This paper was presented at the Association for Education in Journalism
and Mass Communication in Toronto, Canada, August 2004.
If you have questions about this paper, please contact the author
directly. If you have questions about the archives, email
[log in to unmask] For an explanation of the subject line, send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
(Oct 2004)
Thank you.
Elliott Parker
************************************************************************
Framing SARS in Toronto:
A comparative analysis of the World Journal and the Toronto Star
David C. Oh and Wanfeng Zhou
S.I. Newhouse School of Public Communications
Syracuse University
Syracuse, New York
Mailing Address:
David C. Oh & Wanfeng Zhou
Newhouse School
Syracuse, NY 13244
Phone: 315-443-2370
Email: [log in to unmask] & [log in to unmask]
Paper submitted to the Mass Communication and Society Division and
in consideration for the Leslie J. Moeller Award,
Association for Education in Journalism and Mass Communication,
April 2004
Abstract
Our framing analysis compared the frames used in the World Journal and the
Toronto Star. We used four of the five frames identified by Semetko and
Valkenburg (2000): conflict, responsibility, economic consequences, and
human-interest. We also studied the use of risk frames. The Toronto Star
used the conflict, responsibility, economic consequences, and
human-interest frames more than the World Journal; whereas, the World
Journal used the risk frame more than the Toronto Star.
Framing SARS in Toronto: A comparative analysis
of the World Journal and the Toronto Star
The spread of Severe Acute Respiratory Syndrome (SARS) around the globe has
created international alarm and generated significant media coverage. This
attention appears warranted because SARS is an atypical pneumonia that is
fatal in a small proportion of the population. Perhaps, more alarming is
the speed in which SARS spread from the Guangdong Province in southern
China to 29 nations around the world (Summary, 2003). The nation greatest
affected was China with SARS reported in about two-thirds of China's
provinces (Update 42, 2003) and killing more than 829 people in China alone
(Summary, 2003).
Although SARS' origins are located primarily in Southeast Asia, communities
across the globe have been affected. Toronto was one such area. In fact, it
has the distinction of being the only non-Asian region to have a serious
outbreak. Conservative estimates of the spread of SARS in Toronto is
roughly 251 cases and more than 41 deaths (Summary, 2003) with liberal
estimates at more than 200 cases and almost 35 deaths (Meinbarids, 2003;
Travel, 2003). The outbreak was so serious the World Health Organization
(WHO) issued a travel advisory against Toronto on April 23, 2003 (Travel,
2003). This was the first instance of a travel advisory against a North
American city in the 55-year history of the WHO.
On top of the health emergency caused by SARS, there were also serious
economic repercussions. In Canada, Finance Minister John Manley said SARS
has had a significant negative effect on Canada's economic growth
(Meinbardis, 2003). Conservative estimates of SARS's impact on the economy
have been reported as more than US $2 billion in lost growth (Ferguson,
2003). Toronto's medical community alone is projected to endure an
estimated cost of US $100 million to treat the disease (Carey, 2003).
The merchants and restaurant owners in Toronto also saw significant drops
in revenue as the fear of SARS spread and as individuals mistakenly linked
ethnic Chinese individuals and institutions with SARS. Many of the 400,000
ethnic Chinese residents in Toronto were hurt economically as Toronto
residents avoided businesses and restaurants in Chinatown. Some Chinese
Canadian businesses saw a drop of between 30-70% in revenues (Sorenson,
2003). This prompted the Toronto Chinese Business Association to hold a
press conference to correct the mistaken belief that SARS was prevalent in
Chinatown (Keung, 2003).
For citizens of Toronto, perhaps because they suffered under the stress of
SARS more than other communities outside Asia, the reaction was more
visceral and negative. Asian Canadians were not only shunned, but they
endured racist tirades as well (Sorenson, 2003). One particularly hurtful
insult was that of Chinese as dirty, which is reminiscent of the "Yellow
Peril" days when Chinese Americans in San Francisco were accused of
spreading exotic diseases (Rider, 2003). Even Energy Minister John Baird
was reported as having made an insensitive joke after coughing, stating
sarcastically that he had enjoyed his recent trip to Asia (Rider, 2003).
Also hurtful have been false e-mail messages that accused certain
businesses of spreading the disease (Keung, 2003). During this period,
Asian Canadian leaders also were critical of the perceived silence of
government officials, saying this allowed for growing "anti-Chinese"
sentiment (Rider, 2003).
While the government has been criticized by the Asian Canadian activists,
the question is whether media played a role in exaggerating fears of SARS
and in the indirect participation of breeding racism. It has been indicated
in the literature on race that mass-market media's coverage of minorities
tends to be negative (Vergeer, Lubbers, & Scheepers, 2000) and that at
least in terms of U.S. news, it reinforces the dominant position of the
White majority (Shah & Thornton, 1994). Research into the portrayal of
Asians and Asian Americans shows that media continue to promote stereotypes
of Asians as violent, mysterious, sexually deviant, foreign, and so on.
(Hamamoto, 1994; Orbe, Seymour, & Kang, 1998; Shim, 1998; Wilson &
Gutierrez, 1995).
The research on ethnic media is not clear on its portrayal of the ingroup
ethnic group. For example, there is little research about the portrayals of
the Chinese Canadian community by the Chinese Canadian ethnic press. Most
research is in the uses and gratifications tradition, suggesting that
ethnic media is used to maintain ethnic identity (Jeffres, 2000). This does
not explicitly point to coverage in ethnic media; however, it can be
extrapolated that if individuals are using the ethnic media to maintain
ethnic identity that the content is fundamentally different from
mass-market media to the extent that it is used to maintain ethnic
identity. Because of the different audiences and purposes of mass-market
and ethnic press, we suspect that coverage of SARS will be framed
differently. We also suspect that these differences will be exacerbated in
times when ingroup-outgroup reactions in the dominant society are
strongest, including the initial outbreak, the WHO travel advisory, and the
second outbreak.
We believe this study is worthy of investigation for a number of reasons.
The first is there is a general dearth of research on ethnic Asian
populations in the West. This invisibility would suggest that this is an
unaffected or unimportant group, and we disagree vehemently with either
assumption. Therefore, we hope to contribute to this literature as well as
to the larger conversation on race and media. Further, there are relatively
few studies on the ethnic press, and of those studies, most have used
surveys and are focused on media use and its effects on ethnic identity. We
believe that more research on the content of ethnic media is necessary as
this is an important resource for ethnic communities. Further, we were
unable to find any comparative studies of ethnic and mass-market media, and
we believe our research will take a step towards filling that void.
Finally, SARS has had a major impact on Toronto, and we believe it is
important to investigate how media covered it. This leads us to the
following question.
RQ: To what extent does the framing of SARS in the World Journal, an ethnic
Chinese newspaper and the Toronto Star, a mass-market newspaper, differ?
THEORY
Media frames
Journalists use media frames to construct news stories. Because frames
provide a thematic structure for a story, frames affect not only the
journalists' construction of news but also the readers' understanding of
the event (Pan & Kosicki, 1993). Therefore, it is important to understand
how frames are used.
Entman (1993) writes that frames have not been clearly conceptualized and
that there needs to be a movement to unify the framing paradigm. Though
Scheufele (1999) proposed a process model of framing, which links both the
sociological processes of producing media frames and the ways in which
frames affect the audience member, there is still no clear consensus on
what the frames are. In a content analysis of European politics in news,
Semetko and Valkenburg (2000) took an important step in doing this,
identifying five frames employed frequently in news. These include
conflict, responsibility, economic consequences, morality, and human-interest.
The authors define the conflict frame as a frame that "emphasizes conflict
between individuals, groups, or institutions as a means of capturing
audience interest" (Semetko & Valkenburg, 2000, p. 95). The human-interest
frame is defined as a "frame that brings a human face or an emotional angle
to the presentation of an event, issue, or problem" (Semetko & Valkenburg,
2000, p. 95). Third is the economic consequences frame, which is defined as
a frame that "reports an event, problem, or issue in terms of the
consequences it will have economically on an individual, group,
institution, region, or country" (Semetko & Valkenburg, 2000, p. 96). The
next is the morality frame, which is defined as a frame that "puts the
event, problem, or issue in the context of religious or moral
prescriptions. Finally, the responsibility frame is defined as one that
"presents an issue or problem in such a way to attribute responsibility for
its cause or solution to either the government or to an individual or
group" (Semetko & Valkenburg, 2000, p. 96). The only other frames we
identified in the literature had to do with types of frames such as
episodic and thematic (Iyengar, 1991) and syntactic, script, and rhetorical
frames (Pan & Kosicki, 1993).
In relation to minority racial and ethnic groups, there has not been much
literature on the framing of news stories. Most of the scholarship is
content analyses or textual analyses of media messages. This belies the
importance of framing in the coverage of ethnic and racial minorities.
Gandy, Kopp, Hands, Frazer, & Phillips (1997) have found that frames shape
impressions of racial minorities, which then leads to the perceived
fairness of public policy. Their research also suggests that individual
attribution of fault leads to little public support, that it is problematic
for members of groups to be identified primarily as victims, and that
subtle differences in framing can influence perceived risks (Gandy, et al.,
1997). With SARS, the worldwide spread of the disease was believed to have
originated in China. In the case of Toronto, a Chinese Canadian, who had
visited Hong Kong, was believed to have spread the virus. As stated
earlier, this association of SARS and China had harmful effects on the
Chinese Canadians living in Toronto.
Along with the racial overtones of SARS, there are also health risks that
are framed by the mass-market press. Lupton (1993) writes that the press is
highly influential in the sociology of risk. Despite this important role,
journalists tend to simplify the analysis of health risks (Wilkins &
Patterson, 1987). They typically cover stories that are event-focused and
novel (Gandy, et al., 1997; Wilkins and Patterson, 1987). Further, because
stories are not often placed in a broader context, the stories create
confusion about the relative risks, often overrepresenting threat (Lupton,
1993; Wilkins & Patteron, 1987). Finally, the blame for health risks is
typically attributed to the stigmatized individual (Lupton, 1993).
Ethnic Chinese and the press
Ethnic media are thought to serve two purposes: assimilation and pluralism
(Viswanath & Arora, 2000; Lam, 1980). Ethnic media provide information for
new arrivals to help them adapt to living in their host country, fulfilling
an assimilationist function (Lam, 1980; Lee & Tse, 1994). Ethnic media also
maintain pluralism by reinforcing and sustaining ethnic identity (Jeffres,
2000) and help satisfy their readers' need to maintain ties to their
homeland (Lee & Tse, 1994). For ethnic Chinese in Canada, this is an
important source of information and entertainment. In fact, ethnic Chinese
have a higher readership rate than any other immigrant ethnic group,
spending 41 percent of their media time on ethnic media (Lam, 1980; Lee &
Tse, 1994).
Another reason Chinese Canadians turn to ethnic media might be because of
the problematic treatment of ethnic minorities in the mass-market press.
Research has shown that they are either marginalized or misrepresented
(Dunn & Mahtani, 2001; Mahtani, 2001; Ungerleider, 1991). Misrepresentation
of ethnic minorities includes negative stereotypes of minority groups as
mysterious and incompatible with the dominant culture (Mahtani, 2001) and
portrayals of minority groups as threats with an overt positioning of "us"
and "them" (Fleras & Kunz, 2001).
Historically, newspaper portrayals in the 1800s were clearly anti-Chinese
(Dunn & Mahtani, 2001; Mahtani, 2001). More recently, Asian Canadians have
been viewed as a threat with stories about Asian gangs and with portrayals
of Asian Canadians as a social problem facing "real" Canadians (Dunn &
Mahtani, 2001). All of these examples have an undertone of the "yellow
peril" coverage in the 1800s in Californian newspapers. Greenburg (2000)
found this as well in an analysis of the "opinion" discourses surrounding
the 1999 arrival of Fujianese refugees in British Columbia, writing that
the papers were critical of the refugees with the key themes that the
refugees are carriers of disease and danger and that they pose a
significant threat to the moral, physical, and economic well-being of
"legitimate" Canadians.
Explanations
There are several competing explanations for the marginalization and
harmful portrayals in the mass-market press. A social psychological
perspective focuses on the dynamics of ingroups and outgroups. In social
identity theory, there are two primary processes (Stets & Burke, 2000). The
first is self-categorization. This is when individuals accentuate the
similarities with their ingroup and the differences with outgroups. The
other process is self-comparison in which individuals judge the
similarities for the ingroup as positive and the differences for outgroups
as negative. One of the clearest dividing lines of ingroups and outgroups
in North American cultures is race. Given that ownership of mass-market
media belongs predominantly to ethnic Europeans in Canada, it would not be
surprising that media coverage would favor the ingroup of European Canadians.
In fact, this is what the research has shown. The earlier examples point
to the marginalization of outgroup minority populations and the negative
portrayals when they are covered in the Canadian news media, while research
in the United States has also shown similar findings. Ingroup members of
the dominant society find themselves highly individuated and impossible to
stereotype (Entman, 1992). This is in contrast to the perception that
outgroups are a homogenized group and that there is truth to the
stereotypes (Entman, 1992). Further, outgroups are seen as having interests
that conflict with the ingroup and being worthy of blame for their problems
(Romer, Jamieson, & de Coteau, 1998).
Anthropologists in the middle of the twentieth century argued that
societies are not closed systems but, rather, are dynamic systems in which
conflict and change are the norms (Lewellen, 1992). This is especially true
for multiethnic societies, where conflict defines the nature of social
relationships. Realistic group conflict theory states that disruptions and
threats such as competition for scarce resources can escalate the conflict
between groups (Tajfel & Turner, 1979). It is likely that the introduction
of a disease like SARS also exacerbates conflict because of the threat
posed. This leads to our first set of hypotheses.
H1a: The Toronto Star will use the conflict frame more than the World Journal.
H1b: During times of crisis, including the initial outbreak of SARS, the
WHO travel advisory, and the second outbreak of SARS, the greater use of
the conflict frame by the Toronto Star will be more pronounced.
Realistic group conflict theory suggests that the conflict between the
dominant White majority and the Chinese ethnic minority will be disrupted
with the introduction of the SARS threat. Because this disruption is
thought to have been introduced by an ethnic Chinese woman and because the
mass-market press reflects the interest of the dominant group, mass-market
newspapers will employ the conflict frame more often, especially in times
of exaggerated conflict, e.g. the start of the SARS outbreak in Toronto,
the WHO travel advisory, and the second outbreak. The ethnic minority, on
the other hand, will try to minimize conflict. Because of the ethnic
press's assimilationist role, it is expected that conflicts within the
Toronto community and between Canada and China will be downplayed.
Conceptually, we define the use of the conflict frame as a frame that
portrays two or more parties in conflict. Operationally, this includes
direct conflict in the form of physical, verbal, emotional, or
psychological confrontation as well as indirect conflict in the form of
ideological, economic, personal, professional, and religious conflict.
For hypothesis 1b, we define the SARS outbreak in Toronto as March 16,
2003, when the first case of SARS was reported in an elderly Chinese
Canadian resident of Toronto who had visited Hong Kong. The WHO travel
advisory occurred during the timeframe between April 23, 2003, and April
30, 2003. The second outbreak of SARS began on May 26, 2003. All of these
are in accordance with information found on the WHO homepage.
Operationally, we define times of crisis periods as coverage which follows
one week after the initial outbreak, the WHO travel advisory, and the
second outbreak.
H2a: The Toronto Star will use the responsibility frame more than the World
Journal.
H2b: The Toronto Star will blame those people and institutions that are
ethnically Chinese more than the World Journal.
H2c: The World Journal will blame those people and institutions that are
not ethnically Chinese more than the Toronto Star.
The literature on health and risk states that the press seeks to assign
blame. Social identity theory suggests that ingroups tend to blame
outgroups for social problems. Because SARS was traced to China and because
SARS in Toronto was traced to an ethnic Chinese Toronto resident who had
visited Hong Kong, we believe the press will assign blame to China. Because
Chinese are an ethnic outgroup in Toronto, it is expected that the Toronto
Star will use this frame more often. However, because Chinese are an ethnic
ingroup for the ethnic press and because Chinese Canadians seek to minimize
conflict, it is expected that the responsibility frame will be employed
less often. The World Journal, because it reflects the values of its
readership, will not assign responsibility to China or Chinese often.
We define the use of the responsibility frame as a frame that either shows
one party assigning responsibility to another(s) for an issue or problem or
a frame in which the author assigns blame. Ethnic Chinese is defined as
those individuals, groups, institutions, and governmental bodies that are
of Chinese origin. This includes the nation of China, any Chinese
governmental bodies or institutions, any Chinese industries, and any ethnic
Chinese individuals, groups, and businesses located outside of China.
Non-ethnic Chinese are simply any individuals, groups, institutions, and
governmental bodies that are not of Chinese origin.
H3a: The World Journal will use the economic consequences frame more than
the Toronto Star.
H3b: The World Journal will use the economic consequences frame more on the
micro level than the Toronto Star.
H3c: The Toronto Star will use the economic consequences frame more on the
macro level than the World Journal.
While SARS had negative economic consequences for the dominant majority,
those most personally affected on an economic level have been Chinese
Canadians. As indicated earlier, Chinese Canadians in Chinatown saw
revenues drop anywhere from 30-70%. Though many in the dominant group were
affected as well, especially in the hospitality and tourism sectors, the
dominant group, on the whole, was not as negatively affected. Therefore,
the ethnic press will use this frame more often because of ingroup
interests and strengthening of ethnic bonds, thus serving its other major
function, pluralism.
The economic consequences frame is one in which a news story is written,
describing how, to what extent, and in what ways an event, issue, or
problem affects the economy or finances of a nation(s), group(s),
institution(s), industry(ies), or individual(s). We define the micro level
as economic consequences that occur below the citywide level. This includes
local groups, local industries, local institutions, neighborhoods, and
individuals.
H4: The World Journal will use the human-interest frame more often than the
Toronto Star.
Because Chinese Canadians were victims of racist acts and were economically
damaged and verbally assaulted as a group because of SARS's association to
China, the ethnic press will employ greater use of the human-interest
frame. This is consistent with its pluralistic role in strengthening group
identity. Further, the ethnic press will use this frame more often because
of its assimilationist role in order to solidify the fact that ethnic
Chinese, like the community at large, are similarly affected by SARS.
We define this frame as one in which an individual or group of individuals
are highlighted for no other purpose than to bring an emotional dimension
to the story. In other words, stories in which their inclusion or exclusion
does not change the factual nature of the story are coded as human-interest.
H5a: The Toronto Star will use the risk frame more often than the World
Journal.
H5b: The Toronto Star will more often describe the risk as increasing than the
World Journal.
H5c: The World Journal will more often describe the risk as decreasing than
the Toronto Star.
Because mass-market newspapers reflect dominant interests, it is believed
that the Toronto Star will overestimate the risks associated with SARS by
producing more stories about the risks of SARS and especially stories about
the increasing threat. The World Journal, on the other hand, will
underestimate the risks of SARS because of the Chinese Canadian community's
need to reduce anxiety and conflict.
We define the risk frame as one that describes SARS as either becoming more
or less of a health threat. This threat can be real or perceived. For
example, increasing of decreasing numbers of infected or hospitalized
patients or increasing deaths constitute real changes. Perceived risks are
those such as increased quarantines or possible infections. To constitute a
coding decision as risk, there had to be a change stated. For instance,
simply listing the number of deaths was not coded as risk unless given in
context.
METHODS
We analyzed all stories about SARS carried by the Toronto Star and the
Toronto edition of the World Journal between March 16, 2003, and July 31,
2003. March 16 is the date when the first case of SARS was reported in the
Toronto Star. We chose July 31 because this is one full month after the
World Health Organization lifted its travel warning against Beijing.
Several factors were considered in the selection of newspapers, including
circulation, national and regional influence, and language. The Toronto
Star and the World Journal were chosen because the Toronto Star has the
largest circulation in the area, and the World Journal is the largest
ethnic Chinese newspaper in the area. Founded in 1892, the Toronto Star is
Canada's largest metropolitan daily newspaper. With an average daily
circulation of well over half a million, the Toronto Star has a wide base
of readers.
The World Journal is a member of Taiwan's United Daily News Group and is
the largest Chinese newspaper in North America. The World Journal was
founded in 1976 and is one of the few daily Chinese-language newspapers
published nationwide in North America with independent publishing
operations in New York, Los Angeles, San Francisco, Chicago, Houston,
Vancouver, Toronto, and business offices throughout North America. The
newspaper publishes roughly 60-100 pages per day. It claims a circulation
of about 320,000 in North America, 115,500 in Canada, and 22,000 in
Toronto. Coverage of the World Journal highlights important international,
U.S., and Canadian news with emphasis on news related to China, Taiwan,
Hong Kong, Southeast Asia, and local ethnic Chinese communities.
Using "SARS" as the search term in "headlines, lead paragraphs and terms,"
all stories published in the Toronto Star during this period were retrieved
from LexisNexis. All stories run in the World Journal were photocopied from
the Toronto Reference Library in October 2003. Editorials were included in
the study, but reader's letters, news briefs, and stories with less than
200 words were excluded. A total of 735 stories from the Toronto Star and
1,747 stories from the World Journal were coded with the unit of analysis
as the individual news story.
Framing measures
We tested for the presence of five frames identified by Semetko and
Valkenburg (2000), which include conflict, responsibility, economic
consequences, morality, and human interest. In our pretest, we had little
variability in the morality frame and excluded it from our study. Instead,
we added a risk frame because many of the stories in our pretest did not
adhere to Semetko & Valkenburg's (2000) five frames but instead were framed
around the dimension of risk.
Further, we believe stories often have more than one frame, so instead of
using a binary yes-no coding strategy, we measured to what degree the frame
occurred in the story (no use, some use, and much use) "Much use" of the
frame was defined as instances when the frame is the primary organizing
structure of the story. "Some use" of the frame refers to stories in which
the frame was secondary in nature. In our coding scheme, each story can
only have one frame coded as "much use." However, it is possible for
multiple frames to be coded as "some use" within a single news story.
In addition, we investigated the context in which these frames were used.
For example, in terms of the responsibility frame, we analyzed whether it
is Chinese or non-Chinese actors or institutions who are cited as being at
fault. When the economic consequences frame was used, we coded whether the
use was on the macro or micro level. When the risk frame was used, we
examined whether the risk was decreasing or increasing. If the context in
which a certain frame was used changed in the same story, we only coded the
one that first appeared.
Intercoder reliability
As a cross-cultural study, the different language use and different styles
of the two newspapers made intercoder reliability more complicated. Peter
and Lauf (2002) suggested that intercoder reliability assessed for coding
in one particular country is not comparable with that assessed in another
country because reliability tests are not conducted with identical units of
content. In addition, the difference in native language between the two
researchers also posed difficulties in the reliability test (English and
Chinese). Peter and Lauf (2002) concluded that in cross-national studies,
reliabilities are comparable across all country groups when "first, the
coder trainers agree in their coding with one another; second, the coders
within a country group agree with one another; and, third, the coders agree
with the coding of their trainer" (p. 827).
Therefore, in this study, we followed two steps in achieving reliability.
First, we reached agreement in coding stories written in English by coding
all stories published by The Toronto Sun. The Toronto Sun's coverage is
similar to the Toronto Star except it relied more heavily on wire stories
and had fewer stories about SARS. The reason we did not sample from the
Toronto Star is because we planned on examining every article from the
Toronto Star, so we wanted to avoid coding the same articles in our pretest
that would be later included in our study. Intercoder reliability was
computed with Holsti's coefficient of reliability with reliabilities of at
least 80% for the five frames and their specific dimensions (i.e. for the
economic frame, we also coded whether it was on the macro or microeconomic
level).
In the second step, one researcher, whose native language is Chinese,
trained another Chinese coder. They coded 150 Chinese-language stories
randomly selected from the World Journal to achieve reliability. Intercoder
reliability for the key variables was at least 85%, using Holsti's
coefficient of reliability. After reliability was established, one coder,
whose native language is English, coded all stories from the Toronto Star,
and the other researcher, whose native language is Chinese, coded all
stories from the World Journal.
RESULTS
Descriptive data
Our sample consisted of 2,466 stories: 1,731 from the World Journal and 735
from the Toronto Star (see Table 1). These stories were coded on a
three-point scale with a score of 1 representing no use, 2 representing
some use, and 3 representing no use. Of the 2,466 stories, the most used
frames combined from both papers are the risk and responsibility frames
(mean = 1.82 and 1.75, respectively). There was also substantial use of the
economic consequences frame (mean = 1.57). The least used frame was the
human-interest frame (mean = 1.23) and the conflict frame (mean = 1.22).
The means for each newspaper yield similar results (see table 2). Both
newspapers still use the responsibility and risk frames the most
prominently. The World Journal used the risk frame (mean = 1.90) followed
by the responsibility frame (mean = 1.72) most, while the Toronto Star used
the responsibility frame (mean = 1.81) the most, followed by the risk frame
(1.64). Both newspapers used the economic consequences frame often with a
mean of 1.54 for the World Journal and 1.62 for the Toronto Star. The next
most prominent frames for the Toronto Star is the human-interest frame
(mean = 1.44) and the conflict frame (1.39). For the World Journal, the
human-interest and conflict frames both have means of 1.15. Overall, the
two papers featured the same frames in almost the same rank order.
Hypothesis tests
To test for the five main hypotheses, we ran independent t-tests to
determine if there was a significant difference between the frames used in
each newspaper (see Table 2). In our first hypothesis, we predicted the
Toronto Star would use the conflict frame more than the World Journal.
Using an independent t test, there was a significant score of -8.59 (df =
1080, p<.001) with a mean difference of .24 (mean=1.15, sd=.51 for the
World Journal and mean=1.39, sd=.69 for the Toronto Star). Therefore, there
was a significant difference in the direction predicted, and the hypothesis
is supported.
Further, we predicted during times of crisis that the Toronto Star's
greater use of the conflict frame would be more pronounced. A Chi-square
test produced a non-significant score of 0.41 (df = 1). Also, we found that
the direction of our hypothesis was not supported. A cross tabulation of
this frame in crisis and non-crisis periods revealed that during non-crisis
periods, the Toronto Star used the frame 3.1% more often than the World
Journal (74.2% and 71.1%, respectively). We had predicted that this
difference would increase in size, but, instead, the difference moved in
favor of the World Journal. That is, the World Journal used the conflict
frame during times of crisis 4.1% more often than the Toronto Star (28.9%
and 25.8%, respectively). Therefore, our hypothesis both moves in the
opposite direction of our predictions and also is not statistically
significant, providing no support for this hypothesis.
Next, we hypothesized that the Toronto Star would use the responsibility
frame more than the World Journal. According to an independent t-test, we
found a significant difference between the two newspapers. As predicted,
the Toronto Star (M = 1.81) used this frame slightly more than the World
Journal (M = 1.72).
Moreover, we predicted that the Toronto Star would blame ethnic Chinese
institutions and people more than the World Journal, while the World
Journal would assign the blame more to non-Chinese institutions and people
than the Toronto Star. Again, chi-square was run to test for these
differences (see Table 4). Though we did get a significant relationship, _2
of 287.82 (df = 1, p<.05), the findings were in the opposite direction of
our hypothesis. The Toronto Star blamed ethnic Chinese institutions and
people in only 22.4% of its responsibility frame usage and non-Chinese
institutions and people in the 77.5%. The World Journal, on the other hand,
blamed ethnic Chinese institutions and people 73.4% and non-Chinese
institutions and people only 26.6% in its responsibility frames.
Third, we predicted that the World Journal would use the economic
consequences frame more than the Toronto Star. Again, we found a
significant difference, which shows that the two papers featured this frame
almost equally as prominently (see Table 2). At any rate, the difference is
not in the predicted direction. Therefore, this hypothesis is not supported.
Along with this overall prediction, we expected that the World Journal
would report on the microeconomic level more prominently than the Toronto
Star when using the economic consequences frame, while the Toronto Star
will report on the macroeconomic level more prominently than the World
Journal (see Table 5). A Chi-square test (8.94, df = 1, p<.05) reveals that
there is a significant difference in the use of the different economic
levels. More importantly, these differences are in the predicted direction.
The World Journal reported more often on the microeconomic level (44.9%)
than the Toronto Star (34.5%), while, the Toronto Star reported on the
macroeconomic level (65.5%) more than the World Journal (55.1%). Thus, our
hypotheses about the reportage of economic level are supported.
Next, we hypothesized that the World Journal would use the human-interest
frame more prominently than the Toronto Star. Contrary to our predictions,
the Toronto Star used this frame significantly more (t = -9.65, df = 1001,
p<.001) than the World Journal with a mean difference of .29 (1.15 for the
World Journal and 1.44 for the Toronto Star), which is the largest mean
difference of any of our frames. Therefore, this hypothesis is not supported.
Finally, we predicted that the Toronto Star would use the risk frame more
than the World Journal (see Table 2). Contrary to expectations, the World
Journal used the risk frame significantly more than the Toronto Star (t =
6.56, df = 1530, p<.001). In fact, the mean difference was .26 (1.90 for
the World Journal and 1.64 for the Toronto Star). This is the second
largest mean difference of the frames studied. Therefore, this hypothesis
is not supported.
In addition, we hypothesized that the Toronto Star would report on
increasing risk more often than the World Journal and that the World
Journal would report on decreasing risk more often than the Toronto Star. A
Chi-square test (25.96, df = 1, p<.05) reveals a significant difference,
but this difference is not in the hypothesized direction. In fact, the
World Journal reported on increasing risk more often than the Toronto Star
(78.5% and 63.4% respectively). Also inconsistent with our hypothesis is
the Toronto Star's more frequent reportage of the decreasing risk of SARS
than the World Journal (36.6% and 21.5%, respectively). Therefore, none of
the hypotheses in this set are supported.
Overall, four of our 12 hypotheses are supported. Our predictions that the
Toronto Star would use the conflict and responsibility frames more
prominently are supported. However, our predictions on the direction of
blame are not supported. The hypothesis that the World Journal would use
the economic consequences frame more was not supported. However, our
predictions on the economic levels reported for each newspaper are
supported. The hypothesis that the World Journal would use the
human-interest frame more often is not supported. Further, our prediction
that the Toronto Star would use the risk frame more was not supported, and
neither was our prediction that the Toronto Star would report on increasing
risks more than the World Journal.
DISCUSSION
Toronto was the only area outside of Asia to have a serious outbreak of
SARS. The cost of this outbreak on Toronto was enormous. SARS claimed 41
lives in the city (Summary, 2003), caused many more to become ill, and led
to thousands being quarantined. Further, SARS took a heavy toll on the
economic health of the city. Toronto's tourism industry was paralyzed for
several weeks and its health industry was stretched thin because of SARS,
costing the city hundreds of millions in dollars spent and potentially lost.
The press widely reported that the source of the infection was an elderly
Chinese Canadian resident of Toronto, who had recently visited Hong Kong
and become sick. She then spread the virus to her family and to health care
workers, which led to the initial outbreak in the city. Literature would
suggest that with a minority group source, the ethnic Chinese population
might be targeted and blamed as the cause of the disease. In fact, there
were some hate crimes, vandalism of Chinese Canadian-owned property, and
editorials denouncing the racially-infused discourse around SARS.
Given the literature on the Canadian press's coverage of Chinese Canadians
in the past and the research that indicates mass-market newspapers'
reflection of the dominant society, it was expected that The Toronto Star
would use more frames that emphasize blame of Chinese Canadians and the
Chinese government and the increasing risk of SARS. On the other hand, it
was expected that The World Journal, an ethnic Chinese newspaper, would
frame SARS less often as being because of Chinese Canadians or the Chinese
government and that it would emphasize its decreasing risk, especially
given the impact SARS was having on the personal and economic lives of its
Chinese Canadian readership. In fact, Riggins (1992) writes "exercising
considerable self-censorship, they [ethnic media] concentrate on topics
flattering to the minority group."
In general, however, the results were opposite of the hypothesized
directions. One of the hypotheses supported was that the Toronto Star would
use more conflict frames than the World Journal. While this may imply the
conflicts were with ethnic Chinese institutions or organizations, the
Toronto Star tended to avoid these associations. In fact, much of the
conflict was framed around internal conflicts between institutions and
people within Canada. There were far fewer conflict frames that included
individuals of Chinese descent or Chinese institutions. A reporter from the
Toronto Star said, "I think we went out of our way – I hope we did – to
ensure that we didn't contribute, in any way, to the stigmatizing of the
Chinese people." It appears, in fact, that the newspaper accomplished this
stated intention. This may be why our hypothesis that the Toronto Star
would report on conflict more often during times of crisis was not supported.
Further, our second hypothesis that the Toronto Star would use the
responsibility frame more frequently was supported, but our prediction that
the Toronto Star would place the blame on ethnic Chinese institutions and
people, while the World Journal would blame non-Chinese institutions and
people was not supported. There are several possible explanations for this.
First, the reporters and editorial staff at the Toronto Star may have been
conscientious not to connect SARS to the Chinese Canadians living in
Toronto, while being more critical of its local and federal leadership.
Second, our operationalization of responsibility, which only included those
times when there were clear attributions of blame or when an individual or
organization was shown to have led to an infection or an increased spread
of SARS, may have underestimated the implicit links. The many references in
The Toronto Star to the origins of SARS in Guangdong Province were not
included because it did not fit this criterion. However, this implicit
association may have led readers to explicitly blame those organizations
and individuals with connections to China.
Further, because the World Journal's readership is primarily ethnic
Chinese, it includes many more stories about SARS in China than the Toronto
Star. This alone may have increased the use of the responsibility frame as
there was more news of the virus spreading to different Chinese provinces
and to different nations. Finally, because the World Journal is owned by a
Taiwanese-based news organization, it is possible it was especially
critical of mainland China's culpability in allowing the export of SARS to
other nations.
For the hypothesis on the economic consequences frame, there were mixed
results. We had hypothesized the World Journal would use this frame more
often because more of its readership was economically affected more
severely than the general population. However, it was the Toronto Star that
used the economic consequences frame more prominently. The overall economic
effects of SARS were so widespread that the Toronto Star was compelled to
use the economic consequences frame often. Despite this incorrect
hypothesis, our sub-hypotheses were supported. The Toronto Star did, in
fact, use the macro level economic consequences frame more often, and the
World Journal used the micro level economic consequences frame more often
because of the many Chinese Canadian small business owners affected by SARS.
Interestingly, however, the World Journal did not use more of the
human-interest frame. Because of the economic repercussions on the Toronto
Chinatown and because of the sometimes caustic anti-Chinese discrimination,
we had expected the World Journal to incorporate more human-interest
stories. However, it was the Toronto Star that had more human-interest
stories. Many of these stories had to do with individual heroics during the
crisis, remembering those who had died, and the experience of the quarantined.
A reason this may be is that because so many were quarantined or otherwise
affected by SARS, the residents of Toronto may have rallied together as a
single, cohesive ingroup, leading to more reportage of human-interest
angles. When discussing this issue withsome reporters from the Toronto
Star, they emphasized their desire to write about human-interest stories.
One reporter said, "I would try to do personal stories, as much as you
could, because you didn't often know who the people were."
Another reason may have to do with routines. It is possible the routines of
the different newspapers affected the ability to use the human-interest
frame. Stories in the World Journal were significantly shorter, and the
reporters do not have as much time to investigate stories. Reporters at the
World Journal are expected to write three stories and 2000 words per day,
according to an editor from the newspaper. Because of these time
constraints, reporters for the World Journal may not have had the
opportunity to uncover and develop human-interest stories.
Finally, contrary to expectations, the World Journal used the risk frame
more frequently than the Toronto Star. This is most likely because the
World Journal included many stories from China, Hong Kong, Taiwan, and
Southeast Asia, where SARS had its most devastating impact. Further, the
World Journal's Taiwanese ownership group may have written about the
increasing risk as a tacit editorial attack on Chinese culpability. Also,
the Toronto Star may have been more careful in its coverage of the risks as
the entire city's population may have coalesced into a larger ingroup that
is fighting against SARS and organizations that threaten the livelihood of
Torontonians, in general. In fact, several reporters indicated that they
were careful in their coverage of the health risks as they felt it was
their responsibility not to create undue alarm and because some readers
complained that the Toronto Star was "hyping" the disease.
Overall, several major hypotheses are supported, but many of our findings
are opposite of what we expected. These are interesting findings as there
has been very little research comparing ethnic and mass-market media. The
results in this paper suggest the ethnic media may play a stronger
assimilationist role than previously believed by criticizing their ingroup
and not creating many sympathetic human-interest figures. On the other
hand, the Toronto Star's coverage may have indeed been fair and may have
even taken on an advocacy role to prevent increasing ingroup-outgroup
tension. Despite the increasing anti-Chinese sentiment that appeared after
SARS, the Toronto Star appeared not to have played a direct role in
fomenting these feelings. In fact, it may have played a role in mitigating
anti-Chinese sentiment by focusing on internal conflicts and blame.
Limitations and future directions
The World Journal in Toronto is an independent branch but fills a
substantial part of its news hole from branches around the globe. This may
have affected the analysis by presenting a disproportionate number of
frames that assign blame to Chinese institutions and individuals and an
overrepresentation of the increasing risk of SARS. It would be worthwhile
to specifically look at news coverage that is centered in Toronto. Results
may differ based on this coverage alone.
Further, we used different forms of the two media. The Toronto Star was
retrieved from LexisNexis, so it loses contextuality, while the World
Journal was photocopied from library collections. Looking at both media in
the original form would have been helpful as it allows an observation of
the placement and images that are associated with the stories. This would
strengthen our sense of the context in which the frames were used.
Finally, our paper studied two specific newspapers in a specific location
during a specific health crisis. In other words, caution should be taken in
generalizing these results. We can only observe these two newspapers'
framing of SARS. Future research should also look at news from a random
sample of mass market and ethnic newspapers when there is not a major
crisis like SARS. This may change the use of frames dramatically for both
the Toronto Star and the World Journal. It would also be beneficial to
observe ethnic newspapers that serve different populations. Cultural
differences may affect the use of frames by ethnic newspapers. For example,
the African American ethnic press's use of frames may be different than the
use of the Chinese or Spanish ethnic press.
Works Cited
Carey, E. (2003, April 25). Hospitals facing huge SARS bill. Toronto Star,
p. B03.
Chinatown. Toronto Convention and Visitors Association: The Official
Tourism Web Site of the City of Toronto. Retrieved June 24, 2003 from
http://www.torontotourism.com/ about/neighborhoods_chinatown.asp
Dunn, K.M. & Mahtani, M., (2001) Media representations of ethnic
minorities. Progress in Planning, 55(3), 163-72.
Entman, R.M. (Summer 1992). Blacks in the news: Television, modern racism
and cultural change. Journalism Quarterly, 69(2), 341-361.
Entman, R. M. (1993). Framing: Toward clarification of a fractured
paradigm. Journal of Communication, 43(4), 51-58.
Ferguson, R. (2003, April 26). SARS tab may top $2 billion: Bank. Toronto
Star, p. D01.
Fleras, A. & Kunz J.L. (2001). Media and Minorities: Representing Diversity
in a Multicultural Canada. Toronto: Thompson Educational Publishing.
Gandy, O.H., Kopp, K., Hands, T., Frazer, K., & Phillips, D. (Spring 1997).
Race and risk: Factors affecting the framing of stories about inequality,
discrimination, and just plain bad luck. Public Opinion Quarterly, 61(1),
158-182.
Greenberg, J. 2000. Opinion discourse and Canadian newspapers: The case of
the Chinese "Boat People". Canadian Journal of Communication, 25 (4), 517-37.
Hamamoto, D.Y. (1994). Monitored Peril. Minneapolis, MN: University of
Minnesota Press.
Iyengar, S. (1991). Is Anyone Responsible? Chicago: University of Chicago
Press.
Jeffres, L.W. (2000). Ethnicity and ethnic media use: A panel study.
Communication Research, 27, 496-535.
Keung, N. (2003, March 29). Business down at Chinese malls. Toronto Star.
p. A27.
Lam, L. (1980). The role of ethnic media for immigrants: A case study of
Chinese immigrants and their media in Toronto. Canadian Ethnic Studies,
12(1), 75-92.
Lee, W. & Tse, D. (1994). Changing media consumption in a new home:
Acculturation patterns among Hong Kong immigrants to Canada. Journal of
Advertising, 23, 57-69.
Lewellen, T.C. (1992). Political Anthropology: An Introduction. Westport,
CT: Bergin & Garvey.
Lupton, D. (1993). AIDS risk and heterosexuality in the Australian press.
Discourse & Society, 4(3), 307-328.
Mahtani, Minelle. 2001. Representing Minorities: Canadian Media and
Minority Identities. Canadian Ethnic Studies, 33(3), 99-133.
Meinbardis, R. (2003, June 19). Canada's Manley says economy hurt by SARS,
mad cow. Reuters. Retrieved June 20, 2003 from http://reuters.com/
printerFriendlyPopup.jhtml?type=worldNews&storyID= 2958812
Orbe, M.P., Seymour, R., & Kang, M. (1998) Ethnic humor and
ingroup/outgroup positioning: Explicating viewer perceptions of
All-American Girl. In Kamalipour, Y.R. & Carilli, T. (Eds.), Cultural
Diversity and the U.S. Media, 125-136. State University of New York Press.
Pan, Z. & Kosicki (1996). Assessing news media influences on the formation
of Whites' racial policy preferences. Communication Research, 23(2), 147-178.
Peter, J. & Lauf, E. (2002). Reliability in cross-national content
analysis. Journalism and Mass Communication Quarterly, 79(4), 815-832.
Rider, D. (2003, April 4). Fear of virus fuels racism: Ontario must do more
to stop return to days of 'Yellow Peril,' Asian leaders say. Ottawa
Citizen. p. A10.
Riggins, S.H. (1992). The promise and limits of ethnic minority media. In
S.H. Riggins (Ed.), Ethnic minority media: An international perspective.
Thousand Oaks, CA: Sage Publications.
Romer, D. Jamieson, K.H. & de Coteau, N.J. (1998). The treatment of persons
of color in local television news: Ethnic blame discourse or realistic
group conflict? Communication Research. 25 (3), 286-305.
Scheufele, D. (1999). Framing as a theory of media effects. Journal of
Communication, 49, 103-122.
Shah, H. & Thornton, M.C. (1994). Racial ideology in U.S. mainstream news
magazine coverage of Black-Latino interaction, 1980-1992. Critical studies
in Mass Communication, 11, 141-161.
Shim, D. (1998). From yellow peril through model minority to renewed yellow
peril: Asians in popular media: Constructing (mis)representations. Journal
of Communication Inquiry, 22, 385-410.
Semetko, H.A. & Valkenburg, P.M. (Spring 2000). Framing European politics:
A content analysis of press and television news. Journal of Communication,
93-109.
Sorenson, C. (2003, April 4). Chinese Canadians feeling backlash. Toronto
Star. p. A06.
Summary table of SARS cases by country, 1 November 2002-7 August 2003.
(2003). Retrieved March 28, 2004, from
http://www.who.int/csr/sars/country/en/country2003_08_15.pdf
Travel alert widened as SARS toll grows. (2003, April 27). The Washington
Post. p. A03.
Turner, J.C., Brown, R.J., Tajfel, H. (1979). Social comparison and group
interest in ingroup favouritism. European Journal of Social Psychology,
9(2), 187-204.
Ungerleider, C.S. (1991). Media, minorities, and misconceptions: The
portrayal of minorities in Canadian news media. Canadian Ethnic Studies,
23(3), 158-164.
Update 42 – Travel advice for Toronto, effective tomorrow. (2003, April
29). Retrieved June 21, 2003 from http://www.who.int/csr/sars/archive/
2003_04_29/en
Update 84: Can SARS be eradicated or eliminated? (2003, June 19). Retrieved
June 20, 2003 from http://www.who.int/csr/don/2003_06_19/en
Vergeer, M., Lubbers, M. & Scheepers, P. (2000). Exposure to newspapers and
attitudes toward ethnic minorities: A longitudinal analysis. The Howard
Journal of Communication, 11(2), 127-143.
Viswanath, K. & Arora, P. (2000). Ethnic media in the United States: An
essay on their role in integration, assimilation, and social control. Mass
Communication & Society, 3(1), 39-56.
Tajfel, H. & Turner, J.C. (1979). An integrative theory of social conflict.
In W.G. Austin & Worchel (eds.), pp. 33-47. The Social Psychology of
Intergroup Relations. Monterey, CA: Brooks/Cole Publishing Company.
Wilkins, L. & Patterson, P. (1987). Risk analysis and the construction of
news. Journal of Communication, 37(3), 80-92.
Wilson, C.C. & Gutierrez, F. (1995) Race, multiculturalism, and the media:
From Mass to class communication (2nd ed.). Thousand Oaks, CA: Sage
Publications.
Table 1. Means and standard deviations for conflict, responsibility,
economic consequences, human interest, and risk frames for World Journal
and Toronto Star combined
Variables Mean
SD N
Conflict 1.22 0.58
2476
Responsibility 1.75 0.81 2476
Economic Consequences 1.23 0.61 2476
Human Interest 1.57 0.84 2476
Risk 1.82 0.93 2476
Note. Frames were coded on a three-point scale: 1 = no use, 2 = some use, 3
= much use.
Table 2. Independent t tests for use of five frames by World Journal and
Toronto Star
Variables World Journal Toronto Star t
value df significance
(Mean & SD) (Mean & SD)
(N=1741) (N=735)
Conflict 1.15 1.39 -8.59 1080
p < .001
(.51) (.69)
Responsibility 1.72 1.81 -2.56 1328
p < .01
(.80) (.84)
Economic 1.54 1.62 -2.34 1583
p < .05
Consequences (.87) (.75)
Human
interest 1.15 1.44 -.9.65 1001
p < .001
(.50) (.78)
Risk 1.90 1.64 6.56 1530 p
< .001
(.95) (.86)
Note. Frames were coded on a three-point scale: 1 = no use, 2 = some use, 3
= much use.
Table 3. Cross-tabulation of conflict frame in crisis and non-crisis period
by newspaper type
Newspapers World Journal Toronto Star
Crisis period 28.9% 25.8%
Non-Crisis period 71.1% 74.2%
100.0% 100.0%
(N = 142) (N = 198)
Note. X2 = 0.407, df = 1, ns
Cramer's V = 0.035
Table 4. Cross-tabulation of party being at fault in responsibility frame
by newspaper type
Newspapers World Journal Toronto Star
Chinese 73.4% 22.4%
Non-Chinese 26.6% 77.5%
100.0% 100.0%
(N = 867) (N = 396)
Note: X2 = 287.819, df = 1, p < .05
Cramer's V = 0.477
Table 5. Cross-tabulation of economic level in economic consequences frame
by newspaper type
Newspapers World Journal Toronto Star
Micro 44.9% 34.5%
Macro 55.1% 65.5%
100.0% 100.0%
(N = 499) (N = 336)
Note: X2 = 8.938, df = 1, p < .05
Cramer's V = 0.103
Table 6. Cross-tabulation of direction in risk frame by newspaper type
Newspapers World Journal Toronto Star
Increase 78.5% 63.4%
Decrease 21.5% 36.6%
(N = 848) (N = 287)
Note: X2 = 25.958, df = 1, p < .05
Cramer's V = 0.151
Appendix A – Coding Guide
Variable Definition
Code Number on article (i.e. 1,2,…25, etc.)
Type Code the type of newspaper
1 = The World Journal
2 = The Toronto Star
Date Code the date and month of the article as MMDD.
Story origin Code whether the newspaper produces the article or whether it
is a wire story.
1 = Toronto-based story written by staff at the paper and
national/international story written by staff at the paper; editorials are
also included in this category.
2 = Western wire story
3 = Chinese wire story
Conflict Code to what degree the conflict frame is used in the story.
0 = No use of the frame
1 = Some use of the frame
2 = Much use of the frame
Responsibility Code to what degree the responsibility frame is used in the
story.
0 = No use of the frame
1 = Some use of the frame
2 = Much use of the frame
Fault Code whether it is Chinese or non-Chinese actors and institutions who
are cited as being at fault.
1 = Chinese
2 = Other
99 = not
applicable (when responsibility frame = 0)
Economic consequences Code to what degree the economic consequences frame
is used in the story.
0 = No use of the frame
1 = Some use of the frame
2 = Much use of the frame
Level Code whether the use of the economic consequences frame is on the
macro or micro level. The macro level is considered to be above the
city-wide level; whereas the micro level is considered to be anything below
and at the city-wide level, including groups, neighborhoods, individuals,
local industries, etc.
1 = Micro
2 = Macro
99 = not
applicable (when economic consequences frame = 0)
Human-interest Code to what degree the human-interest frame is used in the
story.
0 = No use of the frame
1 = Some use of the frame
2 = Much use of the frame
Risk Code to what
degree the risk frame is used in the
Story
0 = No use of the frame
1 = Some use of the frame
2 = Much use of the frame
Direction Code whether the risk is
decreasing or increasing.
1 = increasing
2 = decreasing
99= not
applicable (when risk frame = 0)
|
