Anticipation, Letdown

So there I was, all set to play my new game. My friends recommended the game, and it seemed really cool. After all of the hundreds of dollars poured into the system, and games, and the connection is no good. Oh well, a few more days won't kill me. Guess I should be studying anyway.

Don't Sleep on Your Research

In this modern information age, the dissemination of information is often misunderstood, and the research methodology is often neglected. Web sites like "Wikipedia" have replaced dictionaries and encyclopedias. Many of us take this information for granted, and do not look at it critically. Look deeper into many articles, and you may find different conclusions than the article's own "conclusion". Take, for example, these articles on "sleep"

One study that was published that attempts to relate sleep habits to the likelihood of contracting the common cold.¹ Common sense would tell us as much, but what does the study actually tell us? The study had "only" 153 participants (relatively small), and exposed them to the virus after their sleep pattern was recorded. The next study we look at relates sleep duration to obesity and weight gain.² One can easily read into it that amount of sleep causes obesity or weight gain, but that is not what the study says. It actually relates the two, but does nothing to say which causes which. Obesity can cause sleeping problems such as sleep apnea, and is in fact one of the primary causes, according to helpguide.org.³ Then there is another study that follows sleep patterns of college students.⁴ All that this study can conclude is (perhaps) "Many students have sleep problems...", but it admits that the "survey may be limited by under reporting". Last but not least we have a fine study on caffeine or melatonin on sleep and sleepiness throughout travel that spans multiple time zones.⁵ Here we find a study that is very specific, controlled, and makes no assumptions.

Don't sleep on your research, and let the blogosphere/wikipedia/myspace/etc. world turn you into a machine. There is no (proof) of artificial intelligence, so use your own. (Not to say tha any of the aforementioned data is, or is not true)

References

1. Sleep Habits and Susceptibility to the common cold. Sheldon Cohen, PhD; William J Doyle, PhD; Cuneyt M. Alper, MD; Denise Janicki-Deverts, PhD; Ronald B. Turner, MD. Archives of Internal Medicine, Vol. 169 No. 1, January 12, 2009
2. Sleep duration, general and abdominal obesity, and weight change among te older adult population of Spain. Esther Lopez-Garcia, Raquel Faubel, Luz Leon-Munoz, Maria C Zuluaga, Jose R Banegas and Fernando Rodriguez-Artalejo. American Journal of Clinical Nutrition, No. 2, 310-316, February 2008
3. Sleep Apnea, Symptoms, Causes, Diagnosis and treatment. http://www.helpguide.org/life/sleep_apnea.htm
4. Sleep Patterns of College Students at a Public University. LeAnne M. Forquer, PhD; Adrian E. Camden, BS; Krista M. Gabrrau, BS; C. Merle Johnson, PhD. Journal of American College Health, Vol. 56, NO. 5
5. Caffeine or melatonin effects on sleep and sleepiness after rapid eastward transmeridian travel. M. Beaumont, D. Batejat, C. Pierard, P. Van Beers, J.B. Denis, O. Coste, P. Doireau, F. Chauffard, J. French, and D. Lagarde. Journal of Applied Physiology 96: 50-58, 2004

Math in Real Life

Wouldn't you know, math really works in real life. In my math class my professor brought up This really great article which basically explains mathematically why human nature leads to the economic crisis's we are experiencing right now. I will sum it up for you as simple as possible.

Lets assume you work at an investment bank. Lets say, for simplicity sake that you and 99 other colleagues have 10 million dollars each to invest in "a coin flip". Heads you double your money, tails you go broke. Now your coin is special. Your coin has a 3/4 chance of coming up heads. Do you bet on your coin or on theirs? If you bet on theirs it is 1/2 you will make money and get a big bonus, 1/2 you go broke. What if you bet on yours? If you win and everyone else wins you get your bonus (3/8 chance..3/4 yours*1/2 theirs). However if you lose and everyone else wins the company is bankrupt, and still no bonus for you.

Would you take the "Smarter risk" that gives you less chance of making money?
Really?

Let's Make a Deal

After reading an obscure nonfiction book a few years ago "The Man Who Loved Only Numbers"
click here for a book review I came across the "Monty Hall Problem" which at first glance seems counter intuitive. It struck me as so simple, but yet so not obvious that I helped my nephew, who was 8 years old at the time, make a science project on it. Lo and behold, I was flipping through the channels, and the movie "21" was on (which I never watched because the book "Bringing Down the House" was so much better), and the scene was exactly this infamous problem. Small World! If you clicked on the second link and still are not convinced try this one .

Do they always have it?

Once again I was playing a game, and the only way I could possibly lose is if my opponent drew the 1 card he needed. First of all he had to have the card. Secondly he had to draw it. Lastly the game state had to be exactly as it was. This seems to happen to my friends and myself far more than sadistically probably, but is that the case? My answer is that no, it is not the case, but our own limited observation makes it seem so. If you are in Atlantic City and you see 10 black in a row, or 10 red in a row it will catch your eye, even though the probability is 1 in 500 of that happening, but do we notice the other 499 times when it does not happen? Another example would be that it seems like it always rains on the fourth of July, and New Years eve is the coldest night. If we look at the numbers it is, in fact, more along the norm.

Introduction: Sleep

After thousands of years of civilization, all of the breakthroughs that we have had in science, and scientists still cannot agree on why we sleep, and exactly what it does for us. You can ask ten different people how much sleep they get a night, or how much sleep they should get a night and you can probably get ten different responses. There are numerous studies on the subject, and there is a huge misunderstanding as to what each study means. When I look at a study, I try to dig deeper and see what is the real conclusion. Being a mathematical person, I usually look into the numbers to see if a study is large enough to account for statistical anomalies. The next thing I do is try to figure out if a proper scientific method is used to show cause and effect. Many studies take data and throw the numbers together, to see what the correlation is, neglecting to come up with a reason. Once I find a study that meets these criteria, it amazes me that people do not try to do what is recommended. Sleep, for example, is such a seemingly simple topic. Almost any study will show you that you need 6-9 hours of quality sleep per night, but then there is a wild discrepancy as to what will happen if you do not stay within those guidelines for prolonged periods of time. Most Americans neglect these basic standards.

Break it Down

Earlier I posted reservations I have with studies that are published, declaring that cause and effect are often not taken into true account. In response I was asked "Should we throw out all epidemiological studies because "some" are flawed? Or should we, which I think is the better choice, wade through the difficulties and attempt to find the truth?" My answer is that these studies should be broken down into categories.

Category I would be studies that show cause and effect, are done with strict guidelines like double blind studies, and are confirmed independently.

Category II would be studies that have a conclusion that is based in cause and effect, but the relationship is not proven, but seems probable

Category III would be studies that take data and are evaluated, and a conclusion is drawn strictly from that data, with no real hypothesis for WHY. These are by far the most dangerous, and should try to be avoided, lest they lead you down the wrong path.