What's Wrong With Fat? (34 page)

NEWS MEDIA CODING

All of the news articles were coded for about 200 different variables or themes, of which only a fraction are discussed in this book. I used a slightly different coding schema for the national and issue comparative news samples, on one hand, compared to the science reporting news samples, on the other. The full coding protocol for all of the news media samples can be found at: Here, I review only those codes that are discussed in the book.

In initial “practice” coding, three researchers coded the same articles and discussed differences as a way of arriving at shared agreement. Two coders coded 10 percent of the articles to test for inter-coder reliability, which was very high. The coefficient of reliability (the ratio of coding agreements to the total number of coding decisions) was more than 0.95. 9 In almost all cases coding required noting simply whether or not a particular aspect was mentioned at all. 10 Using the program Microsoft Access, we entered a “1” when the aspect in question was mentioned by the journalist or a news source or a “0” if it was not mentioned. We later imported these numbers into statistical software for analysis. For instance, if the article suggested that obesity was caused by biological factors, even if it also discussed other causal factors, that article would be coded as “1” for biology. If it did not discuss biological factors, it would be coded as “0” for biology. Each article could be coded as “1” on many different variables. In other words, codes are independent of one another. Coders were not asked to determine which themes dominated the article, only if they were present at all.

Doing this kind of systematic coding is tedious and time consuming. Yet, it allows one to systematically quantify broad patterns that are likely to be missed in a more informal reading. By reading a lot of scientific studies and news articles, one can get a sense of general patterns, but this kind of reading is likely to be influenced by selective attention to details that confirm our preexisting expectations. The systematic coding that my research assistants and I did kept us honest, in that it forced us to ask the same questions of each article and to use the same criteria in answering those questions.

Based on the reading I had done and interviews I had conducted, I came up with several codes. For instance, we coded each article for whether or not it suggested that obesity/overweight was a public crisis, represented an epidemic, or used war metaphors (e.g., “battle of the bulge” or “time bomb”). We also coded each article for whether it blurred the lines between different weight categories. A common example of this was an article that discussed people with BMI over 40 as representative of the larger problem of “overweight,” when, in fact, only 2 percent of the U.S. population has a BMI over 40.
By using extreme examples in this way, these articles give an exaggerated impression of population weight. We coded each article for whether it discussed the medical risks associated with being in the underweight, overweight, obese, or morbid obese categories and for whether it linked overweight and/or obesity to specific diseases, including type 2 diabetes, heart disease, stroke, and cancer, or risk factors, including hypertension or high levels of bad cholesterol. We coded for whether the article discussed one of three debates or controversies, including the extent to which there are health risks associated with obesity, what constitutes appropriate cutoff marks for obesity, or whether one can be “fat and fit.”

To evaluate how scientific and news reports assign blame, we coded articles for whether they emphasized personal responsibility, sociocultural factors, or biological factors as causing obesity. 11 For instance, the following article blames an individual for his weight gain, writing “he could look back on decades of binge eating and failed diets,” and quoting him as saying, “I was killing myself.” 12 The following would be taken as evidence of blaming socio-cultural factors: “In many low-income minority neighborhoods, fried carryout is a cinch to find, but affordable fresh produce and nutritious food are not.” 13
Cultural factors, including mainstream cultural emphasis on thinness, ethnic culinary practices, or cultural attitudes toward body size, were treated as a subset of sociocultural factors. This means that any article coded as 1 for cultural factors was also coded as 1 for sociocultural factors.

We also specifically coded for whether fatness was attributed to bad food choices, to sedentary lifestyles, or to the food industry. We coded for whether each article discussed potential solutions, including behavioral changes, policy changes, weight-loss drugs, and weight-loss surgery. During analyses, we computed a composite variable for any medical intervention. We coded each article for whether it specifically mentioned individuals changing “lifestyles,” individuals increasing exercise, individuals eating better, governmental regulation of food advertisements, or removing vending machines from schools or other places.

We coded articles for whether they discussed weight-based discrimination or whether it was possible to be fat and healthy. Finally, to account for how, and to what extent, these issues are associated with different groups, we coded articles for whether they explicitly mentioned specific demographic groups, including men or women; the poor, middle class, or rich; and whites, blacks, Latinos, Asians, and other races. During analyses, we computed composite variables, including “Blacks, Latinos, or the Poor.” Using statistical software, my students and I calculated the frequency that specific variables are mentioned in different articles.

My students and I also coded news coverage of the Eating-to-Death and Fat-OK studies for whether they quoted outside researchers who contested the study’s validity or outside researchers who were supportive of the study’s findings. We also coded these articles for whether they suggested that the study confirmed what was already known and for whether the study’s findings were treated as surprising.

In addition to the quantitative analysis, we used discourse analysis to get at the subtleties of news reports, including the choice of words and ideologies evident in news reports. 14 We created theme sheets that included lengthy quotes that illustrated key themes, such as blame, responsibility, and moral judgment. The quantitative data allow us to test for statistical significance of differences in reporting across these issues, while the qualitative data permit us to dig deeper into the nuances of reporting.

EXPERIMENTS

Chapter 5 summarizes a few key results from seven experiments that I conducted with David Frederick and Kjerstin Gruys. Table A.4 provides the sample sizes and frames tested for each of these seven experiments.

TABLE A.4.
Sample Sizes and Frames Tested in Seven Experiments

ACKNOWLEDGMENTS

1. This claim was based on a 1993 study that estimated the number of excess deaths
associated with “diet and activity patterns,” not overweight and obesity, but was
typically discussed as pertaining to overweight and obesity. J. Michael McGinnis
and William H. Foege, “Actual Causes of Death in the United States,”
Journal of the
American Medical Association
270, no. 18 (1993): 2207–12 ; and “The Obesity
Problem,”
New England Journal of Medicine
338 (1998): 1157.

CHAPTER 1

1. Paul Kramer,
Maggie Goes on a Diet
(Boise, ID: Aloha Publishers, 2011).

2. Karen Kaplan, “‘Maggie Goes on a Diet’ the Sensible Way in Children’s Book,”
Los
Angeles Times
, August 23, 2011.

3. Amazon, customer discussions for
Maggie Goes on a Diet,
http://www.amazon.
com/Look-an-angry-mob/forum/FxPLSMT70BJ700/Tx3U56QEE8NICLK/1/
ref=cm_cd_dp_ef_tft_tp?_encoding=UTF8&asin=0981974554.

4. Wayt Gibbs, “Obesity: An Overblown Epidemic?”
Scientific American
, June 2005,
72–77.

5. Amazon,
Maggie Goes on a Diet
, http://www.amazon.com/Maggie-Goes-Diet-
Paul-Kramer/dp/0981974554/ref=sr_1_1?ie=UTF8&qid=1314655773&sr=8-1.

6. Amazon, customer discussions for
Maggie Goes on a Diet
, http://www.amazon.
com/Maggie-Goes-On-A-Diet/forum/FxPLSMT70BJ700/Tx1V60OXP437GQ9/1/
ref=cm_cd_dp_ef_tft_tp?_encoding=UTF8&asin=0981974554.

7. Erving Goff man,
Frame Analysis: An Essay on the Organization of Experience
(New
York: Harper Colophon, 1974).

8. William James,
Th
e Principles of Psychology
(Cambridge, MA: Harvard University
Press, 1981), 462.

9. David Snow and Robert D. Benford, “Ideology, Frame Resonance and Participant
Mobilization,”
International Social Movement Research
1 (1988): 198 ; see also
Sidney Tarrow, “Mentalities, Political Cultures, and Collective Action Frames:
Constructing Meanings through Action,” in
Frontiers in Social Movement Theory
,
ed. Aldon D. Morris and Carol McClurg Mueller (New Haven, CT: Yale University
Press, 1992) ; David A. Snow, Rens Vliegenthart, and Catherine Corrigall-Brown,
“Framing the French Riots: A Comparative Study of Frame Variation,”
Social Forces
85, no. 2 (2007): 385–415.

10. David A. Snow et al., “Frame Alignment Processes, Microbilization, and Movement
Participation,”
American Sociological Review
51, no. 4 (1986): 464–81.

11. For example, Robert M. Entman, “Framing: Toward Clarification of a Fractured
Paradigm,”
Journal of Communication
43, no. 4 (1993): 51–58 ; William Gamson,
Talking Politics
(Cambridge: Cambridge University Press, 1992).

12. It is difficult to come up with a form of human variation that does not have some
implication for social hierarchy. Many traits that, on first blush, seem neutral are
not so after closer examination. This is true for height, in that tall people receive
social advantages, and also for skin tone, hair texture, and even eye color, which
are all associated with race and ethnicity, or shoe size, which has gendered conno
tations.
Steven L. Gortmaker et al., “Social and Economic Consequences of
Overweight in Adolescence and Young Adulthood,”
New England Journal of
Medicine
329, no. 14 (1993): 1008–12.

13. Rebecca Popenoe, “Ideal,” in
Fat: Th
e Anthropology of an Obsession
, ed. Don Kulick
and Anne Meneley (New York: Tarcher/Penguin, 2005).

14. Emily Martin, “The Egg and the Sperm: How Science Has Constructed a Romance
Based on Stereotypical Male-Female Roles,”
Signs
16, no. 3 (1991): 485–501.

15. Julie Guthman,
Weighing In: Obesity, Food Justice, and the Limits of Capitalism
(Berkeley: University of California Press, 2011).

16. Charles Schroeder,
Fat Is Not a Four Letter Word
(Minneapolis, MN: Chronimed
Publishing, 1992) ; Marilyn Wann,
FAT!SO?: Because You Don’t Have to Apologize for
Your Size
(Berkeley, CA: Ten Speed Press, 1999) ; Charlotte Cooper,
Fat and Proud:
Th
e Politics of Size
(London: Women’s Press, 1998).

17. Schroeder,
Fat Is Not a Four Letter Word
.

18. Wann,
FAT!SO?

19. David B. Guralnik, ed.,
Webster’s New World Dictionary
, 2nd ed. (New York: Simon
and Schuster, 1984), 980. http://www.merriam-webster.com/.

20. World Health Organization, “Physical Status: The Use and Interpretation of
Anthropometry,” in
WHO Techincal Report Series No.
854 (1995), 274. It did, how
ever, use BMI cut-off s to establish three grades of “over
weight
” among adults,
including “grade 1 overweight” as 25 to 29.99, “grade 2 overweight” as BMI 30 to
39.99, and “grade 3 overweight” as BMI 40 or more.