Things that make me CRAY-zy
*****UPDATE 7/13*****
Okay-yyy... Folks, I'm not saying I don't know why this isn't a good study (again: at least it isn't as reported) or that I have some weird button-sorting theory. There is a solution to this post's puzzler, and you'll find it in comment #2, below. As always, you're welcome to comment pretty much whatever you want (Exhibit A: Bill), but please read comment #2 before showcasing your social sciences acumen or breathtaking lack thereof. Though it's definitely more amusing if you don't.
*****UPDATE 7/13*****
According to this and 126 other "news" stories, keeping a food diary can help you lose weight. In what I can only fervently hope was either an act of misquoting or misconstruing (I haven't read the actual study), the WaPo represents one researcher as saying: "The more food records they kept, the more they lost. Those who kept no food records lost about 9 pounds, and those who kept six or more per week lost about 18 pounds. That's a whopping difference."
I don't really need to even point out the flaw here, do I?
Let's just say this: Give me a hundred dieters. Allow me to ask each of them to sort a bunch of buttons into different piles based on shape and color, three or four times a day, every day. I can, with full confidence, guarantee you that after 20 weeks the wanna-be weight-losers who managed to complete the tasks each and every day will have lost more pounds than the ones who crapped out.
Can we now please start publishing articles about how sorting buttons causes you to lose weight?
Sheesh.
Relatedly.
Okay-yyy... Folks, I'm not saying I don't know why this isn't a good study (again: at least it isn't as reported) or that I have some weird button-sorting theory. There is a solution to this post's puzzler, and you'll find it in comment #2, below. As always, you're welcome to comment pretty much whatever you want (Exhibit A: Bill), but please read comment #2 before showcasing your social sciences acumen or breathtaking lack thereof. Though it's definitely more amusing if you don't.
*****UPDATE 7/13*****
According to this and 126 other "news" stories, keeping a food diary can help you lose weight. In what I can only fervently hope was either an act of misquoting or misconstruing (I haven't read the actual study), the WaPo represents one researcher as saying: "The more food records they kept, the more they lost. Those who kept no food records lost about 9 pounds, and those who kept six or more per week lost about 18 pounds. That's a whopping difference."
I don't really need to even point out the flaw here, do I?
Let's just say this: Give me a hundred dieters. Allow me to ask each of them to sort a bunch of buttons into different piles based on shape and color, three or four times a day, every day. I can, with full confidence, guarantee you that after 20 weeks the wanna-be weight-losers who managed to complete the tasks each and every day will have lost more pounds than the ones who crapped out.
Can we now please start publishing articles about how sorting buttons causes you to lose weight?
Sheesh.
Relatedly.
10 Comments:
-
Anonymous said...
-
- 2:22 AM
-
Anonymous said...
-
- 8:05 AM
-
Anonymous said...
-
- 10:40 PM
-
Anonymous said...
-
- 2:37 AM
-
Anonymous said...
-
- 2:45 AM
-
Anonymous said...
-
- 11:35 AM
-
esquiver said...
-
- 6:22 PM
-
esquiver said...
-
- 6:41 PM
-
Anonymous said...
-
- 12:12 AM
-
Anonymous said...
-
- 2:31 AM
Post a CommentSorry to be dense, but is the flaw related to "survivorship bias" in financial-markets research? (i.e., participants' weight loss gets measured only up to the point that they drop out of the study)
You're over-thinking it, my dear! As the study is reported (but, I hope, was not actually designed), people were assigned to the keeping-a-diary category and those who assiduously followed instructions to complete a boring and arbitrary task (that is, recording their meals) several times a day for 20 weeks were compared to those who were assigned to this category but who wimped out or were half-assed about completion.
In other words, the study took a heterogeneous group of people and then used a task that, by its nature, picked out the ones who had willpower, commitment, and self-discipline. And those are the ones who lost weight? Whoa. Knock me over with a feather.
Maybe the study is flawed. I haven't read it. Other studies have shown that underweight or thin people tend to overestimate the amount of calories they eat. OTOH, overweight people tend to underestimate the amount of calories they eat.
Most fat people go into a state of shock when they realize exactly how many calories they are eating. "You mean half-a-box of pasta is four servings rather than one or two?!"
So it could be that knowledge is weight loss.
Ah, of course.
You're too kind. (I was being dense, not "over-thinking" the issue.) Your explanation makes perfect sense.
PS I know it's the issue of flawed science-related reporting that constitutes the essence of this post, but FWIW I've just finished Kluge by Gary Marcus, who has something to say about how brains work (or don't).
While not ostensibly a self-help book, the last chapter of Kluge has a few "tips" on overcoming the brain's natural tendency to revert to instinctive (rather than rational) behaviors when we need those rational behaviors the most.
One of these tips suggests committing to a series of if-then conditions to achieve a goal, rather than committing to a goal itself. He even uses weight loss as an example: rather than recommending a hearty, "I will lose six pounds," he recommends, "If I see french fries, then I will say no to them."
There's nothing magical or genius about this recommendation per se, but it brings home the point that breaking down a goal from the abstract to the concrete is key to success.
I really am curious now as to whether anyone's ever done the button-sorting experiment and published its results.
The button-sorting diet would probably work, because people would be so busy in an activity using their hands, they'd be unable to eat. That's a classic strategy often suggested for people trying to lose weight in dealing with cravings etc. Engage in alternative activities that are incompatible with eating...As for the record keeping, don't be so quick to smugly dismiss this. This is not the first study to come to this (rather common sense) conclusion. Alot of people eat "on automatic" and are woefully ignorant (or self-delusional) about the quantity of food they eat. Actually keeping track of it can be a much needed reality check to help people make (and stick with) the long term changes needed in order to lose weight. The more in denial people stay the more they will tend to view the task as hopeless and the more they will attribute their weight problem to other things such as 'genetics', or 'fat virus' and other things outside of their control. That's one of the reasons why bizarre 'solutions' like gastric bypass has become so popular...
Italo -
I hadn't heard about that book, and I'll have a look for it (while browsing Powell's this week -- yay). There's a new "your brain: a user's guide" book out, too, which normally I'm sort of iffy about, but the interviews I've heard on NPR with the neuroscientist-authors (Aamodt and Wang) have all been quite good.
Anons 1 and 2:
I'm only being snippy with you because I'm flashing back uncomfortably to walking undergrads through a complex and ingeniously-designed study, painfully and step-by-step, only to have them, at the end, say "yeah, but I just don't buy it."
I am being smug, but that's because I'm right. You are entitled to refute me. You do not, however, accomplish that merely by calling me smug.
Please do note that I'm not casting aspersions on the ability of a food diary to help in weight loss per se. I'm merely pointing out that this study (or, more likely, the way it's misreported) does not/not establish that. A far, far better study would have Group 1 keep a food diary. Group 2 would follow an identical weight-loss regime but would be asked, after every meal, to spend an identical amount of time as Group 1 recording their non-food activities that day -- that is, to keep a regular diary. After 20 weeks, only the weights of those individuals who successfully completed the instructions in each group would be compared.
Now that's a study.
The funny thing is, the basic problem with the original study (as described) is that it's correlational, though the results are erroneously reported in causal terms. And, as frequently happens in correlational research, yet another hidden factor -- in this case, willpower/self-discipline -- also correlates with the study's variables but adds more explanatory power.
It's very easy to teach people that correlation doesn't equal causation -- and, in fact, if you only have time to learn three concepts before taking the GREs, that one will get you through the analytical writing section pretty well.
It's far, far harder, I've found, to get people to see that any given situation is merely correlational. We loooove our causation, for reasons I'm not terribly clear about, but I'm hoping those books Italo and I were just discussing will shed some light.
I'm Anon #1. Your snippiness is silly, given that I didn't argue with you. If you had read what I wrote (instead of what you wish you had been written), you would have noticed the following: "Maybe the study is flawed. I haven't read it." In other words, I was not defending the study!
I did point out, the study's flaws notwithstanding, that "overweight people tend to underestimate the amount of calories they eat."
Feel free to be smug, or snippy. But you should actually read first to ensure that you're entitled to your smugness.
Hey, listen, Indiana: You're rather conveniently forgetting to re-quote your last line. What you actually said, essentially, is "Study, schmudy. Right, wrong; who knows? But, ignoring everything you've said, here's why I think they're right."
As for your appointing yourself arbiter of my right to smugness...oh, hell, this doesn't merit a rant. I've lost interest in the conversation already.
<< Home