Month: July 2013

San Jose State University has put its massive open online course (MOOC) project on hold because more than half the students in the pilot program failed their online courses, with pass rates for some classes as low as 20 percent.  George Anders suggests that Silicon Valley philosophies don’t translate well into higher education:

Will Oremus wonders what will come next:

It’s a sure bet that somehow, at some point, online instruction will indeed reshape higher education, if perhaps in more modest ways than its most ardent backers assume. Missteps are part of the process. Still, this is not the first heavily hyped online-learning venture to make headlines for going dramatically awry. The question is, what university will be eager to offer up its students as the next lab rats in what amounts to a massive pedagogical R&D program by for-profit Silicon Valley startups?

Douthat notes that fewer and fewer Americans are marrying fellow churchgoers or neighbors. More and more are meeting online:

[T]he data on unions formed online looks pretty encouraging, and it’s possible that the internet is helping to compensate for the eclipse of other forms of community, rather than contributing directly to those other forms’ eclipse.

But it seems fair to assume that there are still a lot of people who would prefer to meet their future spouse the old fashioned way — through initial flesh-and-blood encounters embedded in a larger pre-existing social network. If that’s your preference, the university campus is one of the few flesh-and-blood arenas that seems to be holding its own as a place to form lasting attachments.

Millman expands on the possible social consequences of widespread online dating:

The hidden costs of internet dating aren’t some much to the romantic “market” as to the rest of society. How many young people don’t go to church or synagogue because there’s no reason to go there to meet marital prospects? And once that dynamic starts, it inevitably accelerates, as the residual group who does show up is increasingly untenable romantically (because they are there for other reasons – or can’t find an “adequate” mate digitally). Then there’s the problem that if you meet someone digitally, you’re probably meeting someone with a distinct social circle – as opposed to someone from within your own social world. That’s good if the relationship doesn’t work out – less risk of collateral damage. But it creates complications if the relationship results in marriage, as now you have two circles to navigate that don’t overlap well. That happens plenty if you meet in meatspace, of course, but it’s much more certain to happen if you meet digitally than if you meet, say, at church.

Several readers have voiced skepticism over our coverage of this map and the accompanying piece from the NYT:

I would be interested to see that map compared to a map of “chances anyone in this county is an earner in the top fifth.” Some parts of the country simply have no or very few opportunities that earn at those levels, so it should be no surprise that you find few people “rise” to that level – with the inverse true for places like NYC, SF and Seattle. To study mobility, unobscured by regional differences in wealth, one would need to re-calibrate to the quintiles within each county, rather than comparing local samples to national averages.

Another reader:

Quite clearly the map shows (and the original paper makes clear) the importance of race, whether that be African-American or Native American. But this does not mean that somehow the Red States only care about taking care of their “white, GOP brethren,” as one reader implied. What does the data (pdf) from the paper say? Check out Table 5. Two of the best negative correlations with intergenerational mobility are the county’s rate of divorce and its share of single moms – two things that disproportionately affect African-Americans and Native Americans.

Utah, meanwhile, confounds your “two Americas” reading. Look at the blue area in Utah in the map you posted. This is because of Mormonism. Check out this blog post, for an interesting discussion on the similarities between Utah and Denmark, two places of high intergenerational mobility.

Another goes into much more detail on the previous points:

The graphs you’re presenting from the New York Times article are HIGHLY misleading.

First off, they look at social mobility without taking into account the regional differences in cost of living. From the chart: “The top 5th is equal to family income of more than $70,000 for the child by age 30, or more than $100,000 by age 45.” Anyone who’s anyone knows that living on $70,000 per year income on the Upper West Side of Manhattan is vastly different than living on $70,000 per year in the rural South.  In one place you are “poor” and the other you can live like a king.  I’ll gladly earn $65,000 and live in Tuscaloosa, Alabama vs. earning $75,000 and living on the Upper West Side of Manhattan. Yet, in one case I haven’t reached the top 5% of income earners while the other I have.  This is clearly not true in reality.

Without a cost of living adjustment, the data is essentially meaningless. Income tells you next to nothing. What you can buy with that income is what matters! And that varies greatly by region. The chart essentially tells you where the dollar is worth more and where it is worth less.

Secondly, look at the people presented in the article as the “poor” struggling to attain a middle-class lifestyle. Stacey Calvin is presented as a “poor” person struggling to enter the middle class in suburban Atlanta. She has three kids. Is a single mom. Works only part time. Yet, from the picture of her, has the latest in smartphone technology, a three-bedroom home, and remote-control cable television. She and her kids are wearing fashionable clothing. This is NOT the picture of someone mired in poverty.

The next example is Michael Novajovsky, who is an educated network engineer.  He is married, three kids, has a beautiful house (as far as one can tell from the picture), earns $27/hour for a temporary job. From his Facebook page, you can see he isn’t mired in poverty either.

Lastly, how about breaking the data down by zip code or neighborhood.  New York City is not homogenous.  It includes Harlem, the Upper West Side, etc.  Simply aggregating numbers of very different neighborhoods as one is highly misleading.  Especially compared to a place like Atlanta that is spread much further about geographically.

Another:

I happen to live in Atlanta and recognize the difficulty in getting from the south side of the metro Atlanta area to where jobs are located to the north side of the metro Atlanta area if you rely on public  transportation. However, I respectfully submit citing the NYT article in support of your contention “the South” (which I guess does not include Red State/low-tax Texas by your assessment of the NYT map) seems at times to be a different country is cherry picking items from the story. As evidenced by the map itself, the NYT reporter stated “Climbing the income ladder occurs less often in the Southeast and industrial Midwest, the data shows, with the odds notably low in Atlanta, Charlotte, Memphis, Raleigh, Indianapolis, Cincinnati and Columbus.” Would be interested in your thoughts on why the same problem being present in the industrial Midwest fits with your interpretation of the article.

I think the map can bear many different interpretations. I focused on the obvious fact that race is a factor, and that the band in the South correlates very closely to slave-ownership patterns in the past. There may well be other explanations for other regions – but the Southern aspect is what immediately struck me, and, I’d say, anyone who looked at it. Update from a reader:

I wish you would stop saying that race explains the differences in social mobility found in the recent study.  The authors, in their summary of their findings (pdf), explicitly state that the differences remain after controlling for race.

Unfiltered feedback on our Facebook page.

A reader adds anecdotal evidence to the research:

I’ve fought the weight battle my whole life.  Now I’m down 37 pounds and seem to be maintaining at that new level. One of the reasons is marijuana. I’ve dieted frequently over the last couple of decades but no matter what I did I couldn’t seem to push below 239. It was as if my body just said “Thus far, no further.” I was tired of losing the battle, and my doctor was warning that diabetes could be in my future, so I concluded that I needed to shake things up.  I began by swapping marijuana for alcohol.  I’m now at 223, no longer technically obese, though I could certainly lose a few more pounds.

What’s stranger, though, is how easily I’m maintaining at 223. In the past the regaining of weight began almost immediately post-diet. Yes, I’m being reasonably disciplined, but not very. I’m smoking a moderate amount of legal medical weed once a day, in the evening when the work is done and I know I won’t be driving – same rules I had for alcohol consumption.  So far so good.  Not only is my weight down but my pulse and blood pressure are those of a much younger, fitter man than I’ve ever been.  I almost feel as if I’m cheating.

No idea why it works.  But it seems to.

Update from a reader:

“No idea why it works”? I do. Not drinking is the easiest way to lose and keep weight off. I would argue it has nothing to do with he/she replaced it with. My four-pack turned into an eight-pack when I stopped drinking for a few months. I’m very fit and exercise regularly. Those last five pounds that I couldn’t lose before, gone. They creep back over time if I drink even though I’m not a big drinker, 1-2 drinks/per week. When people who see me with my shirt off ask how I did it I simply say I quit drinking. The most common reply is “oh, I could never do that”.

Another writes:

Your post on lower rates of obesity and diabetes in regular marijuana users doesn’t surprise me at all, especially since I have seen so many cases of cannabinoid hyperemesis syndrome (also known as the “anti-munchies”) in my patients over the last three years.  This condition is characterized by periodic cycles of abdominal pain and vomiting in frequent marijuana users, and many of them have had extensive and costly medical testing without any definitive results.

The only definitive relief comes from abstaining from cannabis, although there are several unusual features of the condition (patients state that their nausea and pain are relieved by hot showers or baths) that help in diagnosis.  The mechanism of this disorder may simply be due to the brain “pushing back” against a constant flood of cannabinoids in the blood, similarly to the way chronic alcoholics or opiate addicts will experience tolerance and then withdrawal to those particular drugs.  Not every regular marijuana user develops this condition, but perhaps increasing rates of the condition are contributing to an overall decrease in obesity rates in this population.

A doctor is skeptical of the research we posted:

In general, I get frustrated with general reporting of scientific data.  It’s unfortunate that most people who write these articles 1) don’t understand science or research and 2) don’t understand statistics. Unfortunately, many scientists themselves don’t understand statistics, since the vast minority actually have masters degrees in clinical epidemiology.

So, what we get is post that links to a recent epidemiological study and claims that the data supporting cannabis and weight loss “are clear.” In fact, the study you referenced tells us nothing of the sort (despite the author’s claims in the discussion).  The cohort sampled showed a direct overlap between patients who used tobacco and those who used marijuana.  I don’t care if the epidemiologists claim to be able to use fancy statistical methods to control for tobacco use, the fact is that they can’t actually do that.  All this study tells us is that there MAY be an association between cannabis and weight loss, but frankly drawing that conclusion is a stretch.