Crafting A Survey
Imagine for a moment that you are a scientist. You are an expert in your field and are about to engage in a long-term study of fat people. You are able to come up with the questions that the volunteers will be answering as truthfully as possible and also anonymously. What kind of questions would you ask? In a more broad sense, what would you look into, specifically — fatness and socioeconomic status or education or access to healthcare/fitness options or the history of fatness in the family or all of the above?
I was thinking about doing a small survey in my local area and was going to do ask volunteers to answer some questions. I am interested in all of the above things and more. So how about we do a good old- (or new-) fashioned FFF study, eh?! I’ll post some of the questions I plan on asking, and if you have any suggestions as to what I should add or modify, comment below. Then throughout the next week or two I will gather as many volunteers to answer them as possible and will report back here.
- How old are you?
- How much do you make annually?
- What is your weight and height?
- Do you follow any special dietary requirements? If so, what are they?
- Do you have a family history of “obesity”?
- Do you have access to a gym or place of fitness? If no, why not? If so, do you have a membership to a gym or are involved in fitness activities, either recreationally or in your job?
- What is your highest education achieved? What are your parents’ highest education?
- Do you have access to affordable healthcare or have health insurance?
- Do you have a primary physician?
- How often are you sick or need medical care?
- How many children do you have?
- Out of these three options, where do you most accurately live: urban, suburban, rural?
- Do you work? If so, about how many hours do you work?
- Do you eat fast food and if so, how often?
After everyone has contributed (or not) to the list, I will post a SurveyMonkey link down in the comments so that our patrons can answer if they wish. I look forward to extrapolating data!
Fierce, Freethinking Fatties 
What about access to employment? I turned down a really well paying job once because there was travel on planes involved and, well, no thank you. I don’t know if that opens a can of worms you’re not looking to study, but things like “Do you think you have been turned down for a job because of your weight” “Do you know you have been turned down for a job because of your weight?” Do you feel you eat too much, not enough, or the right amount might be good if you’re asking about fast food. I’m not a scientist so this might all be not helpful and if so, please just know I was trying to help. 🙂 I’ll answer the survey once it’s out for sure. I love helping people gather data!
I think that your questions are good ones to ask, but perhaps in a different study because discrimination isn’t what I want to look at here. Maybe I can make a different study soon to ask about discrimination, but I want to see how well this one will work before I do something that personal and intense. But thank you!!! ❤
What, if any, medical issues are you being treated for/ What, if any, medications are you currently on?
Also, Kitsune, what qualifications and training do you have around doing research? Do you have an ethics board to submit your proposal to? Have you thought of appropriate disclaimers and possible follow-up resources?
I have access to my Science Department that has a board of ethics and were I can get a sponsor to help get more attention to the study locally. I can ask the college to send out a mass email to all the enrolled students and try to catch enough participants. Once I put together the questionnaire I can go to the Dept Head and follow the established procedures.
Also, I have about two years worth of college experience in this field because I was going to enter the science field as my career. I am not, however, as well versed as bwwcarolyn is. I would be considered better than a Joe Shmow and less experienced than a pro.
I’m a data analyst and part of my work is for an IRB (Institutional Research Board). I have some thoughts:
1) Make as many of your answers as possible a set of checkboxes or radio buttons. It makes compiling your results much easier. It’s hard to compile and analyze free-form comments.
2) The question about income can be sensitive. I would definitely use check-boxes next to ranges here.
3) The question about family history of “obesity” begs for a good definition. I would ask it something like “Of your grandparents, parents, siblings and cousins – how many are fat (or very large size/overweight, if you want to avoid the word fat which can set off some people even if it shouldn’t), with the answers something like none, a few, half, most or all.
4) I think it would be really cool to include a question relating to whether they have heard of the fat acceptance movement and do they self-identify as part of it. My theory would be that people who self-identify as part of FA are more likely to have reasonable eating habits and get regular exercise.
I think your idea for a survey is really cool and I wish you luck with it. And I look forward to hearing the results!
Extending on bbwcarolyn’s #3, a “don’t know” option that would cover the various reasons a person would have for not knowing what their blood relatives look like.
This is great, I will definitely use it! Thank you.
Off topic, but I’d love to know more about your job! I work in research right now and am considering grad school/future options. Working with the IRB sounds really interesting.
I have a BS in chemistry and a lot of training in databases and computer programming. My work with the IRB is primarily to help researchers find data. Basically, I design queries. As an example, the researchers here often do retrospective chart reviews. They would ask me for a list of patients who meet a given criteria, I would query our database and give them a list. They then get permissions from the patients to look at their info and proceed. Sometimes they are looking at the costs/outcomes, trying to find treatment paths that give better outcomes at lower cost and I give them data to try to determine that. (All this is subject to some fairly rigids controls designed to protect patients and their privacy).
The Board members are senior researchers, doctors and nurses; they do the heavy lifting of looking over proposed research to protect patients from undue risk and insure the research is probably scientifically useful. The IRB staff keep track of the paperwork and that the researcher is complying with all state and federal regs.
Now I’m way off topic. I hope I answered your question; I could go on and on, if you let me.
Sorry for taking us so off topic – I really appreciate the info though! I’ve got a BA in psychology and I’m trying to find alternative options where I can work in research at a higher level without having to do a phD.
I was thinking a radio or checkbox option would be best, honestly. It would go well with sentitive questions like income or how many people in someone’s family is fat (great suggestion btw!). With the “obesity” bit, I wanted to try and stay as neurtal as possible so that it wasn’t leading, but it looks like it didn’t work. I like what you suggested much better. And I agree that adding a question about FA would be interesting. Thank you for sharing! ❤
I would suggest asking if the person has children, *then* how many, and then their ages. Little kids versus adult children that have flown the nest make a difference in lifestyle, and there’s the further possibility that children aren’t in the equation at all.
This is great, thank you. I didn’t even think about asking the age of a person’s children.
What is the purpose of the survey?
It’s supposed to be a data gathering study to determine if poverty impacts weight and also whether weight has anything to do with healthy habits (which I suspect are linked to poverty or the amount of money a person makes as well). I hypothesize that poverty and the ability and access to healthy habits are directly link, that those who are poor will engage in less healthy habits regardless of weight. I also hypothesize that weight will also correlate with education, making a U-curve, with those who are the most uneducated and the most educating being the ends.
Poverty does impact weight.
But not for the reasons many researchers cling to.
I have written several articles about poverty and being fat on my blog.
http://fivehundredpoundpeeps.blogspot.com/search?q=poverty
As far as education goes, not everyone who is educated is wealthy today. Be careful that in today’s economy that education equals cash automatically. Not anymore.
any endocrine illness or on weight influencing medications {prednisone, etc}
Any major weight changes in a short period of time?
Food allergies?
Levels of hunger?
Great suggestions. I do have a question… why ask about food allergies and levels of hunger?
I wouldn’t personally consider asking about weight influencing medications. The list of medications that have weight loss or gain as a potential side effect is nearly endless. You can’t count on survey participants to know whether or not each of what could be many medications might have such side effects, so you’re going to have wildly varying response rates and measurement error, nor could you possibly list all such relevant medications. So unless you have a very specific medication question in mind (for example perhaps a certain class of anti-depressants), I would not ask such a question.
I think asking about illnesses is more acceptable, if the questions are very specific and accessible.
I believe fat people are hungrier.
It is a physiological difference that needs explored. I have hunger pain right now and am delaying lunch purposefully time wise, because other wise too many calories will go in for the day.
http://fivehundredpoundpeeps.blogspot.com/search?q=hunger+pain
I have a concern about the “how much do you weigh” question. I refuse to step on a scale and my doctor is perfectly fine with weighing me while I stand backwards and not telling me a number. My health is not related to a number so I see no reason to have that number, unless I want to trigger an eating disorder or a deep depression/anxiety spiral. It’s not denial, it’s more of a survival technique.
i agree with this.
one way to get around it is perhaps by asking something like “when clothing shopping, are you relegated to the plus size sections of mainstream stores, plus size sections of online stores, or extended sizes online”… or something worded much more coherent and compassionate than i just said. basically i’m trying to get at something like do you fit the “mainstream” definition of fat, etc., etc.
The problem though is that the focus of my study is whether poverty and education affects weight and also this will (hopefully) be answered by people of all body sizes. I could make it so the participants can make their best guess, but if I make it optional, many people won’t answer and it will make the results limited.
I struggled with this too. 😦
i think the question about fitness centers/gyms should be worded a little different and/or expanded. for example, i do technically have access to a gym at my apartment complex but i choose not to utilize it for a variety of reasons. (it’s not as accessible as i’d like because i’m a wheelchair user, i don’t want to work out with my judgy neighbors, i don’t want to be exposed to allergens, etc.) if you’re specifically looking to see if someone is engaging in enjoyable movement, you could ask about that, and maybe have tickboxes for the types of activity accessible to them that they utilize.
Looking for movement is one of the objectives, but the other part really is whether people have /access/ in and of itself; financially and physically. It also ties in with financial stability and the question of poverty – can you (person) afford to go to a gym or have the option of taking advantage of recreational movement. A lot of poorer people do not have access to healthy options and that is what I am looking to either validate or discredit.
I never could afford a gym when I was fighting my weight gain. Also explore how people simply cannot afford the same amount of leisure and social activities which burn calories. Being poor often means being stuck at home. Life for the poor especially in our broken up communities is no longer the same. It often means isolation. Maybe tie that in. I am poor myself. And have been Very poor. A rare person at this socioeconomic level with computer access-I have nice friends. I believe poverty and stress are connected via cortisol levels in the way that American society now functions.
on the education thing–I have a 4 year college degree and almost completed paralegal school in a post bacc. program. I consider myself mostly self-educated though.
I agree with carolyn re: the check boxes and income ranges.
Additionally, are you going to survey just fat people, or a wide range of body types? As Janelle pointed out, what is the purpose of the study? to describe characteristics of fat people, or to compare them to thin people? It all depends on what your goal is.
Either way it sounds interesting 🙂 There needs to be more research out there on fatness in general beyond trying to get people thin.
I like the checkbox option, which is why I will end up just making the survey available online via SurveyMonkey. The survey will (try to) target people of all body sizes and also as wide of a financial bracket as possible.
The study supposed to be a data gathering study to determine if poverty impacts weight and also whether weight has anything to do with healthy habits (which I suspect are linked to poverty or the amount of money a person makes as well). I hypothesize that poverty and the ability and access to healthy habits are directly link, that those who are poor will engage in less healthy habits regardless of weight. I also hypothesize that weight will also correlate with education, making a U-curve, with those who are the most uneducated and the most educating being the ends.
With the education thing be careful of Charles Murray like conclusions. I read his last book on “white people” from the library. Yeah I know his dubious history but I read all sorts of things I disagree with. He had this elitist idea that those who are the wealthiest are the smartest. [read most educated] That is nonsense especially in a society where conformity is being tossed to the side. If anything being a high intelligently or out of the box thinker means you may find yourself on skid row unless you are good in computers nowadays. We have so many people who have slid down the ladder student loans and all, the education assumptions have limitations.
I’d like to know things like:
What are your hobbies?
How many hours of TV per week do you watch?
I agree that the financial stuff should be “choose a range” and I think the “how much do you weight” should be more like “choose a weight range” or “choose of the following descriptions of your body type.
With the weight ranges, I was thinking about that. The problem I see with selecting a body type is that people will want to either lie or choose an adjective that is most to their liking. With the weight range, I was wondering what increments I should put them as if I do go that route. What do you think?
Here’s what I would pick (for whatever that’s worth): 501.
Well, that didn’t work; try again. Less than 100, 101 to 125, 126 to 150, 151 to 175, 176 to 200, 201 to 250, 251 to 300, 301 to 350, 351 to 400, 400 to 500, greater than 501.
I agree with a couple of the commenters above. In no particular order:
Ethical issues are really important. And as noted by a few people, what is the purpose of the study? Re the exercise question – what are you trying to find out? Access to gyms only? Or whether they engage in physical activity – that should not be relegated to a gym. Walking, dancing in your living room, kayaking etc. With the more sensitive questions (i.e. most of the ones on your study), ethically, you should add a ‘refused’ option where people can choose not to answer. What will people get for participating? How will you recruit them? Will the sample be biased in some way – will the results be generisable?
When it comes to asking about weight, it could be worded as:
Do you consider yourself to be: 1. 15-30 lbs over your ‘ideal’ weight? 2. 30-45 lbs over your ‘ideal’ weight? 3. 45-60 lbs over? Or 4. more than 60 lbs over ‘ideal’ weight?
Of course the number ranges can be changed depending on what kind of info you’re looking for… or if you are surveying ALL body types, even thinner people, then one could start out with “Somewhat under your ‘ideal’ weight” and “At just about your ideal weight” and progress up from there.
The problem I have with this is that a person’s “ideal weight” will be different than another person of the same size. I am 350ish and I don’t have an ideal weight. On the other hand, my mother is 250 and she is 60-80lbs over her “ideal” weight. I hate to ask for weight, but with this particular study, I can’t avoid it.
I agree with you. While there might be an urge to make questions more sensitive, ultimately the more subjective you make the questions, the less you can distill from your statistical results. You can say that a parent’s weight has a certain association with number of children, but with the suggested question here, you’re saying that a person’s perceived state of overweight has an association with their number of children. And ultimately, what does that really mean? Does it mean anything? It’s certainly no longer anything about actual weight. With a survey like this, you are stuck with self report, and it’s my understanding that self report of weight is not that accurate as it is. I think making the question even more subjective would sink anything you wanted to claim based upon your results.
I don’t know how much SurveyMonkey experience you have. When you build your survey, unless you want people to either abandon it or make stuff up, do not make all the questions required. Can’t tell you how many online surveys I’ve quit because someone’s done that.
This I can agree with. The question then becomes, what questions do I make optional; all of them, only a few?
Non-response is expected on surveys, but the more accommodating methods to adjust for non-response in your analysis depend on estimates of the probability of survey selection (who gets the survey) and estimates of the probability of response. With this kind of survey however, you are not working with any design variables, so you are not guaranteed full data for any variable. Requiring basic demographic questions as mandatory would give you full data for those variables, but ultimately it potentially introduces bias because if someone chooses not to answer those demographic questions then they do not submit a survey at all. Either way, I would not make a long internet based survey with all questions required, but survey data is in general not my cup of tea so I am not an expert.
When I build surveys, I usually make all or almost all questions optional. I’ll only make an answer required if I’d rather have the person quit than not answer. About the only time I end up making a question required is if I’m doing some kind of skip logic where I only want people with Attribute A to answer all the survey questions, while those with Attribute B would be taken to the end of the survey and thanked for their time. Sort of like when political pollsters call and first ask for the oldest male or oldest female in the house before they start questioning and if they don’t get the person they want, they thank you and go on their merry way. Does that make any sense? I’m not sure I’m describing this very well.
You might find some helpful resources here as you put this together. http://www.livebinders.com/play/play?id=1051691
A little late to the party here, but you might also consider examining whether the person is a caregiver to others or not. Some people may not have children at home, but they may have elder family members that they’re caring for or even adult children with disabilities. These situations don’t typically show up in employment history, but can strongly impact the person’s stress level and ability to take time to care for themselves.