Subjectivity and science

I am becoming a scientist. I am trained in research methods, statistics, and in the critical analysis of empirical data. As a researcher, I want to study mental health and physical health; mental illness in developing countries; immigrant communities in the United States; and religion and well-being. It would seem like I would think twice before taking on another research topic. Yet that is exactly what I did this week, as I prepared a brief paper on eating disorder classification and the revision of the Diagnostic and Statistical Manual of Mental Disorders. For once, I wanted to study eating disorders face to face, peruse the most recent literature, and put all of my scientific training on a topic that is perilously close to home.

What have I learned from this experiment/experience? I am a bit disappointed at the outcome. My scientific training has prepared me for reading correlation tables and understanding casual relationships, but it has not taught me what to do when the topic that I am reading about pertains to my own life.

The issue of "bias" crept up on me as I read about eating disorders, but not in the way that I expected. I have strong views about eating disorders based upon my own experience, but I am willing to put those aside in the face of strong empirical to the contrary. It was not this kind of bias that I struggled with as I put together my paper, but rather the kind of bias that directed my attention towards those topics that had the most personal relevance.

Usually, when I read scientific articles, I pay attention to things like the methods used, the sample size, the statistical analysis, the significance of the results. With eating disorder research, my initial reaction was different: I read much more selectively. For example: I often struggle with anxiety, so when I came across a few articles that linked eating disorders and anxiety through family studies, I was eager to understand everything that I could. There are similar studies that show connections between eating disorders and obsessive compulsive disorder, and eating disorders and mood disorders -- but because I have been trying to get my own anxiety under control, those were the articles that most caught my attention. This kind of selective reading is most decidedly not the objective approach I have been taught in school, and yet I wonder if it isn't helpful, in its own way.

One drawback to empirical studies is that they often lack a clear sense of their own meaning. What does it matter if people with anorexia are more likely to have anxiety disorders than your average Susan? To scientists, such connections teach us more about the origins of mental illness, but to individuals who suffer from anorexia, knowing that anxiety may be par for the course may be immensely reassuring, in its own way. Going back to my own example: I have recovered from my eating disorder, but I feel that I still have something to gain from addressing my anxiety. Knowing that anxiety often accompanies (or precedes) eating disorders is reassuring in that it tells me that there is an order to this mess of my mind, that it may behave in predictable ways even as it feels chaotic to me. In fact, knowing that anxiety is linked to eating disorders makes me more inclined to treat my anxiety seriously, not just as a remnant of my eating disordered past, but as something that might be worth understanding better in the here and the now.

I am curious: if you have had an eating disorder, how do you feel about reading research about them? Do you have trouble drawing the line between personal experience and scientific evidence and, if so, how do you reconcile the two? I suspect that my attitude towards ED research will change as I learn more about it -- in the brief span of writing this paper I became noticeably more focused on the data, and less on my own experience -- but I would like to know how others deal with this problem, too.


TwistedBarbie said...

Im not going to lie. I am insanely biased. I am more inclined to be annoyed at groupings of people with EDs, wanting to know more about the individuals. I also would rather read something that applies to me, citing the rest as not good because it doesnt apply well across the board when it is assumed to.
Maybe this will change as my client population shifts...

Anonymous said...

Hi Ai Lu I hope it's ok that I leave you a comment....

I think this post raises some v. interesting issues.
I do social science (undergrad) and when I started studying I was almost disappointed by the fact that scientific data was so far removed from how it subjectively feels to experience an ED or whatever other issue was at hand.

i have definitely come to appreciate the scientific method more but at the same time I think subjectivity does have an important place in the analysis of EDs.

A feminist method of analysis gives prominence to people's descriptions, in their own words, of their experiences. Feminist methods have become highly influential since they have begun to be used. And of course, food and eating disorders are feminist issues.

Another idea, it seems to me that dominant understandings of EDs in the past were influenced quite heavily by Freud but a lot of the older ideas have now been discredited... I think the fact that Freud had such an influence for such a long time, even though (its probably fair to say that) he didn't really understand EDs, illustrates the way that ideas can come to be seen as true and achieve dominance even if they are not fully true. It seems like subjective research with ED sufferers might have revealed that what they were experiencing was more complex than the desire to remain a child, or similar Freudian ideas. (I haven't read much Freud and I hope that I'm not misrepresenting what I have read here, my main point is the way in which ideas achieve dominance and influence thinking about a particular issue.)

One other thing that I've just read which seems relevant to this is the sociologist Pierre Bourdieu's work on the importance of reflexivity in research. One aspect is the way in which the richness of a description of experiencing something is often not allowed for because of the research methods used (statistics, questionnaires, data cleaning, etc) This can lead to a gap between lived experience and academic understanding of an issue. Another aspect of the theory of reflexivity is for the researcher to be continually aware of how her own experience may affect the research.

Its clear that you were aware of the possibility of bias affecting your work Ai Lu, so you were being reflexive, which is good research practice and I agree with you that subjectivity may be helpful in its own way.

hope this wasn't too rambly,
sara-j :)

Anonymous said...

I find myself trying to relate how the study is relevant to the people being studied. I think, "Why is this important?" and come to my own conclusions based on what I know. The thing is, I have read research related to eating disorders for four years and counting; very few studies are revealing anything new in my opinion. I'm kind of bored with them and wish they would focus on things like eating disorders in males, the growing population of children with eating disorders, and continue studies of women who stay in some disordered state their entire lives. Now granted, some of this research is happening, but not on a huge scale yet. I guess that's where my personal bias comes in: I want to know more about these topics in particular, but I know there's a shit-ton of other things researchers can study about eating disorders.

Ai Lu said...

Twisted Barbie:

I agree that it is frustrating to read about groupings of people with ED, instead of individuals. Take the following comment that I found in an article lately: "With one exception, there are no effective treatments for patients with eating disorders such as anorexia and bulimia nervosa. As a consequence, patients who are discharged from treatment are never free of symptoms. While they may be in remission from their eating disorder, they typically display symptoms such as perfectionism and anxiety, and most often, they relapse within a year or less."

So many gross, ugly generalizations here! To say that there are no effective treatments for ED may mean that there are none that work for everyone, but people can and do recover from eating disorders. As for the comment about other symptoms persisting, that would suggest to me that current treatments should do more than merely target body- and food-related symptoms, as most cognitive-behavioral therapies limit themselves to. Heck, if it means that I don't have an eating disorder anymore, I'm happy to be merely a perfectionist!

Hello, Sara J:

Thanks for pointing me to Bourdieu! I haven't read his work on reflexivity but I think it sounds very appropriate to the problem that I am struggling with. That "gap" between lived experience and empirical data often seems insurmountable and overwhelming. I need all of the grounding that I can get! And I think that, here, the researcher (me) is far too aware of her own experience, though I tend to look at it as a negative aspect, a "bias," instead of providing me with insight (which might be a better attitude.


It is interesting to hear from someone who has spent more time reading about EDs than I have -- you must have such a broad perspective by this point! Also, it is striking that you find such research boring, because that has been my initial reaction, even with knowing very little about what else is out there. Reading yet another article about BMI and remission is just so monotonous! I agree that a more diversified approach to research would be helpful. In my paper for school I actually wrote about why the current categories of anorexia, bulimia, and eating disorder not otherwise specified are highly problematic when most patients are classified as "not otherwise specified" AND most people go from one category to another over the course of their illness. It is hard for me to think that these categories aren't more useful to researchers, who define themselves as "anorexia people" or "bulimia experts," than to the actual people who suffer from such disorders.

Whew! I love these kinds of conversation, women!

Ai Lu

TwistedBarbie said...

Exactly! AND that has been proven WRONG by much research already. GRR.
Youre in NYC, right?
Im amidst organizing an ED related event that you might be excited about. Email me at twistedbarbie@gmail.com for more info... I dont know what yours is at this point. I would love to have you involved!

Anonymous said...

Wow, Ai Lu, guess what: I conducted a similar research study for a psychology class I took in 2004! I basically set out to prove that the criteria for anorexia nervosa was too rigid -- and proved it right, no less. I showed that the criteria for amenorrhea was a huge factor in whether or not a patient was considered AN (anorexia nervosa) or EDNOS and how this had a negative effect on those with EDNOS seeking treatment. As I imagine you already know, insurance companies are stupid and don't consider those with EDNOS "sick enough" to warrant inpatient treatment. Having dealt with this problem myself is what prompted me to conduct my study, and I ended up encountering many ladies like me who had rather horrifying tales of what they went through when seeking intensive treatment.

Anyway, I digress in making my point: I think the labels are far more important to doctors and insurance companies, not the people they allegedly belong to. Same conclusion you came to. :)

Tiptoe said...

For some reason, my comment didn't go through the other day. I'll try to remember what I wrote.

I've read eating disorder studies off and on for the last ten years. It's less frequent now but that's mainly due to not having the same access.

From that time, it's kind of been a peak and plateau type in what is studied and discussed. I'd say right now there isn't too much new out there with the most recent findings on the brain and genetics. I agree with both you and Charlynn that in general, the studies do need to be more diversified and more research on EDNOS, emotions post-recovery, use of integrated approaches for treatment to name a few.

For the most part, I think eating disorder studies are difficult. It's hard to really get a representative sample, and the high attrition rate is not helpful in finding true outcomes. That's a fairly common thread with mental illness studies anyway, but the implications are there.

As for subjectivity and objectivity, I certainly gravitate towards studies that interest me and I can perhaps relate to, but I try to get a broad picture. I like to gain different perspectives and learn from that, so that those pieces of information are always in my toolbox.