Monday, November 30, 2015

Extremophiles and auto-immune disease

The extremophiles are at home in near boiling water and the deep crust of the earth. Every ecological niche is colonized by life; and life forms everywhere work to change their ecology to suit themselves.

It is the inevitable logic of natural selection in action.

So then, why should the hot tissues of auto-immune disease be any different? How could there not be life forms evolved to that extreme environment? Life forms that might facilitate it, to defeat their enemies and extend their preferred environment. An ecology that, once created, will host competitors, some liking it hotter, some coder. An ecology with successor species, like any forest.

It seems inevitable.

Hockey skill videos (reference post)

Adult ice hockey is a new hobby. Unsurprisingly there are lots of adult no-checking leagues in the Twin Cities, JMS Hockey is the one I went with. Actually, I should say “we” went with. My #1 son is 18 now, and he is the first identified special needs adult hockey player in our league. We play on the same line — he plays at the top range of the “lower level” games and I’m at the bottom. After more than 10 years of managing his hockey teams and even becoming a level 1 coach (no skills required) I finally got off the bench.

Despite growing up in Montreal when Les Habs ruled hockey I’m a lousy hockey player. Fortunately I’m not a bad skater, though lately somewhat knee impaired. So I’m trying to pick up some stick handling skills.

This blog post is where I’m going to put the video links and references I like — largely from [1] and a UK(!) Hockey Tutorial I’ll update it over the season (assuming I last).

Basic snapshot and wrist shot (because this is so embarrassing right now)

Wrist shot

Most basic shot, what I’m learning now. When learning face the puck.

Snap Shot

Quick shot taken when skating to net, face the net, puck to the side.


[1] There are many books and videos on mountain biking skills. Hockey players don’t generally do that kind of thing.

Wednesday, November 25, 2015

Car stereo and price hiding: Buy from Crutchfield but hire an installer.

Our car stereo died a few weeks ago. Wire cutter liked a model that seemed to fit our needs, the Pioneer DEH-X6800BT (Owners manual, I save these on my iPhone) was $112 at both Best Buy and Crutchfield. I didn’t want to do my own installation, so I bought it at Best Buy.

When it came time to do the installation Best Buy told me I needed to spend $70 for a kit — which they didn’t have. It turns out the same kit is bundled free of charge with the Crutchfield device. I can see why Best Buy exploits information asymmetry to hide the true cost of their services, but I think Crutchfield could market their price advantage a bit more. The Best Buy installers (Geek Squad) tell me they’ll do the install at the same price if I bring in the parts.

So now you know. 

(I’ve ordered the unit from Crutchfield, when it comes I’ll decide if I want to try the install myself. In their favor Best Buy is good at accepting returns, in this case an unopened return.)

Update 12/14/2015

After I studied the Crutchfield installation directions little red lights started going off. Especially given vague descriptions of tools I needed (not mentioned previously) and after I took a look at the wiring harness. So I decided to hire an expert.

Best Buy installers require payment in advance and they were booked out quite far. I found a local place with a single (very good) Yelp rating; same cost, quicker appointment, no advance payment and I felt more confident in them. The install took 45 minutes. I’m very glad I didn’t attempt it myself; the uneasy feeling I got from the Crutchfield online directions was well justified. I wonder if Crutchfield more-or-less expects people to pay someone to do the install.

My installer had no qualms about using the Crutchfield gear, he didn’t need anything else.

So after trial and error I’d recommend: 

  1. Find a local installer and confirm they’ll work with the gear you bring.
  2. Look at Wirecutter recommendations. Emily wanted CD so narrowed it down. 
  3. Order from Crutchfield. They bundle installation gear so was $70 less than Best Buy. 
  4. Pay expert to install. Locally I’d recommend this place.
The car stereo seems fine. I won’t do a full review (no time!), but only negatives were a default color change pattern only a teenage boy would appreciate and finding that the Pandora features don’t work with iPhone if you choose Bluetooth connectivity. The former is fixable in settings; the manual doesn’t mention this but you can easily set the display a neutral white. The latter wasn’t a big deal.
There’s a huge volume shift from Bluetooth to FM Radio; when I switched sources I almost blew my stock speakers out. 
Update 12/16/2015
  • The bluetooth connection is messing up Siri, it’s way less reliable. I know of a similar problem with a different BT user. I’m probably going to disable the BT connection and connect via USB.
  • The faceplate won’t come off, it’s stuck on the left side. My installer class that’s a problem with the (free) Crutchfield cage.
  • The UI is pretty awkward, but I’ll figure it out.
  • My Amazon review, including volume issues. I’m going to switch to using the USB/power cable and see if the volume/Siri problems are better.

See also

The impossible machine

The immune system is an impossible, incomprehensible, machine. It is not even a ’system’ — evolution is not so modular. We name it as a thing so we can model it, but it is not made by a mind. It is made by evolution, so it is bizarre and emergent. Like a machine that pumps water and makes potato chips on the downstroke.

The thing we name, which is in truth not a bounded thing, allows us to exist, briefly, in a seething sea of self-organizing energy. Presumably its antecedents emerged with the bounded sack of water we call a cell. It has grown in complexity since then, a complexity that often resembles the nervous system. It is, after all, a processing machine. Brains must tell lies from truth, the immune system must distinguish friend from enemy from frenemy. It must often attack the non-self, except when the non-self is a fetus. It should not attack the parts of the self, except when those parts are broken or rogue. It ages as the body ages, but even as it grows frail the body grows more rogue.

Sometimes the non-self is a frenemy, at least for the moment. We are walking biomes, ecosystems in motion. Billions of microbes live within us, often helpful, sometimes the enemy of our enemy. Except when they turn on us. The immune system must manage this, even as the enemies and frenemies adopt the face of the self.

We created the idea of the immune system, and we created the idea of diseases of the immune system — though the boundary between disease and individual variation is not sharp. Some immune systems are poor at stopping some enemies — we usually die then. At other times the immune system confuses self and non-self, and it turns on the organism. We call this an auto-immune disease.

We can do very little about auto-immune diseases. System Lupus Erythematosis is a classic of this genre, our treatment has changed very little in thirty years. We have some newer treatments for diseases like rheumatoid and psoriatic arthritis, but our treatments don’t correct the error of the immune system, they merely induce selective immune dysfunction to slow the progress of disease. We know so little. We aren’t even quite sure that there isn’t some bizarre infection lurking in the tissues of affected people; maybe sometimes the immune system has the right idea but the wrong execution.

Auto-immune diseases are common. We used to think osteoarthritis was a disease of aging tissues, of wear and tear. Now we think this name we made contains multitudes, some related to aging, others to an attack of self on self (“erosive inflammatory OA”). We have no truly effective treatments for these conditions. We don’t even know if sleep and exercise are a good idea — what strengthens healing also strengthens the enemy within. The war on joints and tendons wears on the body, inducing metabolic syndromes and accelerating aging.

If we could reverse auto-immunity, if we could re-induce tolerance, we might be able to manage organ transplants and even stop the enemy within. Inducing tolerance is now an active research area — at least in rats. We have a very long way to go. I hope the next 30 years improves on the last 30, but we have had many false starts.

Related RSS feeds (for Feedbin, Feedly, etc).

See also

Sunday, November 22, 2015

Resurrect anthropology

One of my big disappointments of past 15 years is absence of falsifiable models of factors that create and sustain Daesh and kin. 

Anthropology died in the 80s. We need to drag it out of the grave and apply high voltage with a serum of empirical economics, political science and geography. 

Or we can keep floundering.

Thursday, November 19, 2015

Scott Moses' FP Notebook is an astounding, and free, medical reference

Want to double check your doctor’s approach to your cholesterol problem?

Try this Family Practice Notebook (FPN) entry …


… VII. Management: Less than two Cardiac Risk Factors

Cholesterol Management
Goal if LDL Cholesterol below 160 mg/dl (ideally <130)
Low Fat Diet if LDL Cholesterol over 160 mg/dl
Anti-hyperlipidemic if LDL Cholesterol over 190 mg/dl

Desirable lipids: Repeat Lipid panel in 5 years
Borderline lipids: Repeat lipid panel in 1 year
Elevated lipids: Repeat lipid panel in 3-6 months…

Or take a look at the cardiovascular medicine book. (In FPN-speak a “book” is a collection of topics, they are accessible from the top right menu.) Ok, I admit there’s a lot of implicit knowledge in those terse phrases. FP Notebook isn’t aimed at consumers, it’s written for family physicians, though it also works for internists and pediatricians. If you’re not a physician you may have to take my word for it — this is an efficient high quality information source for “up to date” (more on that below) evidence-based US medical practice. It’s also a one man show of epic scale (emphases mine)…

About FP Notebook

… This site is derived from a peripheral brain collection of medical notes and is divided over 5700 topics within over 600 chapters and 31 subspecialty books. Information is gleaned from reputable sources, referenced where possible, taken from lectures and workshops, peer reviewed articles and bulletins, and key texts.

… This site is personally funded by the site author, Scott Moses, MD. Additional funding is obtained via advertising support; all paid advertisements are clearly delineated as such. Our advertising related privacy policy may be reviewed at here

Please let us know if you find any advertising to be distasteful or inappropriate, or which you find dilutes the value or integrity of this web site. Absolutely no content on the site is influenced or authored by advertisers. Content is solely per the discretion of the site author….

… These medical notes began as a few scattered pearls of text stored with the Notetaker application of the HP Palmtop 200LX. Since 1995, notes from conferences, articles, textbooks and colleagues have accumulated to its current state.

As the collection of text grew, so did the complexity of its organization, and a program known as Brains was developed as a stand alone application for notetaking and for compilation of the website.

Brains is able to import raw text from the database, process the outlines, titles and synonyms, as well as images and links.

In its third iteration, Brains is written in C# with a SQL Server database and can output each of the site versions of html in 1-2 hours. This latest version also allows topics to be linked to the UMLS metathesaurus codes, and to be viewed on handheld devices.

… Most images on the site I have created myself. Many of the anatomy images were created using the 3d models from Zyogote. Others use a combination of Adobe Illustrator, Photoshop and Poser software.

When images are used from other sources, they are clearly cited. These include the NIH Virtual Human project and MedPix

… Gray’s Anatomy 20th edition from 1918 (Lewis) is in the public domain following expiration of its patent. Scanned images of the plates are available online at both and Yahoo.

In 2012, I started an anatomy series using the Zygote 3D models. I will provide 2 sizes of these images: 800x600 and 4096x3112 (poster size). I have maintained a margin on poster size images, such that they can be printed at 8x10,16x20, and 20x30 inches without losing content when cropping. Costco prints a 20x30 image for about $8.

You may use these images freely for printing (e.g. classroom poster) as well as presentations (e.g. Powerpoint) as long as they are not re-sold or modified…The author of the Family Practice Notebook, is Scott Moses, MD, a board-certified Family Physician practicing in Lino Lakes, Minnesota.

I think this is his hobby. Some people build cabinets, Dr Moses builds textbooks.

There’s really nothing like this any more - FPNotebook is a relic of a bygone world. Twenty years ago, when he started, there were quite a few free online medical references. I did a small one myself, many better ones were published by medical schools and hospitals. For a while some medical textbooks were available online for low costs, often bundled into services like MD Consult. That’s all gone now; Wolter-Kluwers Up To Date has replace many textbooks and online resources - at $500/year for an individual subscription. Once free references, like American Family Physician, are now buried behind paywalls (thereby raising a bit of revenue but sacrificing much good will).

Even if you have a subscription to Up To Date, perhaps through your institution, I suspect you’d find much to like in FPN. For example, take a look at the updates page including the May 2015 FAST exam update: “Test Sensitivity may be as low as 22% for abdominal free fluid in blunt Trauma”. Want to keep abreast of meaningful changes to American medical care? Forget those expensive newsletters, just subscribe to the (new) updates feed: What a painless way to keep up.

Ok, you get the idea. This is an awesome resource — but how reliable can it possibly be? As Dr Moses notes, it’s entirely dependent on his vigilance and email feedback — nobody has volunteered to do peer review. I can give a partial response — because 10 weeks ago I started doing serious studying for my Family Medicine board exam. I last did family practice in 1994 and I don’t have an institutional subscription to Up To Date, so I used FP Notebook as a supplement to traditional texts, American Family Physician, and examination critiques. Over those weeks I covered a lot of family medicine in FPN; the only error I found was a minor misplaced section heading (an obvious copy/paste error that wouldn’t confuse any physician). I (and many others) found more mistakes in the closely edited ABFM exam critiques. There must be mistakes in 5.700 topics — but that’s an impressive record. I wouldn’t use FPN as my only guide to patient care, but I can testify to its excellence as a study and memory aid — and as a guide to what’s new and important.

FP Notebook is a dinosaur — straight out of Jurassic World, stomping over the puny mammals of the  modern web.

Try it, you’ll like it. (And say thanks — perhaps suggesting Dr Moses add a donation button. If nothing else, a charity donation button?)

Randall Munroe introduces world language and Google Translate training program using charming New Yorker article

XKCD’s Randall Munroe, the notorious interstellar sAI, has published a simplified vocabulary explanation of Special and General Relativity in the New Yorker.

This work is presumably taken from his almost released new book, Thing Explainer ($15 Amazon pre-order). The essay is entertaining and educational; it also promotes his new book and shows he is a smart pants man.

But that’s not the real reason he’s written this. Obviously his true agenda is to create an English dialect of a universal human language with a simplified vocabulary and grammar that is ideally suited to machine translation and, eventually, colloquial conversations with terrestrial AIs (contra the Wolfram Language for AI conversation, see also Marain. Siri-speak 2015 is a crude version of this.)

Let’s see how well his first version works, using the nsAI Google Translate to do round trip translations of a sample paragraph from the original muEnglish to another language and then back again. We’ll start with French, a language related to that of England’s 11th century conquerors, then we’ll do Chinese. I know from past experiments that round-trip translations from English to Chinese and back typically produce incomprehensible gibberish:

Munroe original (muEnglish)

The first idea is called the special idea, because it covers only a few special parts of space and time. The other one—the big idea—covers all the stuff that is left out by the special idea. The big idea is a lot harder to understand than the special one. People who are good at numbers can use the special idea to answer questions pretty easily, but you have to know a lot about numbers to do anything with the big idea. To understand the big idea—the hard one—it helps to understand the special idea first.

French version

La première idée est appelé l'idée particulière, car elle ne couvre que quelques pièces spéciales de l'espace et du temps. Celui-la grande idée-couvre l'autre tous les trucs qui est laissé par l'idée particulière. La grande idée est beaucoup plus difficile à comprendre que le spécial. Les gens qui sont bons à numéros peuvent utiliser l'idée spéciale pour répondre à des questions assez facilement, mais vous devez savoir beaucoup sur les numéros de faire quelque chose avec la grande idée. Pour comprendre la grande idée-le dur-elle aide à comprendre l'idée première spéciale.

French to English

The first idea is called the particular idea because it covers only a few special pieces of space and time. This great idea covers the other all the stuff that is left by the particular idea. The big idea is much harder to understand than the special. People who are good at numbers can use special idea to answer questions fairly easily, but you should know a lot about the numbers to do something with the big idea. To understand the great idea - hard - it helps to understand the first special idea.



Chinese to English

The first idea is the idea of so-called special because the space and time it covers only a few special parts. Another big idea, covering all rest of the stuff from the special idea. Big idea is a lot more difficult to understand than the special one. People who are good at numbers you can use special idea is very easy question to answer, but you know a lot about what the figures do big ideas. To understand the big idea, hard and it helps to understand the idea of a special.

Munroe English (muEnglish) works rather well between French and English. If you’re interested in learning French, you might enjoy reading a future French version of Thing Explainer or simply run the English version through Google Translate (and use speech recognition for verbal work).

The Chinese round-trip experiment almost works, but falls apart grammatically. For example, “you can use special idea is very easy question to answer, but you know a lot about what the figures do big ideas” is missing things like “need” and “to” and a few pronouns. There’s also an unfortunate “numbers” to “figures” word substitution. Given that Munroe is a far more advanced AI than Google this essay will be used to enhance Google’s Chinese translation model (which desperately needs work).

I’m optimistic about this new language and happy that the Munroe is now taking a more active hand in guiding human development. Zorgon knows we need the help.

Update 11/19/2015: There’s a flaw in my logic.

Alas, I didn’t think this through. There’s a reason speech recognition and natural language processing work better with longer, more technical words. It’s because short English words are often homonyms; they have multiple meanings and so can only be understood in context [1]. Big, for example, can refer to size or importance. In order to get under 1000 words Munroe uses many context tricks, including colloquialisms like “good at numbers” (meaning “good at mathematics”). His 1000 word “simple” vocabulary just pushes the meaning problem from words into context and grammar — a much harder challenge for translation than mere vocabulary.

So this essay might be a Google Translate training tool — but it’s no surprise it doesn’t serve the round-trip to Chinese. It is a hard translation challenge, not an easy one.

[1] Scientology’s L Ron Hubbard had a deep loathing for words with multiple or unclear meanings, presumably including homonyms. He banned them from Scientology grade school education. Ironically this is hard to Google because so many people confuse “ad hominem attack” with homonym.

Saturday, November 07, 2015

Mountain biking - crash and ride safety tips from Bicycling.

Bicycling Magazine has a surprisingly strong article aimed especially at mountain biker riding solo:

7 Things EMTs Wish You Knew about Bike Crashes (My instapaper link)

You always need to take an impact to the head seriously.” … call 911 if you or another rider has:
• ... a cracked helmet. That means you’ve hit your head hard.• ... a headache. Not just sore from the initial impact, but you have a headache that isn’t abating or is worsening.• ... lost consciousness. If you pass out, you need to get checked out.• ... confusion. If you don’t know who the president is or why you’re sitting on the side of the road, you need to get checked out.• ... vision changes. If the world doesn’t appear clear and normal, you need medical assistance.

Take a Deep Breath: Difficulty breathing is always an emergency situation. “Too often people crash and think they’ve cracked a rib, but figure ‘Why go to the hospital? They can’t do anything about it,’” says Martin. “But you need to go because those cracked ribs can have sharp edges and if it’s an unstable fracture and it shifts, you can puncture a lung.” If it hurts to take a deep breath, get to the ER.

Give Yourself a Gut Check: There’s a lot of vulnerable soft tissue and plenty of vital organs in your belly that can be damaged by impact with a handlebar. Take your hands and palpate your abdominal area. If you have an area that is more tender than others, you could have internal damage. If your belly becomes distended or firm, that’s a sign that you could have internal bleeding and need medical assistance stat.

Stop the Bleeding: Unless you’re a trained professional, forget what you’ve seen in the movies about fashioning a tourniquet around a limb to stop the bleeding. You risk doing more damage than good. The best way to deal with bleeding is basic first aid—direct pressure (preferably with something clean) on the wound. Keep it there till help arrives.

Be Smart About Your Spine: Neck and back injuries are scary. You can generally tell if you’re okay by checking your fingers and toes. Obviously, you want to be able to feel your fingers and toes, but if you have any numbness and/or tingling, that’s not good. You could have spinal injury. Also try slowly turning your head 45 degrees to the left and right. If you feel discomfort, stop. That’s also a sign of spinal injury. Get to the ER.

Make Your Personal Info Accessible: Whether you use Road ID, dog tags, or place ICE ('in case of emergency') information in your cell phone, having your personal information available for emergency workers can definitely save your life, says Martin. “We need to know your medications and your allergies," he says. "There are a lot of medications we can’t give you if you’re allergic to them… and we won’t give them if we don’t know.” New iPhones come with a Health app that provides a place for you to fill in all your medical information. Emergency personnel can access this information without unlocking your phone. “We know to look for it if you’re out there by yourself, unconscious, after a crash,” says Martin.

Leave a Note, or a Text: Riding alone? Take two seconds to leave a note or shoot a text to a loved one or buddy. “We’re all guilty of this,” says Martin. “We go out for a quick ride and nobody knows where we’re going. Even if you’re just 10 miles away, you might as well be 100 miles away if no one knows where you are.” The more remote of a place you ride, the more important this is.

It’s always safest to have a ride partner, but next best is to ride on well marked and trafficked trails. Riding on lesser traveled wilderness trails kicks the risk up several notches, just as with wilderness hiking or scrambling. Note the unstated implication of these recommendations is that you have a working cell phone and can call for help or advice after injury.

The iPhone Medical ID locked device access feature is obscure. You have to know to swipe to unlock then to tap the Emergency button then to look below the call keyboard and tap “Medical ID”. I hope EMTs are trained to do this. Apple forgot to enable Siri access, “whose phone is this” works on a locked phone, but “show me Medical ID” does not. I enabled Medical ID, but my phone’s lock screen has my contact info, more importantly, my wife’s cell number as Emergency contact. That was easy to do — I filled it out in iMessage, then took a screen shot, then made the screenshot my lock screen background. I need to check that I’ve set it up for my kids.

Sharing location on an iPhone using iMessage is easy, but also a bit obscure. You need to start with an existing message thread then tap the wee “i” icon top right. Just text that at start of a ride somewhere, update if you wish. I have Find Friend enabled, so if that works my family can track me (but they tend not to think of it).

Self-assessment with a head injury is tricky. I’ve had a concussion (inline skating actually), and it doesn’t help one’s judgment. If you whack your head out in the trail you should probably call in to a friend and have them check your thought processes. You may need to ride out before you can do anything more formal (assuming you can ride!).

Good stuff.

Monday, November 02, 2015

Trump explained: Non-college white Americans now have higher middle-aged death rates than black Americans

From today’s NYT Health section:

Death Rates Rising for Middle-Aged White Americans. Gina Kolata Nov 2, 2015

… middle-aged white Americans. Unlike every other age group, unlike every other racial and ethnic group, unlike their counterparts in other rich countries, death rates in this group have been rising, not falling…

… two Princeton economists, Angus Deaton… and Anne Case. Analyzing health and mortality data from the Centers for Disease Control and Prevention and from other sources, they concluded that rising annual death rates among this group are being driven … by an epidemic of suicides and afflictions stemming from substance abuse: alcoholic liver disease and overdoses of heroin and prescription opioids…

… the declining health and fortunes of poorly educated American whites. In middle age, they are dying at such a high rate that they are increasing the death rate for the entire group of middle-aged white Americans…

… The mortality rate for whites 45 to 54 years old with no more than a high school education increased by 134 deaths per 100,000 people from 1999 to 2014.

The article falls apart a bit here. What we want to know is how the absolute death rate for non-college middle-aged white Americans in 2013 and in 1999.  We want to know how the Long Stagnation has changed vulnerable Americans, but Kolata’s article mixes all white Americans with the no-college cohort.

Fortunately the PNAS article PDF is freely available, but unfortunately it explains Kolata’s problem — the data we want seems to be buried in an unlabeled parenthesis in Table 1. From that I think I can reconstruct the key information: [1]. 

YearWhite no collegeBlack (all)White some collegeWhite BA+White All
1999 601 797 291 235 381
2013 736 582 288 178 415

For the no-college White American 1999 was a pretty good year; probably the best ever. That was the era of NASCAR America and the candidacy of GWB, champion of the “regular” white guy. Employment demand was high and wages were rising. Yes, as a white guy without any college you had a shorter lifespan than the minority of white (Americans) with a college degree, but at least black Americans were even worse off. It’s always comforting to have someone to look down on.

After 16 years of the Great Stagnation though, things are different. Suicide and substance abuse have pushed no-college white mortality to the level of 1999 black Americans, yet during the same period black American middle-aged mortality has fallen substantially. White no-college Americans are now at the bottom of the heap [1].

This is why we have the inchoate white rage that thunders through the GOP. This is why we have Donald Trump.

A large and culturally powerful part of America is in crisis. A cohort with lots of guns and a history of violence. Maybe we should pay attention. Trump is a signal.

- fn - 

[1] There was no breakdown of black death rates by education; a 2012 census report said 29% of whites and 18% of blacks had a BA or higher. Since 80%+ of black Americans have no BA it’s likely no-college whites now have higher middle-aged mortality than no-college blacks.

See also

Update 11/4/2015

There’s been considerable coverage of this story, but it’s been disappointing. Both DeLong and Krugman missed the college vs. no-college white middle-age cohort, and I think that’s the important story. There’s also been some discussion of anger as a defining trait of the GOP base, but no connection to the extreme distress of their core voter.

I’ve seen speculation that this is all about narcotic overuse. I find that very hard to believe, but I admit the use of narcotics for pain relief in America has exceeded my expectations. I remember in the 90s when “pain is the new vital sign” and family docs were berated for inadequate use of narcotics. I guess my peers responded well to that feedback.

It has occurred to me that there’s a potential bias we’re missing. Over the past 40 years colleges have gone from predominantly male to predominantly female. The big story here is increasing mortality in the no-college white cohort. But if there’s been a gender shift in that cohort, say from 55% female in 1999 to 45% female in 2013, that will make the no-college numbers even more dramatic. Since mortality has increased even when college grads are included this isn’t the entire story, but it will make the no-college effect more dramatic.