Bias (statistics)

In statistics, the word bias has at least two different senses, one referring to something considered very bad, the other referring to something that is occasionally desirable. Both mean that an estimator for some reason on average over- or underestimate what is being estimated.

Table of contents
1 The bad kind
2 The sometimes-good kind
3 See also

The bad kind

One meaning is involved in what is called a biased sample: If some elements are more likely to be chosen in the sample than others, and those that are have a higher or lower value of the quantity being estimated, the outcome will be higher or lower than the true value.

A famous case of what can go wrong when using a biased sample is found in the 1936 US presidential election polls. The Literary Digest held a poll that forecast that Alfred E. Landon would defeat Franklin Delano Roosevelt by 57% to 43%. George Gallup, using a much smaller sample (300,000 rather than 2,000,000), predicted Roosevelt would win, and he was right. What went wrong with the Literary Digest poll? They had used lists of telephone and automobile owners to select their sample. In those days, these were luxuries, so their sample consisted mainly of middle- and upper-class citizens. These voted in majority for Landon, but the lower classes voted for Roosevelt. Because their sample was biased towards wealthier citizens, their result was incorrect.

This kind of bias is usually regarded as a worse problem than statistical noise: Problems with statistical noise can be lessened by enlarging the sample, but a biased sample will not go away that easily. In particular, a meta-analysis will distill good data for studies that themselves suffer from statistical noise, but a meta-analysis of biased studies will be biased itself.

The sometimes-good kind

Another kind of bias in statistics does not involve biased samples, but does involve the use of a statistic whose average value differs from the value of the quantity being estimated. For example, suppose X1, ..., Xn are independent and identically distributed random variables, each with a normal distribution with expectation μ and variance σ2. Let

be the "sample average", and let
be a "sample variance". Then S2 is a "biased estimator" of σ2 because
However, this biased estimator is, by the commonly used criterion of "mean squared error", actually better (but only very slightly) than the unbiased estimator that results from putting n − 1 in the denominator where n appears in the definition of S2 above. Even then the square root of the unbiased estimator of the population variance is not an unbiased estimator of the population standard deviation; for a non-linear function f and an unbiased estimator U of a parameter p, f(U) is usually not an unbiased estimator of f(p).

A far more extreme case of a biased estimator being better than any unbiased estimator is well-known: Suppose X has a Poisson distribution with expectation λ. It is desired to estimate

The only function of the data constituting an unbiased estimator is
If the observed value of X is 100, then the estimate is 1, although the true value of the quantity being estimated is obviously very likely to be near 0, which is the opposite extreme. And if X is observed to be 101, then the estimate is even more absurd: it is -1, although the quantity being estimated obviously must be positive. The (biased) maximum-likelihood estimator
is much better than this unbiased estimator.

See also



In the News

Brain Cleaner: New Method Holds Promise For Treating Brain Injuries
Brain injury causes glutamate, a substance necessary for proper brain function, to flood areas surrounding the trauma and kill other cells. Weizmann Institute scientists have developed a new way to rid the brain of excess glutamate. This method -- which uses an enzyme found in blood -- could lead to new therapies not only for brain injury, but also for stroke and other conditions, and holds promise for prevention of damage from meningitis or nerve gas.

Changes In Brain, Not Age, Determine One's Ability To Focus On Task
When it comes to focusing on a task amid distractions, some folks more than 60 years old are as mentally sharp as 22-year-olds. Others struggle. Researchers at the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign have shed some light on why that is.

2002 Alaskan Quake Left Seven Areas Of California Stirred But Not Shak
New research has found evidence of tremors along non-subduction zone faults in seven California locations immediately following the magnitude 7.8 Denali earthquake in Alaska on Nov. 3, 2002. The scientists commented that their findings were the opposite of what they had expected.

Tandem Ions May Lead The Way To Better Atomic Clocks
Physicists at the Commerce Department's National Institute of Standards and Technology (NIST) have used the natural oscillations of two different types of charged atoms, or ions, confined together in a single trap, to produce the "ticks"that may power a future atomic clock.

Best Blogfights of 2006
When bloggers weren't busy taking cash bribes or collecting free laptops courtesy of Microsoft, they spent 2006 pissing in the link pool. Here's our look back at the blogosphere's best arguments, attacks, slights and scrums from the past year. By Michael Calore.

Cellphone May Not Have Killed Korean Man After All
The South Korean man whose death is being blamed on an exploding cellphone battery had injuries too extensive to be consistent with that theory, medical examiners say.

The California Backyard Orchard: A University of California Resource f
Includes information on the benefits of home orchards, how to choose a site, planting and maintenance tips, pest control, links to related resources, and more. Covers tree and shrub fruits and nuts, berries, and table grapes. Includes events and maintenance calendars, a glossary, and a guide to using the services of University of California Cooperative Extension (UCCE) Master Gardeners program.

Antibody Reduces Incidence Of Acute Rejection In High-risk Kidney Tran
Nearly 70 percent of kidney transplant patients get short-term drug therapy initially administered during surgery to help prevent rejection. In the first direct comparison of the two drugs most commonly given to ward off acute kidney rejection, an international study led by Washington University School of Medicine in St. Louis shows that one -- anti-thymocyte globulin -- is superior. The results also suggest the drug could potentially save millions of dollars in health care costs.

Overeating And Obesity Triggered By Lack Of One Gene
Neuroscience researchers demonstrate for the first time that brain-derived neurotrophic factor is an essential component of neural circuits which regulate body weight in adult mice and that its expression in two particular brain regions is required to suppress appetite.

Philanthropy Statistics
Recent statistics on charitable giving in the U.S. Topics include percentage of households that give, average amount of contributions, types of giving (such as arts, education, health, and social services), types of corporate donations, number of charitable organizations in the U.S., number of religious congregations, volunteering activities, and related data. From the National Philanthropic Trust.


MP3 Music Downloads

Preview songs, Download Free Music,Burn CDs at ITunes.com
iTunes_RGB_9mm

 


Google




InformationQuickFind.com - Find Information Fast

Links