What's Wrong With Fat? (33 page)

What’s wrong with fat? As this book has shown, the answer to this basic question depends on how fatness is framed. And this has critical social implications.

METHODOLOGICAL APPENDIX

In conducting the research for this book, I have adapted a tripartite approach to culture, considering: (1) the
production
of culture (in this case, ideas about fatness), (2) cultural
texts
themselves (such as news reports and official documents), and (3) the way in which cultural texts are
received
, or interpreted. 1
Whereas much work in cultural studies assumes that the meaning of texts, such as magazines, newspaper reports, novels, or film, is self-evident, cultural sociologists recognize that different people can interpret the same texts in different ways. 2 Cultural sociologists emphasize that it is important to understand the social processes that produce particular accounts.

To get at the ways in which particular accounts of fatness are
produced
, I interviewed leading obesity researchers, health at every size researchers, fat right activists, and journalists and publicists who produce authoritative or influential accounts of fatness. I also conducted participant observation at fat acceptance conferences and list servers. I informally spoke to and corresponded via e-mail with scores of additional researchers and activists to clarify specific points that I make in this book. As part of studying the
content
of these messages, I examined hundreds of journal articles, books, official documents, and news articles. To capture how different accounts of fatness are
received
, I asked interview respondents how they interpret and are affected by dominant and challenging accounts of fatness. In addition, social psychologist and former UCLA PhD candidate David Frederick and I designed seven controlled experiments with adults and college students to see how reading specific news accounts of fatness affected the way they responded to questions about health risks and weight-based discrimination and stigma. Some of how I went about doing this research is described in the preceding chapters.
Here, I provide some additional information for interested readers.

FORMAL INTERVIEWS

For this project, I formally interviewed 35 researchers, activists, and journalists, some of them several times. At the start of these interviews, I presented respondents with an “information sheet” that explained their rights as “human subjects,” including that they were not obligated to answer my questions and could end the interview at any time. The information sheet also contained information about the purpose of the study, interview procedures, potential benefits of the research to society, that there would be no payment for participation, and that, if they so chose, all information in the interview would be kept confidential. Respondents also had the option of being identified, which was the preference of many, given their status as public figures. I used an interview guide, which I provide here. Table A.1 lists all of these respondents.
It does not include several other researchers and activists with whom I spoke informally or exchanged multiple e-mails, often over several years. Nor does it include the countless discussions that I had with others who have struggled with the issues discussed in this book in various ways.

TABLE A.1
Interview Respondents

NEWS MEDIA SAMPLING

This books draws on analyses of several different news media samples. The first three were drawn from news articles and opinion pieces of more than 300 words published in
The New York Times, Newsweek, Le Monde,
or
L’Express
that had the words
obese
/
obesity
,
overweight
/
obesity
,
anorexia
/
anorexic
, or
bulimic
/
bulimia
in the heading or lead paragraphs. This generated four separate samples, three of which were systematically analyzed in chapter 3 and are shown in table A.2. The French articles on eating disorders were not systematically analyzed. Instead, my students and I focused on two comparisons: (1) U.S. versus French reporting on overweight/obesity; and (2) U.S. reporting on obesity/overweight versus U.S. reporting on eating disorders.

While limited to four news outlets over 10 years, these samples allow for the examination of how the elite U.S. and French news media discuss fat bodies over time. The year 1995 is the first year that a sizable number of articles on obesity was published in France. These publications have the methodological advantage of being available on LexisNexis for the entire 1995–2005 time period and of maximizing cross-national comparability.
The New York Times
and
Le Monde
are comparable in their readership and reputation as leading national newspapers of record. Similarly,
Newsweek
and
L’Express
are comparable leading mainstream newsmagazines. 3

There is a consistently moderate level of press attention to eating disorders over this time period. 4 Following several high-profile pronouncements from the Centers for Disease Control, the World Health Organization, and the surgeon general about the “obesity epidemic” at the end of 2001, there is considerably more news reporting in the United States on overweight/obesity
beginning in 2002. Because of this, our American sample is weighted toward the later time period. To assess whether reporting during the high-volume years (2002–2005) was significantly different than the lower-volume years (1995–2001), we ran additional analyses for all of the variables discussed by time period, which are discussed in chapter 3.

TABLE A.2:
National and Issue Comparative News Sample

My students and I sampled news articles and opinion pieces published in these publications between 1995 and 2005 that had the words
obese
/
obesity
/ o
verweight
,
anorexia
/
anorexic
(
s
)/
bulimia
/
bulimic
(
s
), or their French equivalents in the heading or leading paragraphs. We then winnowed down this list using three criteria. First, to manage the extremely high volume of articles on overweight/obesity published in
The
New York Times,
we eliminated the first two of every three articles in the chronological
New York Times
list of articles, reducing this sample by two-thirds. To measure the extent to which this sampling strategy may bias our results, we analyzed the newspaper and newsmagazine articles separately, finding that the general patterns hold up for both. 5

Second, we eliminated articles that were less than 300 words, since it is difficult to develop the themes of interest in such a short article, which reduced the sample further by more than one half. Finally, we eliminated the few articles from the full sample that were completely off topic. This yielded a final sample of 262 U.S. articles and 108 French articles on overweight/obesity, including 174 from
The New York Times
, 88 from
Newsweek
, 86 from
Le Monde
, and 22 from
L’Express
. This sampling strategy produced 70 U.S. articles on anorexia/bulimia, including 64 from
The New York Times
and six from
Newsweek
. The 17 French articles on anorexia/bulimia were not analyzed.

Two theme issues on obesity published in the
Journal of the American Medical Association
(
JAMA
) in 1999 and in 2003 provided the basis for a second pair of news media samples, discussed in chapter 4. This design allowed me to examine which sorts of articles receive the most news media coverage.
The ten major research reports and editorials in each theme issue were also coded using the same coding protocol, allowing for systematic comparisons between the original scientific studies and news reports.

With the help of student research assistants, and after experimenting with different search terms, I searched the LexisNexis database for all news articles published in the three months after the publication of each
JAMA
issue that fulfilled the following search criteria: they contained the word
obesity
or the word
weight
and the three consecutive words
American Medical Association
anywhere in the full text. 6 We limited our search to the following LexisNexis categories: General News, World News, News Wires, Business News, Legal News, University News, and Medical News. We excluded articles with fewer than 500 words and peer-reviewed journal articles, generating a final news sample of 128 news articles, including 69 on the 1999 issue and 59 on the 2003 issue. The sample was heavily weighted toward the General News category (N = 66), followed by the News Wires category, which includes several smaller publications (N = 24), and Business News (N = 21). Relatively few articles fell into the World News (N = 8), Medical News (N = 5), and University News (N = 4) categories, biasing the sample toward U.S. news. At 500 words or more, they are a bit longer than the articles in the other three samples. They also include a wider range of publications.

Finally, I analyzed two media samples representing news coverage of two specific scientific studies. The first, “Actual Causes of Death in the United States, 2000,” was published in 2004 and is described in depth in chapter 4. 7 I refer to this as the Eating-to-Death study. The second study, “Excess Deaths Associated with Underweight, Overweight, and Obesity,” which I refer to as the Fat-OK study, was published in 2005. 8 Searching the Business and Finance, Industry News, Major Papers, Magazine and Journals, Medical and Health News, All Available News Wire reports, and University Wire sources of LexisNexis, I identified all articles published within three months of the research study that included the term
obesity
and either the first author’s last name or both
CDC
and
American Medical Association
anywhere in the full text.
Only news reports that actually discussed the article in question and were 500 words or more were retained. This produced a sample of 35 news reports on the Eating-to-Death study and 61 reports on the Fat-OK study. This pair of samples allowed me to examine how the news media differently represent competing claims about obesity-related health risk, published by equally reputable scientists in equally reputable journals. Table A.3 gives the number of articles in each of the science reporting samples.

TABLE A.3:
Science Reporting Samples