post’ is...?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the lst Post’ is...?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the 1st Post’ is...?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 1st Post’ is...?
I) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

67

 

SECTION 6 - The following questions ask about your perception about the 2nd post.

 

2“ d 00CAT10 (MICHIGAN) wrote yesterday
Hey Dude, if you do not want to smoke passively, it’s just easy. You can go to
Post places where smoking is prohibited. You guys are adults and you can decide
where to go. People CHOOSE either smoking places or non-smoking places.
Thus, NO need for the state—wide ban! If enough people want to go to non—

smoking bars, business owners will set up non—smoking bars

 

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
2nd Post’ are positioned on this issue?

 

‘3 '2 -1 0 1 2 3
fixgiglfg} Neutral Extremely
Smoking— (Equally in favor of
ban— represent Smoking—
supporting bOth ban—opposing

sides) group
group

2. Considering ‘believability’, in your opinion, W is...?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the 2nd Post’ is...?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘faimess’, in your opinion, ‘the 2nd Post’ is...?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 2nd Post’ is...?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

68

 

SECTION 7 - The following questions ask about your perception about the 3rd post.
:( 0116M) it- :i _ 1 1;; y
LOL!!! In many places, there are NO non—smoking bars or restaurants!!! NONE!!
1,55: Unless people decide not to go out with friends and stay home all the time, they
3" canmt avoid passive smoking. Also, workers in smoky work—places NEVER freely
dwoose this if no other jobs are available. Workers are exposed to other people’s
smoke but t couldn’t make choice out of their own will to avoid it.

     
 
  
  

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
3rd Post’ are positioned on this issue?

 

—3 —2 —l 0 1 2 3
Extremely Neutral Extremely
In favor of . 1
. (Equally In favor of ‘
Smokmg— .
ban— represent Smoking—
. both ban—opposing
supportIng sides) group
group
2. Considering ‘believability’, in your opinion, ‘the 3rd Post’ is...?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable

5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the 3rd Post’ is. . .?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the 3rd Post’ is...?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 3rd Post’ is...?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

69

 

 

SECTION 8— The following uestions ask about your perception about the 4th post.
" " zeiesterday ' ,
What about jobs like min ng or the armed forces?? Individuals decide those jobs in
~ smoky work places rather than not having a job at all. A state-wide ban on
" smoking in public places would put many bars, pubs and clubs out of business.
Smokers would not go to these places. My uncle owns a large bar in MN. After the
state—wide ban in MN, his business has slashed by 50% due to the loss of
smoking customers. State—wide ban kills small businesses and all those workers.
Pubs provide jobs for people with few skills. A complete ban is not necessary —
ventilation fans can remove smoke.

  
   

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
4th Post’ are positioned on this issue?

 

—3 —2 —l 0 1 2 3
5223':in Neutral Extremely
Smokin — (Equally in favor of
ban— g represent Smoking.
su ortin b0“! ban—opposing

pp g sides) group
group

2. Considering ‘believability’, in your opinion, ‘the 4th Post’ is...?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the 4th Post’ is. . .?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the 4th Post’ is. . .?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 4th Post’ is...?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

70

 

- The following uestions ask about your perception about the 5th post.

  
   

Every time I go out for dinner with my family, I saw foggy smoke all over the
estaurant. I do not want to smell that crap. Smokers do have right to harm
' themselves but they don’t have right to harm others. It is more important to
protect people's health man to protect businesses. Pubs and clubs should adapt,
for example, by trying to earn more money from selling food. After a ban in New
South Wales (Australia, there was only 9% of small businesses reported a drop in
sales. Some small business owners actually earn much more money by attracting
more and more non—smokers”

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
5th Post’ are positioned on this issue?

 

—3 —2 —1 0 1 2 3
Exgig‘fg Neutral Extremely
Smokin _ (Equally in favor of
ban— g represent Smoking—
supporting bOth ban-opposing
group SIdes) group

2. Considering ‘believability’, in your opinion, ‘the 5th Post’ is. . .?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the 5th Post’ is...?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the 5th Post’ is. ..?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 5th Post’ is. . .?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

71

SECTION 10 - The following questions ask about your perception about the 6th post.

 

   
  

Lagt03d (MICHIGAN) wrote yesterday

That is totally bull shit!! If you live in the city, you inhale the equivalent of 2
cigarettes a day in carcinogen from the SMOG, NOT from the SMOKEH There’s
more carbon monoxide than u get from a cigarette - are people all dying from lung
cancer? ‘NO’ Therefore, state—wide smoking ban never make sense at all! If you
ban smoking, why don’t you ban all the traffics — causing smog??

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify ‘the
6th Post’ are positioned on this issue?

 

—3 —2 —1 0 1 2 3
Extremely Neutral Extremely
In favor of .

. (Equally In favor of
Smokmg— .
ban— represent Smoking—

. both ban-opposing
sup p0 rtIng sides) group
group

2. Considering ‘believability’, in your opinion, ‘the 6th Post’ is...?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the 6th Post’ is...?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the 6th Post’ is...?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

5. Considering ‘depth of knowledge’, in your opinion, ‘the 6th Post’ is...?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

72

 

SECTION 11 — The following questions ask about your perception about the ‘the
content (including brief summary about the topic and ALL arguments), as a whole.

 

How do YOU feel about a ‘Smoking Ban in Public Places, such as
restaurants and bars'? . ”WWW”

Topic: Smgking Bgn in Public Plgcgg — Should smoking be banned in

public places?

 

This discussion Board is for people who wants to argue on the issue, ‘Smoking Ban In
Public Places.’

A ‘Smoking ban in public places’ prohibits tobacco smoking in restaurants, bars,
workplaces and other public spaces to protect people from the effects of second—hand smoke.
While Michigan has long pursued the policy of letting the market rule on whether bars and
restaurants would allow smoking, recently, there have been many debates over whether there
should be a state-wide ban or not.

This issue is VERY controvertial and there are continuous debates between two groups of
people, “‘Smoking—ban—supporting" groups & “Smoking—ban—opposing" groups. Should
Smoking be banned in public places??? What is your oplnion about the issue in Michigan?

 

 

 

Displaying posts 1 — 6 out of 93,135 by 18,627 people.

 

  
 
   

GNGLOG (MICHIGAN) wrote yesterday

Smoking is dangerous not only for you but also for people near you who did not

; CHOOSE to smokell Smoking causes many diseases and harms to people nearby,

"'1 who just breathe in the smoke. Smokers choose to smoke, but people nearby DO
NOT choose to smoke. A complete ban on smoking in public is needed to protect
@gle’s health from assive smoking

' OOCATIO (MICHIGAN) wrote yesterday

Hey Dude, if you do not want to smoke passively, it’s just easy. You can go to

places where smoking is prohibited. You guys are adults and you can decide where

‘ to go. People CHOOSE either smoking places or non-smoking places. Thus, NO
need for the state-wide ban! If enough people want to go to non—smoking bars,
buslness owners will set up non-smoking bars

d THglxD (MICHIGAN) wrote yesterday 3

LOL!!! In many places, there are NO non—smoking bars or restaurants!!! NONEll
bsté, Unless people decide not to go out with friends and stay home all the time, they
’ cannot avoid passive smoking. Also, workers in smoky work-places NEVER freely

choose this if no other jobs are available. Workers are exposed to other people’s
smoke, but thgy couldn’t make choice out of their own will to avoid it.

l‘e (MIG-IIGAN) wrote yesterday

What about jobs like mining or the armed forces?? Individuals decide those jobs in

o . smoky work places rather than not having a job at all. A state-wide ban on

““ smoking in public places would put many bars, pubs and clubs out of business.

Smokers would not go to these places. My uncle owns a large bar in MN. After the

state—wide ban in MN, his business has slashed by 50% due to the loss of smoking _

73

 

  
   
 
   
 

 

   

 

  
   

 

 

 

 

 

 

 

customers. State—wide ban kills small businesses and all those workers. Pubs
provide jobs for people with few skills. A complete ban is not necessary —
ventilation fans can remove smoke.
v danuuz {MIGIGAM wrote yesterday
.. _ Every time I go out for dinner with my family, I saw foggy smoke all over the
-~-;§ restaurant. I do not want to smell that crap. Smokers do have right to harm
themselves but they don’t have right to harm others. It is more important to
protect people's health than to protect businesses. Pubs and clubs should adapt,
for example, by trying to earn more money from selling food. After a ban in New
South Wales (Australia, there was only 9% of small businesses reported a drop in
sales. Some small business owners actually earn much more money by attracting
more and more non—smokers!!
, Lagt03d (MICHIGAN) wrote yesterday
_ That is totally bull shitll If you live in the city, you inhale the equivalent of 2
..g cigarettes a day in carcinogen from the SMOG, NOT from the SMOKEH There’s
more carbon monoxide than u get from a cigarette — are people all dying from lung
cancer? ‘NO’ Therefore, state—wide smoking ban never make sense at all! If you
ban smoking, why don’t you ban all the traffies — causing smog??

 

 

 

 

 

 

1. Considering the issue of “Smoking Ban in public place,” how do you identify “the contentI
as a whole’ is positioned on this issue?

 

‘3 -2 —1 0 1 2 3
5121;512:313; Neutral Extremely
Smokin ban— (Equally in favor of
su ortii— represent Smoking—ban—

pp g both sides) opposing group

group

2. Considering ‘believability’, in your opinion, ‘the content, as a whole’ is. . .?
1) Not at all believable
2) Somewhat believable
3) Moderately believable
4) Very believable
5) Extremely believable

3. Considering ‘accuracy’, in your opinion, ‘the content, as a whole’ is. . .?
1) Not at all accurate
2) Somewhat accurate
3) Moderately accurate
4) Very accurate
5) Extremely accurate

4. Considering ‘fairness’, in your opinion, ‘the conteng as a whole' is. ..?
1) Not at all fair
2) Somewhat fair
3) Moderately fair
4) Very fair
5) Extremely fair

74

 

 

 

 

5. Considering ‘depth of knowledge’, in your opinion, ‘the content, as a whole’ is. . .?
1) Not at all knowledgeable
2) Somewhat knowledgeable
3) Moderately knowledgeable
4) Very knowledgeable
5) Extremely knowledgeable

This ends the survey.
Thank you!

75

 

 

REFERENCES

 

Abrams, D. B., Orleans, C. T., Niaura, R. S., Goldstein, M. G., Prochaska, J. 0., &

Velicer, W. (1996). Integrating individual and public health perspectives for treatment of

tobacco dependence under managed health care: a combined stepped—care and matching

model. Annals of Behavioral Medicine, 18(4), 290—304. ,A‘

Barbeau, E. M., Delaurier, G., Kelder, G., McLellan, D., Sorensen, G., Balbach, E. D., et
al. (2007). A Decade of Work on Organized Labor and Tobacco Control: Reflections on
Research and Coalition Building in the United States. Journal of Public Health Policy,
28(1), 118—135.

Bates, B. R., Romina, S., Ahmed, R., & Hopson, D. (2006). The effect of source
_ credibility on consumers' perceptions of the quality of health information on the Internet.
. Medical Informatics & The Internet in Medicine, 31(1), 45—52.

Bauer, R. A. (1973). The Audience. 8.: 141—152. Pool, 1. de Sola (Hrsg.), Handbook of
Communication. Chicago: Rand McNalIy.

Baumeister, R. F., & Newman, L. S. (1994). Self—regulation of cognitive inference and
decision processes. Personality and Social Psychology Bulletin, 20(1), 3.

Berland, G. K., Elliott, M. N., Morales, L. S., Algazy, J. 1., Kravitz, R. L., Broder, M. S.,
et al. (2001). Health information on the Internet accessibility, quality, and readability in
English and Spanish. Journal of the American Medical Association, 285, 2612—262 1 .

Biek, M., Wood, W., & Chaiken, S. (1996). Working knowledge, cognitive processing,
and attitudes: On the determinants of bias. Personality and Social Psychology Bulletin,
22(6), 547.

 

Bucy, E. P. (2003). Media Credibility Reconsidered: Synergy between On—Air and
Online News. Journalism and Mass Communication Quarterly, 80(2), 247-264.

Bylund, C. L., Sabee, C. M., Imes, R. S., & Sanford, A. A. (2007). Exploration of the
construct of reliance among patients who talk with their providers about intemet
information. Journal of health communication, 12(1), 17—28.

CEPA (Canadian Environmental Protection Act)- Federal. (1999). Provincial Advisory

. Committee Working Group on Air Quality Objectives and Guidelines. National Ambient
Air Quality Objectives for Particulate Matter, Science Assessment Document. Ottawa,

Canada. Health Canada and Environment Canada.

76

 

Chaffee, S. H. (1982). Mass media and interpersonal channels: Competitive, convergent,
or complementary. Inter/Media: Interpersonal communication in a media world, 57.

Chaffee, S. H., Saphir, M. N., Graf, J., Sandvig, C., & Hahn, K. S. (2001). Attention to
counter—attitudinal messages in a state election campaign. Political Communication,
18(3), 247—272.

Chapman, G. B., & Elstein, A. S. (2003). 7 Cognitive Processes and Biases in Medical
Decision Making. Decision making in health care: Theory, psychology, and applications,
183.

Chinn, C. A., & Brewer, W. F. (1993). The role of anomalous data in knowledge
acquisition: A theoretical fi'amework and implications for science instruction. Review of
educational research, 63(1), 1.

Choi, J. H., Watt, J. H., & Lynch, M. (2006). Perceptions of news credibility about the
war in Iraq: Why war opponents perceived the Internet as the most credible medium.
Journal of Computer—Mediated Communication, 12(1), 209—229.

Christen, C. T., Kannaovakun, P., & Gunther, A. C. (2002). Hostile media perceptions:
Partisan assessments of press and public during the 1997 United Parcel Service strike.
Political Communication, 19(4), 423—436.

Cline, R. J. W., & Haynes, K. M. (2001). Consumer health information seeking on the
Internet: the state of the art (Vol. 16, pp. 671—692): Oxford Univ Press.

Dahlberg, L. (2001). The Internet and democratic discourse: Exploring the prospects of
online deliberative forums extending the public sphere. Information, Communication &
Society, 4(4), 615—633.

Dalton, R. J ., Beck, P. A., & Huckfeldt, R. (1998). Partisan cues and the media:
Information flows in the 1992 presidential election. American Political Science Review,
111-126.

Ditto, P. H., & Lopez, D. F. (1992). Motivated skepticism: Use of differential decision
criteria for preferred and nonpreferred conclusions. Journal of Personality and Social
Psychology, 63(4), 568—584.

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes: Harcourt Brace
Jovanovich College Publishers Fort Worth, TX.

Eastin, M. S. (2001). Credibility assessments of online health information: the effects of

source expertise and knowledge of content. Journal of Computer—Mediated
' Communication, 6(4).

77

 

Eighmey, J. (1997). Profiling User Responses to Commercial Web Sites. Journal of
Advertising Research, 3 7(3).

Eng, T. R., Maxfield, A., Patrick, K., Deering, M. J., Ratzan, S. C., & Gustafson, D. H.
(1998). Access to Health Information and Support A Public Highway or a Private Road? ,
Journal of the American Medical Association, 280, 1371—1375.

Eysenbaoh, G., & Kohler, C. (2002). How do consumers search for and appraise health
information on the world wide web? Qualitative study using focus groups, usability tests,
and in—depth interviews. British Medical Journal, 324(7337), 573—5 77.

F lanagin, A. J., & Metzger, M. J. (2001). Internet use in the contemporary media
environment. Human Communication Research, 27(1), 153-181.

Fogg, B. J., Kameda, T., Boyd, J ., Marshall, J., Sethi, R., Sockol, M., et a1. (2002).
Stanford-Makovsky Web Credibility Study 2002: Investigating what makes Web sites
credible today, A Research Report by the Stanford Persuasive Technology Lab &
Makovsky & Company. URL: http://www. webcredibility. org/pdf/Stanford—
Makovsky WebCredStu—ayZOOZ—prelim. pdf (2002).

Fox, S. (2006). Online health search 2006. Pew Internet & American Life Project.

Friedlander, M. L., & Phillips, S. D. (1984). Preventing anchoring errors in clinical
judgment. Journal of consulting and clinical psychology, 52(3), 366—37 1 .

Gaziano, C., & McGrath, K. (1986). Measuring the Concept of Credibility. Journalism
Quarterly.

Giner—Sorolla, R., & Chaiken, S. (1994). The causes of hostile media judgments. Journal
of experimental social psychology/(Print), 30(2), 165—180.

Glowniak, J. V. (1995). Medical resources on the Internet. Annals of Intetnal Medicine,
123, 123—131.

Gunther, A. C. (1992). Biased press or biased public? Attitudes toward media coverage
of social groups. Public Opinion Quarterly, 56(2), 147.

Gunther, A. C., & Christen, C. T. (2002). Projection or persuasive press? Contrary effects
of personal opinion and perceived news coverage on estimates of public opinion. The
Journal of Communication, 52(1), 177—195.

Gunther, A. C., Christen, C. T., Liebhart, J. L., & Chih-Yun Chia, S. (2001). Congenial

Public, Contrary Press, and Biased Estimates of the Climate of Opinion. Public opinion
quarterly, 65(3), 295-320.

78

 

 

Gunther, A. C., & Liebhart, J. L. (2006). Broad reach or biased source? Decomposing the
hostile media effect. Journal of Communication, 56(3), 449—466.

Gunther, A. C., & Schmitt, K. (2004). Mapping boundaries of the hostile media effect.
The Journal of Communication, 54(1), 55-70.

Hagel, J ., & Armstrong, A. G. (1996). The real value of on—line communities. Harvard
Business Review, 74(3), 134—141.

Hogarth, R. M. (1980). Judgement and choice: The psychology of decision: John Wiley
& Sons.

Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on
communication effectiveness. Public Opinion Quarterly, 15(4), 635—650.

Impicciatore, P., Pandolfini, C., Casella, N., & Bonati, M. (1997). Reliability of health
information for the public on the World Wide Web: systematic survey of advice on
' managing fever in children at home, British Medical Journal, 314, 187 5 .

International Agency for Research on, C. (1986). IARC monographs on the evaluation of
the carcinogenic risk of chemicals to humans.

Jadad, A. R., & Gagliardi, A. (1998). Rating Health Information on the Internet
Navigating to Knowledge or to Babel? Jama, 279(8), 611—614.

Johnson, T. J ., & Kaye, B. K. (2002). Webelievability: A path model examining how
convenience and reliance predict online credibility. Journalism and Mass Communication
Quarterly, 79(3), 619—642.

- Jones, A. P. (1999). Indoor air quality and health. Atmospheric Environment, 33(28),
4535—4564.

Katz, E., Blumler, J. G., & Gurevitch, M. (1974). Utilization of mass communication by
the individual. The uses of mass communications: Current perspectives on gratifications

' research, 3, 19—32.

Keselman, A., Browne, A. C., & Kaufman, D. R. (2008). Consumer health information
seeking as hypothesis testing. Journal of the American Medical Informatics Association,
15(4), 484—495.

Kiousis, S. (2001). Public trust or mistrust? Perceptions of media credibility in the
information age. Mass Communication and Society, 4(4), 381—403.

Kiousis, S. (2003). Job approval and favorability: The impact of media attention to the

Monica Lewinsky scandal on public opinion of President Bill Clinton. Mass
Communication and Society, 6(4), 43 5—45 1 .

79

 

 

Kiousis, S. (2006). Exploring the impact of modality on perceptions of credibility for
online news stories. Journalism Studies, 7(2), 348—359.

_ Ko, H., Cho, C. H., & Roberts, M. S. (2005). Internet uses and gratifications: a structural
equation model of interactive advertising. Journal of Advertising, 34(2), 57—70.

Kuehn, S. A. (1994). Computer-mediated communication in instructional settings: A
research agenda. Communication Education, 43(2), 171—183.

Kuhn, D. (2001). How do people know? Psychological Science, 12(1), 1—8.

Kunda, Z. (1990). The case for motivated reasoning. Psychological bulletin, 108(3), 480—
498.

Larkin, M. (1999). What's amiss with web search engines? Lancet, 354(9174), 260.

LaRose, R., Mastro, D., & Eastin, M. S. (2001). Understanding Internet usage: a social—
cognitive approach to uses and gratifications. Social Science Computer Review, 19(4),
395.

Lee, E., & Leets, L. (2002). Persuasive storytelling by hate groups online: Examining its
effects on adolescents. American Behavioral Scientist, 45(6), 927.

Lewis, P. H. (1996). F ollow-up survey reports growth in Internet users. New York Times,
August 14:02.

Lin, C. A. (1996). Standpoint: Looking back: The contribution of Blumler and Katz's
uses of mass communication to communication research. Journal of Broadcasting &
Electronic Media, 40(4), 574—581.

Lodge, M., & Taber, C. (2000). Three steps toward a theory of motivated political
reasoning. Elements of reason: Cognition, choice, and the bounds of rationality, 183—213.

Lorence, D. P., & Greenberg, L. (2006). The zeitgeist of online health search:
Implications for a consumer—centric health system. Journal of General Internal Medicine,
21(2), 134.

Mahoney, M. J ., & DeMonbreun, B. G. (1977). Psychology of the scientist: An analysis
of problem—solving bias. Cognitive Therapy and Research, 1(3), 229—238.

Matheson, K., & Dursun, S. (2001). Social identity precursors to the hostile media

phenomenon: Partisan perceptions of coverage of the Bosnian conflict. Group Processes
& Intergroup Relations, 4(2), 116.

McClung, H. J ., Murray, R. D., & Heitlinger, L. A. (1998). The Internet as a source for

I . current patient information. Pediatrics in Review, 101.

80

 

 

McMullan, M. (2006). Patients using the Internet to obtain health information: How this
affects the patients health professional relationship. Patient Education and Counseling,
63(1—2), 24—28.

Metzger, M. J., Flanagin, A. J ., & Zwarun, L. (2003). College student Web use,
perceptions of information credibility, and verification behavior. Computers & Education,
41(3), 271-290.

Meyer, P. (1988). Defining and measuring credibility of newspapers: Developing an
index. Journalism Quarterly, 65(3), 567—574.

. Morris, M., & Ogan, C. The Internet as mass medium. McQuail's reader in mass
communication theory, 134—145.

National Toxicology Program 2005. Toxicological Evaluation of Nanoscale Materials:
Fact Sheet. Research Triangle Park, NC:National Toxicology Program. Available:
http://ntp.niehs.nih.Q/ntp/htdocg/jLason/fgctsheets/nanoscaleFacts.pdf [accessed 20
January 2005].

Newhagen, J ., & Nass, C. (1989). Differential Criteria for Evaluating Credibility of
Newspapers and TV News. Journalism Quarterly, 66(2), 27 7—284.

Newhagen, J. E., & Rafaeli, S. (1996). Why communication researchers should study the
Internet: A dialogue. Journal of computer—mediated communication, 1(4).

Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises.
Review of General Psychology, 2, 175—220.

Palmgreen, P., Wenner, L. A., & Rosengren, K. E. (1985). Uses and gratifications
research: The past ten years. Media gratifications research: Current perspectives, l l—17.

Perloff, R. M. (1989). Ego—involvement and the third person effect of televised news
coverage. Communication Research, 16(2), 236.

Ries, L. A. G., Eisner, M. P., Kosary, C. L., Hankey, B. F., Miller, B. A., Clegg, L., et al.
SEER cancer statistics review, 1975?001. Bethesda, MD: National Cancer Institute, 2004.
Links.

Rubin, A. M. (2002). The uses—and—gratifications perspective of media effects. Media
effects: Advances in theory and research, 2, 525—548.

Schmitt, K. M., Gunther, A. C., & Liebhart, J. L. (2004). Why partisans see mass media
as biased. Communication Research, 31(6), 623.

81

 

 

 

Schweiger, W. (2000). Media Credibility—Experience or Image?: A Survey on the
Credibility of the World Wide Web in Germany in Comparison to Other Media.
European Journal of Communication, 15(1), 37.

Shaklee, H., & F ischhoff, B. (1982). Strategies of information search in causal analysis.
. Memory & Cognition, 10(6), 520—530.

Silverman, B. G. (1992). Judgment Error and Expert Critics in Forecasting Tasks“.
Decision Sciences, 23(5), 1199—1219.

Skov, R. B., & Sherman, S. J. (1986). Information-gathering processes: diagnosticityl
, hypothesis—confirmatory strategies, and perceived hypothesis confirmation. Journal of
experimental social psychology(Print), 22(2), 93—121.

Slater, M. D., & Rouner, D. (2002). Entertainment—education and elaboration likelihood:
Understanding the processing of narrative persuasion. Communication Theory, 12(2),
173-191.

Smoke, E. T. (1986). Measuring exposures and assessing health effects. National
7 Research Council, Washington DC.

Snyder, M., & Campbell, B. (1980). Testing hypotheses about other people: The role of
the hypothesis. Personality and Social Psychology Bulletin, 6(3), 421.

Snyder, M., & Swann, W. B. (1978). Hypothesis-testing processes in social interaction.
Journal of Personality and Social Psychology, 3 6(1 1), 1202—1212.

Sonnenberg, F. A. (1997). Health information on the Internet. Opportunities and pitfalls.
Archives of Internal Medicine, 157(2), 151—152.

Strauss, K. (1997). Quality of medical information on the intemet. JAM4, 278(8), 632—
' 632.

Sundar, S. S. (1999). Exploring Receivers. Journalism and Mass Communication
Quarterly, 76(2), 373—386.

Sundar, S. S., & Nass, C. (2001). Conceptualizing sources in online news. The Journal of
Communication, 51(1), 52—72.

Tseng, S., & Fogg, B. J. (1999). Credibility and computing technology. Communications
of the ACM, 42(5), 39—44.

Vallone, R. P., Ross, L., & Lepper, M. R. (1985). The hostile media phenomenon: Biased

perception and perceptions of media bias in coverage of the Beirut massacre. Journal of
Psrsonality and Social Psychology, 49(3), 577—585.

82

 

 

‘rwfi4V v

Viswanath, K. (2006). Public communications and its role in reducing and eliminating
health disparities. Examining the health disparities research plan of the National
Institutes of Health: unfinished business. Washington, DC: Institute of Medicine, 215—
253.

Wanta, W. (1997). The public and the national agenda: How people learn about
important issues: Lawrence Erlbaum.

Wason, P. C. (1960). On the failure to eliminate hypotheses in a conceptual task. The
Quarterly Journal of Experimental Psychology, 12(3), 129—140.

Wilson, N., & Thomson, G. (2002). Still dying from second—hand smoke at work: a brief
review of the evidence for smoke—free workplaces in New Zealand. Journal of the New
Zealand Medical Association, 115, 1165.

Wyatt, J. C. (1997). Commentary: measuring quality and impact of the World Wide Web.
British Medical Journal, 314, 1879.

York, S. (2009). The Flint Journal. URL:
http://wwwmlive.com/news/flint/indexssf/2009/O1/as_smoking_ban_battle_again_he.ht
m1.

Zanna, M. P., & Del Vecchio, S. M. (1973). Perceived credibility of television news: A
matter of viewer's attitudes and the position taken by the media. European Journal of
Social Psychology, 3(2).

83