Newton and the mean

Ok this is really an excuse to put Isaac Newton in a post, but still this is kind of fun:

Newton did something unusual, and even, as Alan Shapiro notes, “almost [we would say entirely] unprecedented in the 17th century”: he averaged all of the differences….None of this reached print….Newton certainly avoided hinting in print that his law of arithmetical progression was adduced by anything other than the most skillful and precise of measurements.

….Newton’s “mean”—the average—was the weapon with which he slew the invevitable dragons of sensual errors. It was a most paradoxical weapon for the times, because it amounted to a method by which error seems to be reduced by committing it repeatedly. No such method appears elsewhere at the time, and it would certainly have seemed odd, to say the least, to most practitioners of the period.

….We have no contemporary record of the reasoning by which he justified this unusual method….Yet Newton used averages early on; he used them frequently and, it seems, consistently….Why did Molyneux and Flamsteed, a decade or two later, do so as well?….Is there some evidence as to what underpinned the average, decades before statistical notions became widespread?

Apparently the answer to that last question is no. The authors produce a bit of evidence that Newton thought of the average as akin to measuring a center of gravity, but that’s about it. It appears that Newton never explained himself, but just quietly went ahead with his use of  averages several decades before anyone else. It was the secret behind his famously accurate observations.

I’m not sure this is true, as Wikipedia says that Tycho Brahe did the same thing. I can’t find any primary source that says that Brahe did (in a Google search), so I don’t know if he did. Anyway …. Isaac Newton.

 

 

 

Violence is way down

A report on gun violence came out yesterday. It shows that gun violence is way down, but that’s part of the general picture where violence and crime have dropped considerably since 1993 (violent crime is down about 71% and serious violent crime is down about 75%–from here):

That’s pretty stunning. Homicides have also dropped but not as dramatically (the number is down about 38% but that underestimates the decrease because it isn’t adjusted for population; from here–I found the numbers for total homicides by dividing the number of gun homicides by the percent given in the file, thus the number could be off a bit; note the blip in 2001 is from 9/11):

The numbers aren’t dropping as quickly now as they did from 1993 to 2000, but it’s still dropping–it’s always niceto see some good news.

What exactly is the American Community Survey for?

Via here and here, it seems Republicans have decided that the American Community Survey is too intrusive (the bill is here):

This week, the U.S. House voted 232-190, mostly along party lines, to prohibit the Census Bureau from using federal funds to conduct the survey. All four of the House members from East Tennessee voted in favor of the legislation.

…

The Census Bureau introduced the survey in 2005 to replace the “long form” questionnaire that had been used in the census count taken every 10 years. The idea was that the ongoing survey would provide statistical data on a regular basis, instead of the once-a-decade figures generated under the previous questionnaire.

But Duncan and others argue the survey is so detailed that it amounts to an invasion of privacy.

Duncan said he can’t think of any reason why the government needs to know how Americans get to work, how many bedrooms are in their homes or whether or not they have hot and cold running water — all questions that are posed on the survey.

“It’s just ridiculous how detailed these questions get,” Duncan said. “It seems to me there’s just almost no privacy anymore, and it just keeps getting worse and worse.”

Yes, it’s a mystery why they ask these questions. It’s not like the Census Bureau has a site explaining why questions are included … oh wait they do. Yeah, but is there really a reason to ask how Americans get to work?

Meeting Federal Needs

Basic knowledge about commuting patterns and the characteristics of commuter travel come from responses to these questions. The commuting data are essential for planning highway improvements and developing public transportation services, as well as for designing programs  to ease traffic problems during peak periods, conserve energy, reduce pollution, and estimate and project the demand for alternative-fueled vehicles. These data are required to develop standards for reducing work-related vehicle trips and increasing passenger occupancy during peak periods of travel.

The Bureau of Economic Analysis (BEA) plans to use county-level data in computing gross commuting flows to develop place-of-residence earnings estimates from place of work estimates by industry. In addition, BEA also plans to use these data for state personal income estimates for determining federal fund allocations.

Community Benefits

Transportation

These data form the database used by state departments of transportation and the more than 350 metropolitan planning organizations responsible for comprehensive transportation planning activities.

Metropolitan planning organizations use these data to manage traffic congestion and develop strategies to mitigate congestion, such as carpooling programs and flexible work schedules.

Public transit agencies use these data to plan for transit investments, identify areas needing better transit service, determine the most efficient routes, and plan for services for disabled persons.

Emergency Preparedness

Police and fire departments use data about where people work to plan emergency services in areas of high concentrations of employment.

Employment

Data are used to identify patterns of discrimination in hiring among minorities and other population groups within labor markets.

Banking

Financial institutions use data about commuting patterns and occupation to define market areas for describing lending practices and the effects of bank mergers.

Well, perhaps but there couldn’t possibly be a reason to know whether we have hot and cold running water.

Complete plumbing facilities are defined as hot and cold running water, a flush toilet, and a bathtub or shower. These data are essential components used by the U.S. Department of Housing and Urban Development in the development of Fair Market Rents for all areas of the country. Federal agencies use this item to identify areas eligible for public assistance programs and rehabilitation loans. Public health officials use this item to locate areas in danger of ground water contamination and waterborne diseases.

Ok, but to be fair to Representative Duncan, I actually had to go to the Census Bureau’s web site to find this out. And that sounds like work. I should also note that Duncan didn’t say something as stupid as Rep. Webster:

“This is a program that intrudes on people’s lives, just like the Environmental Protection Agency or the bank regulators,” said Daniel Webster, a first-term Republican congressman from Florida who sponsored the relevant legislation.

“We’re spending $70 per person to fill this out. That’s just not cost effective,” he continued, “especially since in the end this is not a scientific survey. It’s a random survey.”

Really, he seems to think that making a survey random means it’s not scientific. He also stated the ACS is unconstitutional, here’s the Census Bureaus’s response to an earlier such question.

Not to be outdone, the Senate has introduced a bill to make filling out the ACS voluntary.

You can read more about this here, here, and here.

Medicare and Medicaid spending

Hmm, this seems like it could be a big deal:

In its most recent baseline projections, CBO reduced its estimates of spending for the Medicare and Medicaid programs compared with its estimates in the August 2012 baseline. For the 2013–2022 period, projected spending for those programs is now $382 billion (or 3½ percent) below the agency’s estimates in August 2012.

…

In recent years, health care spending has grown much more slowly both nationally and for federal programs than historical rates would have indicated. For example, in 2012, federal spending for Medicare and Medicaid was about 5 percent below the amount that CBO had projected in March 2010.

In response to that slowdown, over the past several years CBO has made a series of downward adjustments to its projections of spending for Medicaid and Medicare. For example, from the March 2010 baseline to the current baseline, technical revisions—mostly reflecting the slower growth in the programs’ spending in recent years—have lowered CBO’s estimates of federal spending for the two programs in 2020 by about $200 billion—by $126 billion for Medicare and by $78 billion for Medicaid, or by roughly 15 percent for each program.

If this isn’t a short-term blip then this is really good news. Healthcare costs are the driver behind increasing deficits, so if these costs start to moderate there will be no long-term deficit problem. Given that this is a good time for deficit spending, perhaps this news will mean politicians will wait to see if there really is a problem before they start making big cuts. Ha ha, sorry, that’s a stupid thought.

Guns and murder

I missed this article when it came out which ranks countries by guns per capita. The US is number one by a lot. Here are the top ten (data from here which got it from a UN survey–the rates are per 100 people)

United States	88.8
Yemen	54.8
Switzerland	45.7
Finland	45.3
Serbia	37.8
Cyprus	36.4
Saudi Arabia	35
Iraq	34.2
Uruguay	31.8
Sweden	31.6

It’s a weird mix: free and unfree, stable and unstable. One thing to consider is that the US is even further ahead in number of weapons–the US has an estimated 270 million guns, while the next country is India with 46 million. The US owns an astounding 35-50% of the world’s guns. These top gun countries are not near the top in murder rates (from here, as a comparison the rate in the US is 4.8 per 100,000):

Honduras	91.6
El Salvador	69.2
Côte d’Ivoire	56.9
Jamaica	52.2
Burundi	50.7
Venezuela	45.1
Belize	41.4
U.S. Virgin Islands	39.2
Guatemala	38.5
Saint Kitts and Nevis	38.2

A few things to consider: first the murder rates here are astounding, the rate in Honduras is almost 20 times that of the US which is one of the highest for developed countries, second, according to the definition, this does not include death in conflicts which means countries like Iraq and Yemen look much better than they should with their murder rates of 2 and 4.2.

I don’t have much to say about this, it’s just data to consider.

Science and gender

The NY Times has a nice graph comparing how 15-year-old girls and boys did on a science test in different countries. It shows that in certain regions boys did better than girls on average (for example, in most of the Americas and Western Europe), but girls did better in other regions (much of Asia and the Middle East). This is another point that seems to imply that much of the difference is probably social, so it’s more evidence that Larry Summers was wrong (to be fair, he said that the difference might be due to a difference in the spread, but there has also been some evidence against that).

More gun data

Slate has a project where they are trying to gather together all the gun related deaths since Newtown. As they note, it’s pretty difficult to get this information so this is just an attempt to get an idea of the scale of the problem. I looked at some of the relations as I did before and I would guess that it is undercounting suicides, because I get a correlation of -.137 between the rate of gun deaths and the Brady campaign’s score for the states (I got a correlation of -.66 (although this was age adjusted) when I looked at the full set of data, but that broke down as -.05 when looking at assault and -.75 for suicide). Still, this is a good project–it’s always better to have more data.

Assault, suicide and guns

This is a very interesting article:

In 2010, the last year for which complete numbers are available, the number of gun deaths by suicide in the United States outnumbered homicides 19,392 to 11,078. If you add up all American gun deaths that year, including accidents, 3 out of 5 people who died from gunshot wounds took their own lives. Those figures are not an anomaly: With just a few exceptions, the majority of gun deaths in the United States have been self-inflicted every year since at least 1920. This is a startling fact, and one that forces us to realize that, no matter what we may believe about the Second Amendment, the debate over how to reduce the death toll from guns is, to a great extent, a debate about suicide prevention.

…

“To some people, it’s just totally counterintuitive, because it’s so obvious that if you want to kill yourself, you can always find something else to kill yourself with,” said Barber. “What they assume is that once you’re suicidal, you remain suicidal.” But a preponderance of evidence, including interviews with suicide survivors, indicates that most suicidal acts come during a surprisingly short period during which the person is suffering an acute crisis.

“When you ask people who’ve made attempts and survived,” Miller said, “even attempts that are life threatening and would have proved lethal [without emergency medical care], what they say is, ‘It was an impulsive act, and I’m glad that I’m alive.’”

The central insight for public health researchers is that a lot of lives could be saved simply by making sure that people don’t have access to an extremely lethal weapon during that high-risk period. One striking illustration of this principle can be seen in the experience of the Israeli Defense Forces, which saw a 40 percent drop in suicides after a new rule was introduced forbidding soldiers from taking their guns home with them over the weekend. Though some soldiers may have tried to kill themselves using some less lethal method instead, it appears that scores of lives were saved.

I then went searching for more information. I found that Harvard has compiled a list of studies on gun research and specifically its connection to suicide. I then found a couple places that had compiled some gun numbers (the first link also looks at the correlation with a bunch of other variables) and that led me to the CDC and its data compiler Wonder.

Once there I played around with the numbers and got a bunch of graphs (All courtesy of: Centers for Disease Control and Prevention, National Center for Health Statistics).

First I looked at the relation between the level of gun rights in a state (as compiled by the Brady Campaign) and the total number of deaths in 2010 caused by guns (assaults, suicide, and accidental):

The correlation is -.66, which is pretty strong (the rates of death are age adjusted, but otherwise I did not try to adjust for possible confounders). It looks like more regulation reduces the number of gun deaths, but because of the article, I also looked at gun rights and suicides by gun:

Here the correlation is -.75, even stronger, so I wondered about the relation between gun regulations and murders (officially this is the number of assault deaths):

which has a correlation of only -.05. That’s pretty stunning to me. The Globe article noted that the probability of a suicide depends on the availability of a gun, so I went to look for the percent of residents who are gun owners. Using the data from here, I looked at its correlation with gun deaths and found:

a correlation of -.77 with gun regulations, a correlation of .83 with suicides by gun, a correlation of .72 with all gun deaths, and a correlation of -.03 with gun assault deaths.

Obviously this is very rough data (using grouped data like this will usually inflate the correlation, I made no attempt to control for other variables, I made no attempt to check the data (I assume the CDC data is ok), didn’t try to look at the type of gun used and made no attempt to see if the same pattern holds for counties or other region types), but I still think this can tell us that guns and suicide might be what we should look at.

Aside: here are some other correlations I found:

all deaths by gun from 1999-2004 and 2005-2010: .97 (and .9 between  1999-2004 and 2010)

assault deaths by gun and by other means: .78

suicides by gun and other: .13 (!)

total number of assault and suicide deaths: -.3 (!)

NRA really wants no facts

As I noted before, facts have an anti-gun bias so we can’t have studies about guns:

Kellermann found people turned those guns on themselves and others in the house far more often than on intruders. “In other words, a gun kept in the home was 43 times more likely to be involved in the death of a member of the household than to be used in self-defense,” he says.

Kellermann says the National Rifle Association and other Second Amendment advocates leaned on his then-employer, Emory University, to stop the research. That didn’t work.

So, he says, “they turned to a softer target, which was the [Centers for Disease Control and Prevention], the organization that was funding much of this work. And although gun injury prevention research was never more than a tiny percentage of the CDC’s research budget, it was enough to bring them under the fire of the NRA.”

Lawmakers — both Democrats and Republicans — held back some money from the CDC and made clear that no federal funds should be used to promote gun control.

…

In 2003, Rep. Todd Tiahrt, a Republican from Kansas, added language to the Justice Department’s annual spending bill. It says the Bureau of Alcohol, Tobacco, Firearms and Explosives can’t release information used to trace guns involved in crime to researchers and members of the public. It also requires the FBI to destroy records on people approved to buy guns within 24 hours.

Separately, guns mean freedom (via here):

Patriots: The following represents a voluntary set of conditions to which every single Patriot who accepts a residence in the Citadel must agree, in writing. This is a voluntary Agreement. The Citadel is a martial endeavor designed to protect Residents in times of peril (natural or man-made).  The Citadel will be built as a fortified bastion of Liberty. The Citadel is not the best housing solution for everyone. Read carefully.

The purpose of this agreement is two-fold. Firstly, it informs all Residents of their obligations to their fellow Patriots. Secondly, this Agreement will assist individuals/families in making an informed decision to opt-out of residency in the Citadel Community in the event our lifestyle is incompatible with you and/or your family.

Preamble: We the People come together in this covenant of our free will and do pledge our Lives, our Fortunes and our Sacred Honor to defend one another and Jefferson’s Rightful Liberty, defined as unobstructed action according to our will within limits drawn around us by the equal rights of others.  We further affirm our intent to abide all Constitutional laws of the United States of America and our State of residence.

One: Once ratified, this list of conditions may only be altered by unanimous consent of all parties governed by this Agreement, i.e. full-time Residents. Full-time residency will be defined by a select panel of full-time Residents who will also establish guidelines for acceptable exceptions for absences.

Two: Every able-bodied Patriot aged 13 and older governed by this Agreement shall annually demonstrate proficiency with the rifle of his/her choice by hitting a man-sized steel target at 100 yards with open sights at the Citadel range. Each Resident shall have 10 shots and must hit the target at least 7 times.

Three: Every able-bodied Patriot aged 13 and older governed by this Agreement shall annually demonstrate proficiency with a handgun of choice by hitting a man-sized steel target at 25 yards with open sights at the Citadel range.  Each Resident shall have 10 shots and must hit the target at least 7 times.

Four:  Every able-bodied Patriot of age within the Citadel will maintain one AR15 variant in 5.56mm NATO, at least 5 magazines and 1,000 rounds of ammunition. The responsibility for maintaining functional arms and ammunition levels for every member of the household shall fall to the head of household.  Every able-bodied Patriot will be responsible for maintaining a Tactical Go Bag or Muster Kit to satisfy the Minuteman concept. Details TBD and posted elsewhere.

Five: Every household within the Citadel will remain stocked with sufficient food, water, and other preparedness essentials (to be detailed elsewhere) to sustain the needs for every member of the household for a duration of one year.

Six: Every able-bodied Patriot aged 13 and older shall pass a class every three years sponsored by the Citadel on basic emergency medical care; the courses shall focus on battlefield and wilderness environments.

Seven: Every child attending Citadel schools — with parental discretion for maturity — shall have as part of every semester’s class curriculum basic marksmanship and firearms safety training leading to the proficiency test on the child’s 13th birthday as a “Coming of Age” rite of passage (a parent may attend the classes or personally train the child if so desired). Children attending schools beyond the Citadel, or home-schooled, are required to meet the same proficiency standards at age 13.

Eight: All Patriots, who are of age and are not legally restricted from bearing firearms, shall agree to remain armed with a loaded sidearm whenever visiting the Citadel Town Center. Firearm shall be on-the-person and under the control of the Resident, not merely stored in a vehicle.

Nine: Each household will provide ONE able-bodied Patriot (aged 13 or older) who shall muster one Saturday per month for Martial/Support Training for neighborhood-level training & musters, as set forth by the Militia Commanders of the Community. No single Patriot shall be required to muster more than once per quarter. In the course of every calendar year every able-bodied Patriot and every full-time resident in each household must participate in at least three musters (one neighborhood-level and two full-scale).  Part-time residents must participate, in good faith, to the best of their ability.

Ten: Twice annually a full-scale Town Defense Drill will be held for all households and all residents (once in winter weather & once in summer weather).

Eleven: The Town Militia will hold an award/recognition ceremony on April 19th each year to celebrate Patriot’s Day.

Twelve: Violations of this Agreement will result in review by an arbitration panel consisting of Citadel Residents with appropriate and proportional disciplinary action taken.  The most severe disciplinary action may include the loss of Lease and expulsion from the community.

Thirteen: All Citadel Citizens agree to accept dispute resolution by an arbitration panel in the event a problem that cannot be resolved between residents.

Facts have an anti-gun bias

This is instructive:

Gun violence is just one of many factors contributing to lower US life expectancy, but the report took on urgency because it comes less than a month after the shooting deaths of 26 people at an elementary school in Newtown, Conn.

The United States has about six violent deaths per 100,000 residents. None of the 16 other countries included in the review came anywhere close to that. Finland was next, with slightly more than two violent deaths per 100,000 residents.

…

The National Rifle Association did not immediately return calls seeking comment about the report, but in the past gun-rights advocates have fought any suggestion that firearms ownership has public health implications, and they have won cuts in the government’s budget for such research.

The NRA is so sure that guns aren’t the problem that they are against research to see if guns are part of the problem. Facts like this are obviously biased:

The researchers said there is little evidence that violent acts occur more frequently in the United States than elsewhere. It’s the lethality of those attacks that stands out.

Election results

I’m at least somewhat happy after last night:

• President Obama won reelection, winning at least 303 electoral votes (right now, he’s ahead in Florida but it hasn’t been called yet) and he’s also ahead in the popular vote (by about 2.6 million votes right now).
• Democrats widened their majority in the Senate –it was 51 Dem, 47 Rep, 2 Other (with both caucusing with Democrats), now it’s 52 D, 45 R, 2 Other (one will continue to caucus with Democrats, the other is new but leans Democratic) with one race still undecided (the Democrat, Heitkamp, is slightly ahead).
• A person who I really like, Elizabeth Warren, won her Senate Race replacing a Republican and becoming the first female Senator in Massachusetts.
• It looks like Democrats will gain a few seats in the House–the breakdown was 193 D, 242 R at the beginning of the last term and 190 D, 240 R before the election (5 seats were vacant); right now it is 192 D, 232 R with not all races decided.

Obviously this is not a great night for a liberal like me, but it’s as good as I expected and better than I feared.

I haven’t looked at all the ballot questions, but I did look at the ones that concerned same-sex marriage and they give good news: same-sex marriage won in Maine and Maryland, is ahead in Washington, and a bill to ban same-sex marriage lost in Minnesota. After Massachusetts opponents railed against activist judges. Then same-sex marriage was approved by some legislatures, so opponents started complaining that there should be put to popular vote because same-sex marriage was always voted down in such votes. Now that some states have approved same-sex marriage with popular vote–I wonder what opponents will say now.

Now look at how two predictors did. Nate Silver predicted that Obama would win 313-225 in electoral votes and 50.8%-48.3% of the popular vote–right now it’s 303-206 and 50-48% (with Silver getting all the states correct, except possibly Florida where Obama leads but hasn’t been declared the winner). How about that guy who decided that the polls were skewed? His predictions: Romeny would win with 275-263 in electoral votes and 51-48% in the popular vote . Oops, I guess the results must be skewed also.

Nate Silver and statistics

I really like Nate Silver’s 538 blog at the NY Times (I don’t always read it though, because I don’t subscribe to the Times). He is a typical professional statistician, which means he will tell you exactly how he gets his estimates (leaving out a few of his details since he is trying to make a living). As Charles Pierce notes, this is a battle between pundits and statisticians in general just as it was when Silver was in baseball:

It was while I was chasing the people who chase the spheroids, oblong and  otherwise, that the brutal hooley sprang up between the Old Baseball Hands and  the new, stat-conscious, SABR-ized people who insisted that Everything You Knew  Was Wrong.

That’s why we get such stupid arguments as this:

Brooks doubled down on this charge in a column last week: “I should treat polls as a fuzzy snapshot of a moment in time. I should not read them, and think I understand the future,” he wrote. “If there’s one thing we know, it’s that even experts with fancy computer models are terrible at predicting human behavior.”

On MSNBC’s “Morning Joe” today, Joe Scarborough took a more direct shot, effectively calling Silver an ideologue and “a joke.”

“Nate Silver says this is a 73.6 percent chance that the president is going to win? Nobody in that campaign thinks they have a 73 percent chance — they think they have a 50.1 percent chance of winning. And you talk to the Romney people, it’s the same thing,” Scarborough said. “Both sides understand that it is close, and it could go either way. And anybody that thinks that this race is anything but a tossup right now is such an ideologue, they should be kept away from typewriters, computers, laptops and microphones for the next 10 days, because they’re jokes.”

and such idiocy as this:

For all the confidence Silver puts in his predictions, he often gives the impression of hedging. Which, given all the variables involved in a presidential election, isn’t surprising. For this reason and others — and this may shock the coffee-drinking NPR types of Seattle, San Francisco and Madison, Wis. — more than a few political pundits and reporters, including some of his own colleagues, believe Silver is highly overrated.

No, you stupid idiot, Silver is using statistics and so there’s always the possibility that he could be wrong, all statisticians hedge–especially when it comes to polling, as he notes:

For a variety of reasons, the magnitude of error associated with elections outcomes is higher than what pollsters usually report. For instance, in polls of Senate elections since 1998 conducted in the final three weeks of the campaign, the average error in predicting the margin between the two candidates has been about 5 points, which would translate into a roughly 6-point margin of error. This may be twice as high as the 3- or 4-percent margins of error that pollsters typically report, which reflects only sample variance, but not other ambiguities inherent to polling. Combining polls together may diminish this margin of error, but their errors are sometimes correlated, and they are nevertheless not as accurate as their margins-of-error would imply.

Joe Scarborough, by the way, just sounds stupid. As an example look at Silver’s prediction for Virgina: 50% for Obama and 49.4% for Romney; probability of Obama winning is 58%. In other words, there is a small difference, but the probability that Obama will win is a fair amount more than 50%–think of it this way, if a coin is unfair where the probability of getting a head is .505 then the probability of getting more than 50% heads is much more than 50.5%.

Update: I wanted to make that last bit a bit clearer. Suppose the probability of a head is .505 then the probability of getting more than half heads in n tosses is:

• .389 for 10 tosses
• .5 for 100 tosses
• .612 for 1000 tosses
• .839 for 10,000 tosses

The idea is that a small difference can still lead to large probabilities (the idea works the other way also, if we tossed a coin 10,000 times and got 5050 heads we would have about 84% confidence that the population percent is more than .5 ).

Global Warming

Via Kevin Drum, it seems some people are trying to lie with statistics. Here’s the picture you need to look at (via here):

Do you see a plateau at the end there? Neither do I. If you’re in the business of saying Global Warming is a hoax, you do and this says more about you than the science.

More jobs

So the jobs picture is a bit better:

The unemployment rate decreased to 7.8 percent in September, and total nonfarm payroll employment rose by 114,000, the U.S. Bureau of Labor Statistics reported today. Employment increased in health care and in transportation and warehousing but changed little in most other major industries.

Of course, the numbers tell a different story because they come from different surveys:

Total employment rose by 873,000 in September, following 3 months of little  change. The employment-population ratio increased by 0.4 percentage point to  58.7 percent, after edging down in the prior 2 months. The overall trend in  the employment-population ratio for this year has been flat. The civilian labor  force rose by 418,000 to 155.1 million in September, while the labor force  participation rate was little changed at 63.6 percent.

So, according to one survey employment went up 114,000 and up 873,000 by the other.  This could be because they measure two different things–more succinctly given here:

Why are there two monthly measures of employment?

The household survey and establishment survey both produce sample-based estimates of employment and both have strengths and limitations. The establishment survey employment series has a smaller margin of error on the measurement of month-to- month change than the household survey because of its much larger sample size. An over-the-month employment change of about 100,000 is statistically significant in the establishment survey, while the threshold for a statistically significant change in the household survey is about 400,000. However, the household survey has a more expansive scope than the establishment survey because it includes the self- employed, unpaid family workers, agricultural workers, and private household workers, who are excluded by the establishment survey. The household survey also provides estimates of employment for demographic groups.

But, of course, the discrepancy has led to complaints by conservatives:

The leader of the “job truther” movement: former GE CEO Jack Welch.

“Unbelievable jobs numbers..these Chicago guys will do anything..can’t debate so change numbers,” he said on Twitter.

He had some friends in Congress too. Rep. Allen West (R-FL) tweeted “I agree with former GE CEO Jack Welch, Chicago style politics is at work here.” He added on Facebook that the jobs report was  “Orwellian to say the least and representative of Saul Alinsky tactics from the book ‘Rules for Radicals.’”

FOX News’ Stuart Varney apparently sensed where his audience was going. Within minutes of their release he told viewers that “there is widespread mistrust of this report and these numbers.”

“How convenient the rate drops below 8% [for the] first time in 43 months, five weeks before the election,” he added later.

You would think if they were cooking the books they wouldn’t get: ‘The overall trend in  the employment-population ratio for this year has been flat’.

Science faculty bias against women

It seems that gender bias is alive and well:

In a randomized double-blind study (n = 127), science faculty from research-intensive universities rated the application materials of a student—who was randomly assigned either a male or female name—for a laboratory manager position. Faculty participants rated the male applicant as significantly more competent and hireable than the (identical) female applicant. These participants also selected a higher starting salary and offered more career mentoring to the male applicant. The gender of the faculty participants did not affect responses, such that female and male faculty were equally likely to exhibit bias against the female student.

This is another data point showing overt bias, as opposed to genetics, is what leads to gender differences.

On polling

As I’ve noted before, political polling is really hard. When you’re trying to predict who will win an election there are so many things that you have to worry about that it’s near impossible to adjust poll numbers. That seems to elude some people:

Basically, what UnskewedPolls does is take a poll and look at the  partisan percentage split that the poll reports from the random sample.  This could be, for instance, a 33 percent Democrat, 28 percent Republican, and  39 percent independent breakdown.

Then, they somehow weight the data so that what comes out is what the sample would look like if the  Democrat/Republican/Independent percentages were equal to the  results found by Rasmussen Reports in the single month of August— 35.4  percent Republican, 34 percent Democrat, and 30.5 percent Independent.

This is wrong for a few reasons, such as:

Even the founder of Rasmussen Reports, whose surveys often show higher Republican numbers, cast doubt on Chambers’ methods: Scott Rasmussen told BuzzFeed in an e-mail that “you cannot compare partisan weighting from one polling firm to another.”

“Different firms ask about partisan affiliation in different ways,” explained Rasmussen. “Some ask how you are registered. Some ask what you consider yourselves. Some push for leaners, others do not. Some ask it at the beginning of a survey which provides a more stable response while others ask it at the end.”

So a Republican for Rasmussen might not be a Republican under another poll.

Another problem is that Rasmussen is an outlier as the Business Insider article shows, so how do we know that that’s the right one? Well, we don’t–that’s why we should trust the average of lots of polls more than any one poll.

Finally, the implication is that most pollsters are skewing things towards Obama which doesn’t make sense. Every time a poll is wrong by a significant amount their reputation goes down (and Unskewed is saying they’re off by more than 10%). Their business rests on them being accurate, do you really think they would risk that to show Obama is ahead in their polls (and this would be almost all pollsters)? I guess you might if you’re a Republican.

Income and insurance

The Census Bureau has a new report out about income, poverty, and insurance coverage. Here are some highlights:

Insurance:

• The number of people with health insurance increased to 260.2 million in 2011 from 256.6 million in 2010, as did the percentage of people with health insurance (84.3 percent in  2011, 83.7 percent in 2010).
• The percentage of people covered by private health insurance in 2011 was not statistically different from 2010, at 63.9 percent. This was the first time in the last 10 years that the rate of private health insurance coverage has not decreased. The  percentage covered by employment-based health insurance in 2011 was not statistically different from 2010, at 55.1 percent.
• The percentage of people covered by government health insurance increased from 31.2 percent to 32.2 percent. The percentage covered by Medicaid increased from 15.8 percent in 2010 to 16.5 percent in 2011. The percentage covered by Medicare also rose over  the period, from 14.6 percent to 15.2 percent. The percentage covered by Medicaid in 2011 was higher than the percentage covered by Medicare.

oh and the percent without insurance in the 19-25 year age group went down by 2.2%–thanks Obamacare.

Income:

• Median family household income declined by 1.7 percent in real terms between 2010 and 2011 to $62,273. The change in the median income of nonfamily households was not statistically significant.
• In 2011, real median household income was 8.1 percent lower than in 2007, the year before the most recent recession, and was 8.9 percent lower than the median household income peak that occurred in 1999. The two percentages are not statistically different from one another.
• Income inequality also increased between 2010 and 2011 when measured by shares of aggregate household income received by quintiles. The aggregate share of income declined for the middle and fourth quintiles. The share of aggregate income increased 1.6 percent for the highest quintile and within the highest quintile, the share of aggregate income for the top 5 percent increased 4.9 percent. The changes in the shares of aggregate income for the lowest two quintiles were not statistically significant.

so the economy is still quite weak–unemployment is still a bigger problem than the deficit. And there is no class-warfare in the US but one group seems to be winning.

Poverty:

• In 2011, the family poverty rate and the number of families in poverty were 11.8 percent and 9.5 million, respectively, both not statistically different from the 2010 estimates.
• In 2011, 6.2 percent of married-couple families, 31.2 percent of families with a female householder and 16.1 percent of families with a male householder lived in poverty. Neither the poverty rates nor the estimates of the number of families  in poverty for these three family types showed any statistically significant change between 2010 and 2011.

That’s pretty pathetic, maybe we can start talking about poverty in the US?

How about a double fake

This is interesting:

Before getting their prescriptions, half of the men were told about a small risk of erectile dysfunction and other sexual side effects; none of the men had a history of such troubles. Of the men who got the warning, about 44 percent developed these symptoms. But among those kept in the dark, the rate was far lower: only about 15 percent.

It’s known as the nocebo effect: When patients are warned of possible pain and unpleasant side effects, it increases the likelihood that they’ll experience them. If the placebo effect—in which positive patient expectations can be therapeutic—is relatively famous and well explored, the nocebo effect is its little-known evil twin.

It is both encouraging and discouraging because it shows how much of healing is affected by the mind. It also shows the usual interplay of two or more ideas that are both thought to be good:

“There was a push in the field to tell patients every possible risk,” says Dr. Frank Chessa, associate director of the ethics and professionalism course at Tufts Medical School. “What do you need to tell the patient? You need to tell the patient everything.”

But what doctors say, it turns out, can have unintended health consequences. Even subjects who get a placebo in drug trials routinely experience precisely those side effects listed on the informed consent documents.

The easy solution would seem to be to not mention side effects unless it’s necessary but, as the article notes, that’s not easy:

side effects do occur independent of a nocebo effect, and many are “mild” only “as long as the patient stays in bed.” Even a small risk of drug-induced drowsiness is serious for a school bus driver, for instance.

From a legal perspective, then, telling a patient about every possible side effect is safer. There are other pragmatic concerns as well: Making nuanced decisions about contextual informed consent rather than routinely telling every patient everything takes time and more personal attention, in an era when doctors’ time per patient is on the decline. “It requires a collaborative relationship with the patient, to explain what the nocebo effect is,” says Dr. Wayne Altman, the director of student education in the department of family medicine at Tufts Medical School, who teaches the importance of encouraging positive thinking in conversations with patients. “Because if they don’t trust what I’m telling them, they can just go home and look it up online in two minutes.”

and thus we get to the really issue: part of being a good doctor is spending time with patients and that’s not allowed in our current system. If doctors were able to spend more time with their patients then this power of the mind could be, perhaps, be directed and reduce the need for strong medication and reduce the number of unwelcome side effects.

Just go to sleep

At first glance, this doesn’t sound very helpful:

Yaffe recently conducted a series of studies evaluating more than 1,300 adults older than 75, initially assessing their sleep patterns and, five years later, their cognitive abilities. She found that those with sleep-disordered breathing or sleep apnea had more than twice the odds of developing dementia years later.

Those who developed disruptions of their circadian rhythm were also at increased risk. So were those who awoke throughout the night, tossing and turning. The findings were presented at the annual conference of the Alzheimer’s Association.

So now, not only do you have the problem of falling to sleep but you also have to worry that not falling asleep might lead to Alzheimer’s–that’ll help you to fall asleep. If you look at it closely though it is good news:

Older adults can be routinely screened for sleep problems. And, if diagnosed early, treatments can help them sleep better and possibly, down the line, reduce the risk of cognitive decline.

There are a few ways the association might work:

• problems sleeping leads to cognitive problems. If this is true, then it’s really good news since there are known and effective treatments of sleep disorders.
• the conditions that lead to cognitive decline also lead to sleep disorders (so, somewhat, cognitive decline causes the sleep disorders) or a third cause (a confounding factor) leads to both cognitive decline and sleep disorder. In this case, the sleep disorder acts as an early warning signal of cognitive decline and treatment for cognitive decline can begin earlier.

Discrimination

Via Digby, this is indeed depressing:

Both whites and blacks agree that anti-black racism has decreased over the last 60 years, according to the study. However, whites believe that anti-white racism  has increased and is now a bigger problem than anti-black racism.

It’s from a fairly small sample (only 417 people in total), but the results are crazy–whites now believe we are now discriminated against more often? Really/

It also led me to the blog by one of the authors (Sam Sommers ) here which is quite interesting and has some relevant articles.

This one notes that we’re often hesitant to believe people are racist so deny that a particular instance is racism:

• Most employers aren’t racists, so racial disparity in a company’s hiring tendencies must result from other factors, like there simply not being enough strong applications from qualified minority candidates.

• Few attorneys or judges are bigots, so what look to be racial differences in  say, how they evaluate potential jurors must result from other, race-neutral considerations in their jury selection calculations.

• The arresting officer used to run diversity training sessions for his colleagues, so race couldn’t have played a role in his decision to arrest the Black professor who was legally inside his own home–the professor must have been a disorderly jerk who warranted arrest.

• Race has nothing to do with it; we just don’t believe that the dark-skinned president with the funny name was born in this country (or that many of his supporters are “real Americans,” for that matter).

and he notes that we can test to see if there is inherent racism in certain cases:

Résumés with Black-sounding names get 50% fewer call-backs than résumés with White-sounding names.  The same juror background is seen more positively by a prosecutor when the juror is White than when the juror is Black.

In another post, he notes that most people don’t think they’re racist so asking whether a person is racist can be counter productive:

 A few years ago I and a colleague published a series of studies looking at how people define “racist.”  The answer?  We set the bar just past where we ourselves are.  So what makes someone a racist?  You may not know, but you do know it’s not you.

and he gives more examples of studies that show that racism still exists:

this is also what the behavioral science does tell us:  That respondents in a study who first see a Black face are more likely than those who first see a White face to mistakenly think an ambiguous object subsequently presented is a gun.  And participants completing a video game-like police simulation perform similarly, becoming more likely to mistakenly push the “shoot” button when an unarmed suspect is Black than when he’s White.  For that matter, subliminal presentation of crime-related images–shown to respondents so quickly that they don’t consciously know what they’ve just seen–makes people pay more attention to Black faces shown next, the mere unconscious suggestion of crime being enough to activate visual processing related to race.

In a partially unrelated post (this one is about sexism), we see that perception isn’t always about reality (the study based on video surveillance found that women parked better in some objective criteria):

And if you ever need a reminder about the power of expectation and assumption with respect to sex difference, just look at the poll at the bottom of the MSNBC story about the parking garage research.  When asked which sex is better at parking, even after reading about the garage study, 38% go with men and only 25% pick women.  When it comes to sex differences, we often let our intuition cloud our judgment, data be damned.