Stories tagged statistics
Opps! There's one!: So I guess it's 17 now.
Courtesy Minnete LayneWell, if you were feeling anxious about there being no more undiscovered sea monsters, chill out. There are still some out there. About 18, to be specific.
See, ever since Science’s parents (Magic and Critical Thought) stopped putting Science’s stuff up on the fridge, Science has really been going out of its way to make sure we all know how special it is.
We get it, Science, you’re great. Take it easy.
As if.
Science, in its latest flailing and pathetic play for attention, has announced that there are indeed more huge, unknown sea creatures out there, and it knows that there are 18 of them.
Okay, Science, whatever you say. Act like you know.
But, no, Science goes on to explain, here’s my reasoning: If we first decide that a body length exceeding 1.8 meters defines a large sea creature (which, by the way, makes JGordon a large sea creature by 3 cm when he goes swimming), we can then look at the rate at which large sea creatures have been discovered in the last 180 years or so. The rate of discovery for large sea creatures remains pretty strong, and if you consider the places large sea creatures could be hiding, deep in the oceans, or under polar ice, say, it’s very likely that there are quite a few of them left to find. Using some flashy statistical modeling, Science predicts that there could be as many as 18 of these large sea creatures still undiscovered.
Science goes on to emphasize that there probably aren’t any monsters hiding out in lakes and lochs, and that accounts of sea serpents and their ilk can probably all be explained by known creatures, like colossal squid, and 30 plus-foot oarfish. Ah, thanks for that, Science.
Still, Science doesn’t hold all the cards. It may know that there are 18 monsters still hiding out there, but I know exactly what they are. Deal with it, Science.
Anguirus
Baragon
Destroyah
Ebirah
Gabara
Ganime
Gigan
Gorosaurus
Kamoebas
King Caesar
King Ghidorah
Kumonga
Megalon
Minilla
Mothra
Rodan
Urkel
Varan
Tags : 
Fair and balanced: Or leaning to the left, as it often does. Journalism often presents science in less-than-accurate ways.
Courtesy nick farnhill
You'll never find any of that here! ;-) But the American Association for the Advancement of Science recently discussed how the public receives and understands science news. The situation is discouraging – there’s a lot of bad information out there, much of it the result of sloppy reporting. One of the big culprits was a misunderstanding and misrepresentation of statistics.
Meanwhile, biochemist Michael White complains about how the human desire to tell a good story often misrepresents how science really works.
A story on CBS news claims that military veterans commit suicide at a much higher rate than the general population. However, blogger Bill Sweetman argues that the report is flawed. It fails to account for the fact that the vast majority of veterans are men, who have a higher suicide rate than average. Most veterans are also young, and young people commit suicide far more often than older people. Once you account for these two factors, the supposed difference in veterans’ suicide rates disappears.
Providing health care for all is no easy task.: Photo by Neimster at Flickr.com
There’s been a lot of talk about the American health care system of late. And there’s going to be a lot more talk in the months ahead, as it becomes a campaign issue in the 2008 Presidential election. Gregory Mankiw, a professor of economics at Harvard, has crunched the numbers on health care, and found that some of the issues aren’t quite what they seem.
Lower life expectancy
Demographically and economically, the United States and Canada are fairly similar. Yet Americans, on average, die about two-and-a-half years sooner than our neighbors to the north. So health care in the US must be worse, right?
Not necessarily. Mankiw found that Americans, especially younger ones, are far more likely to die in an accident or a homicide than a similar Canadian. Take that away, and the difference virtually disappears. As Mankiw states, “Maybe these differences have lessons for traffic laws and gun control, but they teach us nothing about our system of health care.”
Infant mortality
In America, a higher percentage of babies die during infancy than in other countries. Ironically, this is not a sign that American health care is worse, but rather, that it is better.
In many countries, low-weight babies who are born not breathing are considered stillborn—doctors do not try to save them. American doctors do. In fact, America has the best rate of success with low-weight babies, simply because we are willing to take on these high-risk cases. But high risk also means high failure rate: despite the doctors’ best efforts, many of these babies die anyway, raising our infant mortality rate. In other countries, the baby is not counted as ever having been alive at all, making their rate appear low.
Also, Mankiw notes that low birth weight is associated with teen pregnancy, and America has a higher teen pregnancy rate than many other countries. While there are steps we can take to reduce that phenomenon, overhauling the way we pay for health care is not going to have any effect on teenagers’ behavior.
Millions of uninsured
Many politicians have noted that some 47 million Americans – nearly one in six – has no health insurance. Some of these people want health insurance, but can’t afford it, or can’t get it through their jobs. This is a real problem.
However, Mankiw notes that this 47 million includes a lot of other groups. Millions of poor people are already eligible for Medicaid, but have simply never enrolled. Millions more have been offered health insurance through their jobs, but declined. We could reduce the 47 million substantially, without changing a thing, just by getting these folks to sign up.
Mankiw notes that a large number of uninsured are illegal immigrants. Getting these people covered is a matter of immigration reform, not health care reform.
So, when you hear politicians throwing numbers around in the health care debate, remember: the story behind the numbers is often a lot different than the sound bites make it appear.
Meanwhile, here’s a possible solution to providing health care to the uninsured.
Here are some of the most interesting perspectives on the 35W bridge collapse that I have run across in the last few days:
Cell phone network sends ominous signals - Engineers at T-Mobile were alerted that something had gone wrong right after the bridge collapse. They hadn't heard the new yet but saw a sharp change in cell phone activity on their network.
Government spending collapsed as well - A graph of US government spending on infrastructure over the last 55 years.
Historians and engineers have a thing or two to learn from each other - An editorial from 2006 of the history of engineering disasters.
Bridges made from glass - A prescient report from the National Science Foundation on poor infrastructure and the future of bridge technology.
CocaineThe Economist has an interesting graph of cocaine prices around the world. It hits a high of nearly $715/gram in New Zealand, $110/gram in the USA, and only $2/gram in Columbia where most of the world's cocaine is grown.
But in case you're tempted by the high market value of this illegal drug, Gene pointed out another revealing economics/statistics story a couple months ago about, the worst job in America. Drug dealing makes Alaskan crab fishing seems safe.
Barry Bonds: Barry Bonds and the issue of steroids in baseball are back in the news as he chases Hank Aaron's home run mark. New statistical analysis by two college students shows that steroids don't have a huge impact on hitting homers over a long period of time.Barry Bonds, home runs and steroids. They're all in the headlines again.
We’ve debated this topic a lot since last spring – do steroids actually help a baseball player to hit a ball farther?
Now some students from Gustavus Adolphus College in St. Peter, Minn., have crunched the numbers to try to provide some statistical analysis on the matter. Their quick answer to the question is “no.”
Tyler Kramer and Dan Johnson spent their January term analyzing the home run records of all the Major League hitters who’ve had 500 or more homers in their career. They divided those players into two groups: known or suspected steroid users and non-steroid users.
According to the abstract of their research posted in a Gustavus blog: “Based on the data from players that have hit 500 or more career home runs without the assistance of steroids, it is apparent that most major league players peak in their home run production between their sixth and tenth seasons. Players who use (or are accused of using) steroids have a peak much later in their career around their 11th through 17th seasons. Even though they are able to increase the productivity later in their careers there is no statistical evidence that steroid users are able to sustain this level of productivity over an extended period of time.”
In fact, the non-steroid users had a slightly higher home run average than the suspected users. The study found that admitted and presumed steroid users averaged 41.36 homers during their best five years while non-users averaged 43.38 over their best five seasons.
But the study also shows that steroids can provide a short-term burst in home run production. The top six single season home run marks belong to the steroid suspects.
“The probability of a steroid user breaking the record for most home runs in one year is much greater than a non-user,” the students also contend.
Their findings have earned enough national attention that this month they’ll be presenting their findings at the United States Conference On Teaching Statistics at Ohio State University.
What do you think of their conclusions? Have your thoughts on steroids in baseball changed at all through all this debate? Share your thoughts here with other Science Buzz readers.
A study published in the American Journal of Education showed that over one-quarter of black college freshmen were immigrants from Africa, rather than native born – twice as many as in the general population. For Ivy League schools, the figure was 40%.
Diversity initiatives and race-based financial assistance, all designed to help black Americans, have ended up helping foreigners rather than the native-born – precisely because the programs are race-based.
Conservative writer James Taranto in the Wall Street Journal argues
“at least as measured by enrollment in elite universities, black immigrants and their children are succeeding in America far more, on average, than blacks whose families have been in the U.S. for generations--i.e., the descendants of slaves. This is a strong argument against the proposition that black underachievement in America is primarily the result of present-day racism.”
While many would argue against that claim, Taranto has found some unlikely allies. In promoting diversity in higher education, liberal author Eric Alterman feels affirmative action has outlived its usefulness, and has in fact become a political liability. He would rather see it tied to class, rather than race, saying, “I’m interested in helping poor people, not black people, not Chinese people, not American Indians, not anything.”
Do you agree? Is racism still a problem that needs to be addressed directly? Or is the real problem poverty, or lack of opportunity, or something more color-blind like that? Leave a comment.
It seems like every week there’s another medical breakthrough announced in the press – only later to fizzle when additional studies show it didn’t really hold up. Why are there so many false starts?
Dr. Peter Austin of the Institute for Clinical Evaluative Sciences in Toronto says it has to do with the way researchers use statistics. All statistical studies rely on “confidence intervals” – if an event has only a 5% chance of happening at random, then doctors can be 95% confident that it isn’t a random fluke. They assume they’ve discovered a real phenomenon, and start looking for a cause.
(For instance, a coin has about a 3% chance of landing heads five times in a row. If you had a coin that did that the first time you tried it, you’d have good reason to suspect something funny was going on, and conduct more tests.)
But Dr. Austin notes that many studies run multiple tests simultaneously. When you do that, the odds of at least one test giving an unusual result, just by chance, is very high. In our coin-flipping example, if you tested 100 identical coins by tossing each 5 times, it would be perfectly normal for at least one, and probably a few, to land all heads, without anything “funny” about them at all.
The point is, you have to look at the whole test, not just selected parts of it. And doctors – and journalists – need to be more careful when presenting the results of studies, so they don’t report false relationships.
Digging dinosaurs in Utah: Courtesy Bureau of Land Management
That’s how many types of dinosaurs remain to be discovered. According to Steve Wang, a statistician at Swarthmore College, Pennsylvania, and Peter Dodson, a palaeontologist at the University of Pennsylvania, Philadelphia, at least 70% of the dinosaurs that once existed have yet to be found. They arrived at this figure by taking the known dino discoveries and plugging them in to a mathematical model that has proven successful in extrapolating data.
The scientists estimate that about half of the missing dinos will probably never be found. They lived in upland areas where fossilization is rare. Or the rocks that held their bones have been destroyed by glaciers or other Earth processes. But that leaves some 700 types of dinosaurs yet to be discovered.
Dinosaur discovery has accelerated in recent years. Nearly half of all dinosaurs known today have been dug up in just the last 20 years. Countries like China and Argentina – long inaccessible to paleontologists – have been producing many new finds. But there are plenty of other countries, particularly in Africa, that have yet to be fully explored. Wang and Dodson figure that most of the remaining dinosaur discoveries should come to light in the next 100 to 140 years.
Science Buzz and all related activities
Add a new comment