4 ms·

I don't think scaling this "experiment" is a problem anymore. There are millions of people that use calorie counting apps and websites. Among them there must be

by woobar 11y ago

I don't think scaling this "experiment" is a problem anymore. There are millions of people that use calorie counting apps and websites. Among them there must be plenty of people with consistent data. My personal anecdata: I have 5+ years of personal food intake and exercise data in MFP app. It is not hard at all. Data entry takes about 10 minutes a day. And now I know what a cup of salmon looks like. I've also ran some unscientific analysis that confirmed my biases. Major one: when I consume less calories then I burn my weight goes down, and vice versa. ;-)

richardboegli 11y ago
hehe, yep I wrote a book about it. Just Eat Less, its that simple . http://weightloss40kgin40weekswithoutexercise.com/ http://weightloss40kgin40weekswithoutexercise.com/
- nikolay 11y ago
  Loosing weight is not that hard; maintaining the loss is what's hard - the notorious Boomerang Effect.
  - woobar 11y ago
    If you buy into the "calories in/calories out" theory, there is no boomerang effect. Just don't eat more* than you burn. *) add individual adjustment (~10%?) based on your personal data.
    - nikolay 11y ago
      The notorious "calories in, calories out" is so unscientific and is in my view a criminal oversimplification. It ignores things like epigenetics, hormones, thermogenesis, metabolic diseases, digestion problems, microbiota, etc. Yes, counting calories can give you some rough idea about things, but it's so rough that it's useless for the purpose of dieting! Calories from fats and carbs don't get metabolised the same way either! When I see MDs giving such advice, I smell quackery! I cut my dietary intake by over 1,000 from fats (I was putting ghee and MCT/coconut oil in my coffee) without actually compensating or expending less energy. I didn't lose a gram of weight!
TeMPOraL 11y ago
It would be amazing to use this data in a study. I'm all for observing and analyzing unsuspecting people - that's the only reasonable way you can avoid the effect of people changing their behavior when they know they're being studied. I think that the biggest issue here is not with scaling, or even with finding a reliable subset of the data - it's with fitting such experiment into formal and ethical framework of science. Some of the issues I see: - Scientists generally design their studies, trying to amplify the visibility of studied phenomenon while filtering out as much noise as possible. Here you'd have an experiment that's not designed. My guess is that it's something most researchers are not used to nowadays. - It may be hard to reuse this technique. The first team to run this study may have some results that we could trust, but as soon as they publish and everyone knows that apps are now used to study people, users will change behavior. - It creates a whole battlefield of interest. It suddenly is in the interest of scientists and their research institutions to promote commercial apps. It's also in the interest of the app vendors to enter into agreements with researchers, because getting an interesting result will become a huge marketing opportunity. And so there will be pressure to find "proper" results. > I've also ran some unscientific analysis that confirmed my biases. Major one: when I consume less calories then I burn my weight goes down, and vice versa. ;-) I think you should publish :).

by woobar 11y ago

richardboegli 11y ago

hehe, yep I wrote a book about it. Just Eat Less, its that simple . http://weightloss40kgin40weekswithoutexercise.com/ http://weightloss40kgin40weekswithoutexercise.com/

nikolay 11y ago
Loosing weight is not that hard; maintaining the loss is what's hard - the notorious Boomerang Effect.
- woobar 11y ago
  If you buy into the "calories in/calories out" theory, there is no boomerang effect. Just don't eat more* than you burn. *) add individual adjustment (~10%?) based on your personal data.
  - nikolay 11y ago
    The notorious "calories in, calories out" is so unscientific and is in my view a criminal oversimplification. It ignores things like epigenetics, hormones, thermogenesis, metabolic diseases, digestion problems, microbiota, etc. Yes, counting calories can give you some rough idea about things, but it's so rough that it's useless for the purpose of dieting! Calories from fats and carbs don't get metabolised the same way either! When I see MDs giving such advice, I smell quackery! I cut my dietary intake by over 1,000 from fats (I was putting ghee and MCT/coconut oil in my coffee) without actually compensating or expending less energy. I didn't lose a gram of weight!

TeMPOraL 11y ago

It would be amazing to use this data in a study. I'm all for observing and analyzing unsuspecting people - that's the only reasonable way you can avoid the effect of people changing their behavior when they know they're being studied. I think that the biggest issue here is not with scaling, or even with finding a reliable subset of the data - it's with fitting such experiment into formal and ethical framework of science. Some of the issues I see: - Scientists generally design their studies, trying to amplify the visibility of studied phenomenon while filtering out as much noise as possible. Here you'd have an experiment that's not designed. My guess is that it's something most researchers are not used to nowadays. - It may be hard to reuse this technique. The first team to run this study may have some results that we could trust, but as soon as they publish and everyone knows that apps are now used to study people, users will change behavior. - It creates a whole battlefield of interest. It suddenly is in the interest of scientists and their research institutions to promote commercial apps. It's also in the interest of the app vendors to enter into agreements with researchers, because getting an interesting result will become a huge marketing opportunity. And so there will be pressure to find "proper" results. > I've also ran some unscientific analysis that confirmed my biases. Major one: when I consume less calories then I burn my weight goes down, and vice versa. ;-) I think you should publish :).