Saturday, January 19, 2013
What The Daily Beast’s Absurd Vaccine Truther Screed Tells Us About Journalism: pI’m not going to link to the execrable anti-vaccine screed published on The Daily Beast today. I’m not even going to link to the thoughtful, well-written counterpoint they published by a infectious disease specialist. To do either would reward a transparent attempt to gin up a pageview-inducing “controversy.” Moreover, it would treat the two pieces [...]/p
Posted by Statgirl Tells The Truth at 11:18 AM
Thursday, January 3, 2013
Happy 2013! Children born during recessions more likely to exhibit problem behaviors? More scaremongering for all those parents of children born in 2008, 2009, AND those children were unwanted and cared for by unavailable working parents or Spot the Lousy Logic (a fun and interactive game)
WARNING - This made me angry so it may read like a rant...
The article, "Do Recession Babies Grow Up to Be Troubled Teens?" suggests that children born during the recessionary period in the early 1980s had higher rates of substance abuse, theft and other behaviors. All of the children? No. It was only those children from areas with higher levels of unemployment. So was it higher than other areas or compared to another time period? Not clear. Were the children still living in higher than normal unemployment areas at the time the study tracked them? Also not clear if anyone controlled for that.
“The risk for being arrested, joining a gang, smoking pot, stealing, drinking, and smoking were all slightly higher (by 6 to 17 percent) for kids who were born in or spent their first few years in areas with high unemployment rates, even if their families were wealthy or not unemployed -- and even though the U.S. economy was well on the way to recovery by 1997, when the teens surveyed were exhibiting their less-than-stellar behavior”
"It basically went across all socioeconomic strata," Ramanathan said. Since the increase in risky behavior wasn't limited to one area of the country or one socioeconomic class, "From a national level, it seems like everyone is affected," she added.
Anyone spot the two logical inconsistencies here?
1.The risky behavior occurred where the birth was in an area of high unemployment, but because there were pockets of high unemployment in various metro areas, “it seems like everyone is affected?” Sorry, no. That’s a contradiction. The speaker mentions a specific population among whom the behavior was observed and in the enxt breath says, “everyone.” My mother says, “never say never.” I say “never say EVERYONE!” Once you say that, you subject yourself to the now famous White Swan counterfactual.
You can’t get there from those data. It’s a problem in specific areas where unemployment is high – so that means focusing on those high unemployment areas for public policy not lining up kids by age cohort and saying, “Born in 1980? Uh-oh. You’re in trouble.”
2. The conclusion does not follow from the evidence. They are using inductive reasoning (poorly) to arrive at a conclusion not supported by the data. It’s also referred to in statistics as making claims, “outside the range of observation.” Think about water. Imagine we observe water while we heat it from 33 degrees Farenheit to 37 degrees Fahrenheit. Water was a liquid. Can we then conclude that if we continue to heat the water to, say, 300 degrees Fahrenheit the water will still be a liquid? Obviously not. However, without observing water boil assuming that water remains a liquid would be what a researcher might hypothesize, but that would be pure conjecture and obviously wrong.
3. And The Kicker – It MAY assign the wrong DIMENSION to the problem (of course this is my opinion here but it’s just as likely to be true as anything else)
“For every 1 percentage point below the mean regional unemployment rate, kids in affected areas had a 9 percent higher chance of using marijuana, a 7 percent higher chance of smoking tobacco, and a 6 percent higher chance of drinking when they were teenagers. Also higher: Gang affiliation (9 percent), petty theft (6 percent), major theft (11 percent), and the chance of getting arrested (17 percent). More serious problems -- like gun violence, assault, destroying property, and abusing hard drugs -- were not affected by higher unemployment rates.”
STOP. DO NOT PASS TO ANY CONCLUSION. DO NOT COLLECT TENURE FOR PUBLISHING NONSENSE.
“1 percentage point below the regional average.” So that means areas that were the bottom of the region? Aren’t those areas always likely to have the problems associated with high unemployment? Aren’t those also going to have the problem of a lower level of educational attainment among the populace? Higher dropout rates, etc. Greater drug use among parents? If you’re born in the inner city in the US, it’s not about when, it’s about WHERE. The public policy implication is not based on birth cohort, it’s based on whether you are more likely to be born into poverty! Since people more likely to be in poverty or in high unemployment areas tend to more vulnerable economically, maybe those areas get hit even harder than the rest of the areas.
Did they correlate high unemployment at other times? Is the leading variable actually unemployment? Not sure. Think about places which experienced a temporary increase in unemployment and then improved vs. certain US cities that have never recovered from factories closing, etc.
So the best part? Here comes insult to the children and the guilt trip for the working parents.
The insult and guilt trip come from the journalist and other studies looped together in a haphazard and simplistic manner suggesting the writer used Google only as her primary reference tool. Therefore, it’s pure bs editorializing. For example, she assumes the children born during those times were more likely to be unwanted pregnancies! How dare she. Speculating that a child is unplanned (and therefore unwanted – which on its own is a huge leap) is a severe and horrible thing to do. So children born in 2009, during a recession, during a dip in births, should assume that they were unplanned?
The guilt trip then assumes that parents forced into the workplace aren’t available “during children’s early years” is a contradiction (these were areas of HIGHER unemployment – so that makes no sense). The lack of affordable childcare in this country is a social issue (and a major factor in the US’s ability to compete in a global marketplace, but I digress) that does not support the findings of this study. The author places a judgment that children from working parents do not fare as well as other children. Not so. In fact, data show that children in quality childcare fare no different form children who were cared for at home and were more school ready to boot.
What a load of horse manure…
Posted by Statgirl Tells The Truth at 11:56 AM