Results for “cohort” 290 found
Small samples mean statistically significant results should usually be ignored
Genomes Unzipped: In October of 1992, genetics researchers published a potentially groundbreaking finding in Nature: a genetic variant in the angiotensin-converting enzyme ACE appeared to modify an individual’s risk of having a heart attack. This finding was notable at the time for the size of the study, which involved a total of over 500 individuals from four cohorts, and the effect size of the identified variant–in a population initially identified as low-risk for heart attack, the variant had an odds ratio of over 3 (with a corresponding p-value less than 0.0001).
Readers familiar with the history of medical association studies will be unsurprised by what happened over the next few years: initial excitement (this same polymorphism was associated with diabetes! And longevity!) was followed by inconclusive replication studies and, ultimately, disappointment. In 2000, 8 years after the initial report, a large study involving over 5,000 cases and controls found absolutely no detectable effect of the ACE polymorphism on heart attack risk.
The ACE story is not unique to the ACE polymorphism or to medical genetics; the problem is common to most fields of empirical science. If the sample size is small then statistically significant results must have big effect sizes. Combine this with a publication bias toward statistically significant results, plenty of opportunities to subset the data in various ways and lots of researchers looking at lots of data and the result is diminishing effects with increasing confidence, as beautifully shown in the figure.
For more see my post explaining Why Most Published Research Findings are False and Andrew Gelman’s paper on the statistical challenges of estimating small effects.
Addendum: Chris Blattman does his part to reduce bias. Will journal editors follow suit?
Overeducation in the UK
Chevalier and Lindley have a new paper:
During the early Nineties the proportion of UK graduates doubled over a very short period of time. This paper investigates the effect of the expansion on early labour market attainment, focusing on over-education. We define over-education by combining occupation codes and a self-reported measure for the appropriateness of the match between qualification and the job. We therefore define three groups of graduates: matched, apparently over-educated and genuinely over-educated; to compare pre- and post-expansion cohorts of graduates. We find the proportion of over-educated graduates has doubled, even though over-education wage penalties have remained stable. This suggests that the labour market accommodated most of the large expansion of university graduates. Apparently over-educated graduates are mostly indistinguishable from matched graduates, while genuinely over-educated graduates principally lack non-academic skills such as management and leadership. Additionally, genuine over-education increases unemployment by three months but has no impact of the number of jobs held. Individual unobserved heterogeneity differs between the three groups of graduates but controlling for it, does not alter these conclusions.
For the pointer I thank Alan Mattich, a loyal MR reader.
When are signaling and human capital theories of education observationally equivalent?
Going as far back as Andrew Weiss’s survey paper, there are various attempts to argue that the two theories make the same predictions about earnings and education. A randomly elevated individual will earn more money but is this from having learned more or from being pooled with a more productive set of peers?
To explore this, let’s pursue the very good question asked by Bryan Caplan:
Our story begins with a 22-year-old high school graduate with a B average. He knows an unscrupulous nerd who can hack into Harvard’s central computer and give him a fake diploma, complete with transcript. In the U.S. labor market, what is the present discounted value of that fake diploma?
If he can fake a good interview (a big if, but let’s say), and if certification from recommenders is not important in the chosen sector (another big if), he may get a Harvard-quality job for his first placement. If you believe in the signaling theory, however, his marginal product is fairly low, much lower than the wage he will be paid. They will fire him. He’ll come out a bit ahead, if he is not too demoralized, but within a few years he will be paid his marginal product.
In most jobs they figure out your productivity within two or three months after training, if not sooner.
In a one-shot static setting, signaling and human capital theories might have the same empirical implications because the learning and pooling effects can produce similar links between education and wages (again assuming someone can fake an interview). But not over time and of course the wage dispersion for an educational cohort does very much increase with time. The workers don’t keep on receiving their “average marginal product” for very long.
Do not be tricked by those who serve up one-period examples to establish the empirical equivalence of signaling and human capital theories!
To tie this back to the academic literature, if IV-elevated workers enjoy an enduring wage effect comparable to that of the other degreed workers, you should conclude they learned something comparable at school unless you wish to spin an elaborate and enduring W > MP story.
Addendum: There is a less drastic scenario than the one outlined by Bryan. Let’s say there are fourteen classes of workers and a class nine worker is randomly elevated to class seven credentials. He might use that momentary good fortune to learn from smarter peers, work hard to establish a foothold, and so on. His lifetime earnings might end up as roughly those of other class seven workers, despite being of initial type nine. The higher earnings are still based on learning effects (not mainly pooling), though pooling gave that worker temporary access to some new learning and advancement opportunities. In most regards this works like the learning model, not the pooling model, although the period of learning extends beyond schooling narrowly construed.
Dennis is Not More Likely to be a Dentist
Pelham, Mirenberg and Jones (2002) found that the names Jerry, Dennis and Walter were the 39th, 40th, and 41st most frequent male names in the 1990 census (moreover the absolute frequency of (Jerry+Walter)/2 was almost identical to that of Dennis). But in a nationwide search they found 482 dentists named Dennis but just 257 named Walter, and 270 named Jerry, a highly statistical significant difference. Hence the meme was born, “Dennis is more like to be a Dentist.”
The expected number of dentists named Dennis, however, depends not on the frequency of Dennis in the 1990 Census but on the entire stock of people named Dennis over the past ~70 years and similarly for Walter and Jerry. If, for example, no one was ever named Jerry prior to 1989 but in 1990 the name skyrocketed to prominence following the appearance of Seinfeld then there would be no dentists named Jerry despite Jerry being a popular name in the 1990 census.
Following this logic, Uri Simonsohn proposes that instead of comparing the number of dentists named Dennis to those named Jerry or Walter we compare the number of dentists named Dennis to the number of lawyers named Dennis. Making this comparison, Simonsohn finds that Dennis’s are just as overrepresented among lawyers as among dentists, thus the Dennis is a dentist finding is most likely due to a spurious cohort effect.
In addition, to testing the name-profession link Simonsohn reexamines many of the classics of the implicit egoism literature and finds many of them (not all and he does not challenge the experimental results) wanting. Virginia is not more likely to move to Virginia, for example. The Simonsohn paper is impressive and a great resource for anyone wanting to teach the difficulties of doing causal statistical research.
Assorted links
1. Following an entire cohort.
2. New Sichuan hot pot coming to GMU Arlington, right in the GMU building!
3. Systematically biased beliefs about political influence.
4. Anthony Bourdain visits Haiti (TV show).
5. The future of public buildings?
6. Ten movies about food, from Christiana Weber.
What do twin adoption studies show?
"A case in point is provided by the recent study of regular tobacco use among SATSA's twins (24). Heritability was estimated as 60% for men, only 20% for women. Separate analyses were then performed for three distinct age cohorts. For men, the heritability estimates were nearly identical for each cohort. But for women, heritability increased from zero for those born between 1910 and 1924, to 21% for those in the 1925-39 birth cohort, to 64% for the 1940-58 cohort. The authors suggested that the most plausible explanation for this finding was that "a reduction in the social restrictions on smoking in women in Sweden as the 20th century progressed permitted genetic factors increasing the risk for regular tobacco use to express themselves." If purportedly genetic factors can be so readily suppressed by social restrictions, one must ask the question, "For what conceivable purpose is the phenotypic variance being allocated?" This question is not addressed seriously by MISTRA or SATSA. The numbers, and the associated modeling, appear to be ends in themselves."
Why is the suicide rate rising for baby boomers?
Julie A. Phillips, Ashley Robin, Colleen Nugent, and Ellen Idler report (partially gated):
Our analysis of suicide rates among the middle-aged for the period 1979–2005 showed a substantial increase in suicides by men aged 50–59 years and women aged 40–59 years between 1999 and 2005, following a period of stability or decline in rates for these groups. Suicide rates also increased for younger middle-aged men between 1999 and 2005, but we found that this increase was better characterized as a continuation of previous, ongoing trends for this group. The post-1999 increase for all cohorts was found among both married and unmarried members, although the risks were higher for unmarried people. The rise was particularly dramatic for those without a college degree, while those with a college degree appeared largely protected from the trend. The timing of the increase coincided with the complete replacement of the U.S. population’s middle-age strata by the postwar baby boom cohorts, whose youngest members turned 40 years of age in 2004.
Here is a related press release and it mentions substance abuse and chronic disease (more of a rude awakening for them?). Here are some speculations about the rising rate; one possibility is that regulatory warnings have to some extent discouraged anti-depressant drugs. Here is a related paper, on measurement and considering a few possible explanations. It seems there is no comparable rise for African-Americans.
Assorted links
1. Very good blog on antiquity and medieval times.
2. Funeral cakes.
3. Very early Bruce Lee audition (physical stuff near the end)
4. Old color photos of America. Old black and white photos of Moscow.
5. How unemployment rates evolved by county, 2007-2010.
Robin Hanson receives ein Wunsch
German scientists are planning the country’s biggest biomedical study. The National Cohort will be an intensive investigation of the health, lifestyle and genetics of 200,000 people, at an estimated cost of €210m over 10 years.
…They intend to use the National Cohort “to investigate how chronic diseases are conditioned by lifestyle and environmental issues, as well as by genetic predisposition”.
“Technology has now advanced to the point at which we can use a population study to find and evaluate biomarkers and other tools for early detection of disease,” said Prof Kaaks.
These people will receive extensive medical examinations at the beginning and along the way. Here is one example of what will be done:
The German scientists are keen, for example, to discover how exercise protects against disease. “Better quantitative estimates are required of how much protection there is to be had and how much physical activity is required to obtain it,” said Prof Kaaks. That means studying a very large number of participants whose activity can be assessed regularly over many years or even decades.
Germany has less variation in health care access than does the U.S., but still this is another variable which could be studied, given this data.
Quotations by and about Paul Samuelson
You'll find a bunch here. Here is his seminal piece on public goods, three pages long. Here are quotations in appreciation of Samuelson, from Summers, Bernanke, and others. Lucas offered the following:
“Samuelson was the Julia Child of economics, somehow teaching you the basics and giving you the feeling of becoming an insider in a complex culture all at the same time. I loved the Foundations. Like so many others in my cohort, I internalized its view that if I couldn’t formulate a problem in economic theory mathematically, I didn’t know what I was doing. I came to the position that mathematical analysis is not one of many ways of doing economic theory: It is the only way. Economic theory is mathematical analysis. Everything else is just pictures and talk.”
It's mesmerizing to watch the rate at which the Twitter feed is adding messages.
You'll find free pdfs of some of his major articles here.
Assorted links
1. Scott Sumner's most absurd belief: India as #1 in gdp by 2109.
2. Click "play" and watch unemployment grow.
3. Who is Hollywood's most overpaid star, relative to box office returns? Will Farrell is #1 it seems.
4. Markets in everything: NYC McDonald's with sleek Danish furniture.
5. Saddam's strategic thinking.
6. Via Caroline Flyn, China ethnicity of the week, good for a whole year (photos, recommended).
7. The Political Economy of Trust, by Henry Farrell.
The Return of the Puppet Masters
In a post from a few years ago titled, Do you love cats?, I wrote this:
Toxoplasma gondii is a favorite parasite of evolutionary biologists because it has an incredible property. The parasite lives in the guts of cats where it sheds eggs in cat feces that are often eaten by rats. Now how to get back from the rat to the cat? Amazingly, Toxoplasma gondii infects the brains of rats making them change their behavior in a subtle way that increases the genetic fitness of the parasite. Toxoplasma makes the infected rats less scared of cats and so more likely to be eaten!
Now here is the kicker. Toxoplasma gondii also infects a lot of humans.
Now here is the latest research finding;
Toxoplasma gondii infects 20–60% of the population in most countries…We confirmed, using for the first time a prospective cohort study design, increased risk of traffic accidents in Toxoplasma-infected subjects…Our results show that …subjects with high titers of anti-Toxoplasma antibodies had a probability of a traffic accident of about 16.7%, i.e. a more than six times higher rate than Toxoplasma-free… subjects.
People with RhD blood factor have some protection – see the article for more. No word yet on whether this increases the probability of being eaten by cats although I suppose it would have to.
Defining Fat Down
Americans are more overweight than ever but Burke, Heiland and Nadler find:
…that the probability of self-classifying as overweight is significantly
lower on average in the more recent survey, for both men and women, controlling
for objective weight status and other factors….The shifts in self classification are not explained by differences between
surveys in body fatness or waist circumference, nor by shifting demographics. We
interpret the findings as evidence of a generational shift in social norms
related to body weight, and propose various mechanisms to explain such a shift,
including: (1) higher average adult BMI and adult obesity rates in the later
survey cohort, (2) higher childhood obesity rates in the later survey cohort,
and (3) public education campaigns promoting healthy body image. The welfare
implications of the observed trends in self-classification are mixed.
Airlifting Yemeni Jews
Here is a new paper:
This paper estimates the effect of the childhood environment on a large
array of social and economic outcomes lasting almost 60 years, for both
the affected cohorts and for their children. To do this, we exploit a
natural experiment provided by the 1949 Magic Carpet operation, where
over 50,000 Yemenite immigrants were airlifted to Israel. The
Yemenites, who lacked any formal schooling or knowledge of a
western-style culture or bureaucracy, believed that they were being
"redeemed," and put their trust in the Israeli authorities to make
decisions about where they should go and what they should do. As a
result, they were scattered across the country in essentially a random
fashion, and as we show, the environmental conditions faced by
immigrant children were not correlated with other factors that affected
the long-term outcomes of individuals. We construct three summary
measures of the childhood environment: 1) whether the home had running
water, sanitation and electricity; 2) whether the locality of residence
was in an urban environment with a good economic infrastructure; and 3)
whether the locality of residence was a Yemenite enclave. We find that
children who were placed in a good environment (a home with good
sanitary conditions, in a city, and outside of an ethnic enclave) were
more likely to achieve positive long-term outcomes. They were more
likely to obtain higher education, marry at an older age, have fewer
children, work at age 55, be more assimilated into Israeli society, be
less religious, and have more worldly tastes in music and food. These
effects are much more pronounced for women than for men. We find weaker
and somewhat mixed effects on health outcomes, and no effect on
political views. We do find an effect on the next generation – children
who lived in a better environment grew up to have children who achieved
higher educational attainment.
Here is an ungated version.
The future of immigration
Timothy Hatton and Jeffrey Williamson report:
This paper documents a stylized fact not well appreciated in the
literature. The Third World has been undergoing an emigration life
cycle since the 1960s, and, except for Africa, emigration rates have
been level or even declining since a peak in the late 1980s and the
early 1990s. The current economic crisis will serve only to accelerate
those trends. The paper estimates the economic and demographic
fundamentals driving these Third World emigration life cycles to the
United States since 1970 — the income gap between the US and the
sending country, the education gap between the US and the sending
country, the poverty trap, the size of the cohort at risk, and migrant
stock dynamics. It then projects the life cycle up to 2024. The
projections imply that pressure on Third World emigration over the next
two decades will not increase. It also suggests that future US
immigrants will be more African and less Hispanic than in the past.
A non-gated version is available here. A more imaginative title for this post would have been "Steve, the good news is…Steve, the bad news is…" but I'm not sure how many MR readers would get the reference. I am in any case impressed by how much African immigrants have brought to the Washington, D.C. area. Don't forget to visit Abay Market, currently the best Ethiopian place in my area. The menu has moved from having three items — raw beef only, plus fatty lamb soup — to some vegetables and cooked items as well.