Jeremy Ackerman-Yost
Explorer
Science is fun, kids! 
Food, Energy, Waste
There is an adage amongst biologists: "Given tightly controlled conditions of medium, humidity, light, temperature, and air pressure, the organism will do whatever the hell it wants."
Did you make sure you ate the same foods and drank the same liquids at the same times of day in both trials? Was the room temperature the same? The relative humidity? If you didn't control the environment, it may have impacted the results.
Even if you didn't control the environment, your body is a biological organism - it is dynamic, not static. You were not likely to be in the same state the second try as the first, and so maybe you held on to your water better.
We see here why science and statistics go hand in hand. One experiment that happens to go as per the hypothesis is not proof the hypothesis is correct. It may be suggestive, even demonstrative, but it is not proof. Why? Because the real world is not theory. Real world data never matches theoretical predictions to infinite levels of precision. There is variance and scatter. Scientists do experiments over and over to help average out the random elements.
When someone says, "People lose two pounds of water over night," that is probably not a 100% surety. At best it is on some sort of broad average, for some assumed normal conditions. Sometimes it'll be less, sometimes it'll be more. Do the test 100 times, and you can come up with an average for what you lose. Do the test 100 times each for 100 randomly chosen individuals, and you may start to get an average for what people do, in general.
Jeremy Ackerman-Yost
Explorer
As Umbran pointed out, the "noise" can swamp the effect size when you're looking at a particular output of a biological system. Remember, a pound isn't very much relative to your total body weight.
A salty meal or a low pressure center would be enough to throw the numbers off.
A salty meal or a low pressure center would be enough to throw the numbers off.
Whether that variation is small compared to your total body weight may not be meaningful unless the amount of weight you lose is also strongly dependent o how much you weigh.
In science, we try not to explain away anomalous variation from expectations before we have a large amount of data.
Jeremy Ackerman-Yost
Explorer
Well that was dogmatic given the context here. I'm not talking with scientists, about actual science. I'm playing "Bill Nye" here.
Do you really think anyone on this website wants to sit through a lecture on the hormones involved in water balance, how different nutrients and activities impact water balance at half a dozen different levels, body fat as a hormonally active tissue, and even bacteria populations in the gut and their proximate effects on water and nutrient retention?
Hell, I spend leisure time reading about that stuff, and I still don't want to even write that lecture.
That lecture is fun for a select audience (debatably... about 1/2 of the audience will tell you it was great, but were checking their e-mail on their iPhone less than a third of the way in, and were updating their Facebook status by the time you get to the meat of it). While a select audience might find some fun, that same lecture absolutely kills fun for most (normal) people.
Also, for the record, last I checked, body fat and surface-to-volume ratios do, in fact, account for a noticeable chunk of the variance in overnight weight loss. But the direction of the effect varies with other factors, like diet. Fun, wot? Try controlling all those factors in a human trial. This is why people end up relying on terrible self-report from huge populations.
Do you really think anyone on this website wants to sit through a lecture on the hormones involved in water balance, how different nutrients and activities impact water balance at half a dozen different levels, body fat as a hormonally active tissue, and even bacteria populations in the gut and their proximate effects on water and nutrient retention?
Hell, I spend leisure time reading about that stuff, and I still don't want to even write that lecture.
That lecture is fun for a select audience (debatably... about 1/2 of the audience will tell you it was great, but were checking their e-mail on their iPhone less than a third of the way in, and were updating their Facebook status by the time you get to the meat of it). While a select audience might find some fun, that same lecture absolutely kills fun for most (normal) people.
Also, for the record, last I checked, body fat and surface-to-volume ratios do, in fact, account for a noticeable chunk of the variance in overnight weight loss. But the direction of the effect varies with other factors, like diet. Fun, wot? Try controlling all those factors in a human trial. This is why people end up relying on terrible self-report from huge populations.
Jeremy Ackerman-Yost
Explorer
Heh. Maybe that was a little over the top. Finding a conversational tone about these things is tough, though, since the devil is in details the average person doesn't know about or usually want to know about.
I try providing the cocktail party version of some bit of research and one of my fellow scientists always seems to lean into the conversation and start "correcting" me with details. The ensuing exchange between myself and my colleague reliably causes the eyes of my original conversation partner to glaze over. At that point they excuse themselves to get another drink and to find someone who can talk at a party.
This, as much as anything, has decreased the number of academics I invite to parties.
I try providing the cocktail party version of some bit of research and one of my fellow scientists always seems to lean into the conversation and start "correcting" me with details. The ensuing exchange between myself and my colleague reliably causes the eyes of my original conversation partner to glaze over. At that point they excuse themselves to get another drink and to find someone who can talk at a party.
This, as much as anything, has decreased the number of academics I invite to parties.
Similar Threads
