Eating Disorders, the Media, and Skepticism
A few weeks ago, my good friend and fellow skeptic Ben Radford wrote an article for Discovery News asserting that the movie Black Swan “cannot ‘encourage’ anorexia, any more than photographs of depressed patients can ‘encourage’ clinical depression.”
This was in response to Kate Torgovnick’s ABC News article (link here but beware of the streaming video that starts playing) that pointed out that pro-anorexia sites are using images from the film as “thinspiration.”
I suggest you read both (very brief) articles to decide for yourself, but I think Radford has misunderstood Torgovnick’s post. For instance, he writes “Even if seeing thin people could somehow encourage eating disorders in audiences, there’s little reason to blame this film.” However, Torgovnick doesn’t seem to blame the film at all. She says it’s great, and that the movie is actually showing a dystopia “of a subculture riddled with body issues.”
I agree. I saw the film last week and loved it for the way it horrifically portrayed a character who was repressed and obsessive. Like Torgovnick, I can appreciate a great film while also recognizing that some women could see these emaciated women as aspirational.
All that is beside the point I really want to address, though. Radford takes issue with the very idea that images in the media can contribute to eating disorders, even branding it a “myth” and linking to a post he made back in March that I’m sorry I missed the first time around.
In that article, Radford asserts that “The idea that a model, photo of a model, or Web site can “encourage” anorexia is not supported by science or research.” He cites several papers to support that assertion, and the average layperson might be easily convinced that the science is on his side.
When I read that article, though, I was immediately alarmed at this point:
For example, R.A. Botta, writing in his 1999 study, ‘Television Images and Adolescent Girls’ Body Image Disturbance’ in the Journal of Communication, noted that, ‘Concrete evidence is still necessary to show that this existence of thin images and ideals has an effect on viewers. At this point, the relationship between television exposure and body image disturbance remains in question.
First of all, I thought I recognized the surname “Botta” as a researcher who agreed with what appears to me (a layperson, FTR) to be the consensus opinion: media images are directly linked to body image. Second of all, I’m always suspicious when I see that many ellipses.
I brought up my concerns on Radford’s Facebook page where he had posted a link to his article. I didn’t have the full paper Radford got that quotation from, but I did know that many other papers cited Botta’s in support of the hypothesis that media images and body image are linked. When I asked Radford to explain how Botta’s paper supported his conclusions, he continued to quote the above truncated quotation. When I mentioned that I didn’t have the full paper, Radford said he couldn’t send it to me because it was in hard copy form, but he did add two more excerpts (caps and brackets are Radford’s):
The results replicated the study by Stice and his colleagues (1994), WHO DID NOT FIND A SIGNIFICANT RELATIONSHIP between overall media exposure and endorsement of the thin ideal.
There was NO SUPPORT for the notion that increased exposure [to thin images] leads to body image disturbance
Luckily, someone else was viewing that discussion and had access to the full paper. Jenna Marie Griffith sent it to me and I was able to verify exactly what Botta was writing. What I found surprised me.
In Television Images and Adolescent Girls’ Body Image Disturbance, Botta (who, by the way, is a woman despite Radford’s use of the masculine pronouns in the initial article), wrote in conclusion:
Overall, the results suggest that media do have an impact on body image disturbance, both directly through body image processing and indirectly by encouraging adolescent girls to endorse a thin ideal and by establishing what they see as realistic ideals.
All the quotes Radford gave were completely out of context, and I’m tempted to quote the entire paper because every word of it disagrees with Radford’s assertion that media images have no relationship to body image. Instead I’ll just give you longer excerpts with Radford’s chosen quotes unbolded and the rest of the quote in bold:
The results replicated the study by Stice and his colleagues (1994), who did not find a significant relationship between overall media exposure and the endorsement of the thin ideal.
However, by looking deeper into the relationship between media and thin ideal endorsement with more conceptually derived measures, this study revealed that the combined media variables accounted for 33% of the variance in endorsing a thin ideal. As predicted, this was because the measures went beyond mere exposure to include body image processing effects of media images. The impact on endorsing a thin ideal seems to be more about how adolescent girls process thin images than about how much they view those images.
Botta found that girls who saw the women on TV as having realistic bodies were much more likely to be bulimic than girls who saw those images as unrealistic. She also found that for girls who already saw thinness as a positive goal for women, viewing thin images increased their rates of bulimia. For girls who did not feel that women “should” be thin, viewing those images did not necessarily increase rates of bulimia. In other words, these images were incredibly damaging to girls who were processing them the wrong way. Make a mental note of this, because we’ll see it again later.
Here’s another of Radford’s quotes:
The one prediction not supported was H2 about exposure. There was no support for the notion that increased exposure leads to body image disturbance as a result of forced comparisons. This may be due to the media exposure measures in this study. Perhaps daily rather than weekly viewing measures might have been better able to detect this relationship.
H2 was only one of six hypotheses that were tested, and the only one that wasn’t supported by the results:
H1: Forced automatic comparisons will result in a direct impact of exposure on body image disturbance.
H2: Thin ideal endorsement will significantly interact with exposure as a predictor of body image disturbance, because those who are predisposed to thin ideals may be more vulnerable to automatic comparisons.
H3: Making goal-directed comparisons with media images for self and significant others will predict body image disturbance.
H4: Goal-directed processing will predict more variance in body image disturbance than mere exposure.
H5: Outside information and personal standards put to use through critical viewing will result in a significant interaction effect between endorsing a thin ideal and questioning thin media images on body image disturbance.
H6: There will also be a significant interaction effect between seeing media images as realistic and questioning thin media images on body image disturbance.
Botta goes on to suggest better ways to test H2 in the future, and makes it clear that she believes the hypothesis is still valid.
As for the first quotation Radford used in his article to suggest that Botta agreed with him, well, I’d like to show you the context around those ellipses but that would require excerpting three pages of the paper. Here’s the best I can do without potentially violating copyright. The brackets indicate where Radford’s ellipses were:
Concrete evidence is still necessary [however] to show that this existence of thin images and ideals has an effect on viewers. [1,188 words deleted] At this point, the relationship between television exposure and body image disturbance remains in question.
The “however” is required because Botta had just finished listing many studies that show how the thin ideal is getting thinner in the media. The 1,188 words that are deleted between these two sentences detail the many studies that suggest a link between the way adolescents process media images and the way they construct their own body image. Then she examines the experimental studies that have suggested a link. Botta covers a dozen studies that support her hypothesis, and that’s just within the second of Radford’s ellipses. When Botta says in her introduction that concrete evidence is still necessary, she’s making the case for why her study matters.
What of the other papers Radford cites in his Discovery News article? Well, I’m glad I asked. After his misleading sentence about Botta, he wrote:
This view is supported by other researchers including Heidi Posavac, who wrote in her 1998 Sex Roles journal study, ‘Exposure to Media Images of Female Attractiveness and Concern with Body Weight Among Young Women’ that ‘experimental investigations have not found a link between exposure to the media ideal and increased concern with weight.’
Luckily for us, that article is online in full. If you care to read it, you will experience the same realization I did: this paper comes to the exact opposite conclusion Radford claims. It is, in fact, an experimental investigation that did find a link between exposure to the media ideal and increased concern with weight:
Together, the results from the three experiments clearly demonstrate that exposure to media images of female attractiveness is capable of causing increased weight concern among most young women.
The quote that Radford once again takes out of context is in the introduction, which explains exactly what Botta found: that when you lump all women together into one experiment involving total media exposure, you don’t get significant results. When you pay attention to how these women are processing said images, you do. Posavac and her team evaluated their subjects’ body images prior to viewing images, and found that those with high levels of body dissatisfaction (the majority, btw) experienced the greatest negative impact after viewing thin images. Those women who were generally satisfied with their bodies weren’t as affected.
The only other research Radford cites is Cynthia Bulik’s 2007 study The Genetics of Anorexia, in which researchers find evidence that anorexia may have a genetic link. It is accurate that these researchers have compelling evidence to suggest that this one eating disorder is inherited (putting aside the usual concerns about twin studies), but that doesn’t have anything to do with whether or not media images impact body image. For starters, this possible genetic link is about anorexia, not bulimia or body dysmorphia. Additionally, it doesn’t suggest that sociocultural factors can’t trigger or worsen the anorexia. (Bulik does seem to believe that the genetic factors are more important than the sociocultural factors, though.)
And that’s all the research mentioned in Radford’s article. So is it a myth that images in the media are linked to body image as he claims? I’m not convinced, seeing as of his three citations, one is fairly irrelevant and two solidly rebut his claim. In fact, as much as I love Radford’s work and consider him a friend, I have to admit that I’m blown away by how out-of-context those quotes are. Some of those sentences are buried in page-long paragraphs, surrounded by the evidence that refutes what Radford is trying to assert.
In the Facebook thread, I genuinely wanted to see what evidence actually supported the idea that the link between media and body image is a myth, but I got nothing from Radford or anyone observing the thread. So, I’m forced to continue to side with what appears to be the consensus opinion: the media’s portrayal of the thin ideal most likely negatively impacts the body image of those who process those images poorly.
EDIT: Commenter foutch took a look at the paper and offers some more number crunching for the statistic fans out there.