Monday, July 30, 2012

Research Integrity and Politics: The Regnerus Affair

Because many engineers do scientific research and publish in peer-reviewed journals, the matter of research integrity should be a concern of all engineers.  An acquaintance of mine, University of Texas sociology professor Mark Regnerus, has recently found himself in the center of a tornadic controversy over a paper he published last month in the Journal of Social Science Research.

I am not an unbiased observer of this situation.  I met Prof. Regnerus several years ago at a dinner, and he impressed me as a pleasant, sincere Christian (he is a Catholic convert) whose presence in the field of sociology was a welcome one, because sociologists in general tend to be leery of personal commitments to organized religion.  Regnerus is interested in the way sexuality influences and is influenced by social behavior, as evidenced by his earlier Oxford University Press book Forbidden Fruit, an investigation of teenage sexual behavior and attitudes.  But with his latest paper, Regnerus stepped on a political third rail.

The paper describes an extensive research project into the question of whether gay parenting affects the lives of children in measurable ways.  The conventional sociological wisdom, represented by a fairly small number of research papers, says that there is essentially no negative effect of being raised by two mommies or two daddies, as opposed to the conventional mother and father.  This body of work is cited by every judicial decision in favor of things such as adoption by gay parents and the extension of marriage to gay couples.

Regnerus’s study, which he himself admits is not perfect, found otherwise.  There were significant negative consequences of being raised by parents who were gay, according to the study.  I am not going to address the controversial question of defining “gay” or how extensive the negative consequences were or how accurate and scientific the study was. 
Not being a sociologist, I am not qualified to pass judgment on these matters.  What I am qualified to judge is the way the peer-review process has been attacked and corrupted after Regnerus’s paper was published.

The idea behind peer review is that scientific publications should be judged by those most qualified to do so:  namely, other scientists in the same field.  That is exactly how Regnerus’s paper was judged.  As is common practice in some fields, Regnerus was allowed to suggest the names of some reviewers, and as is also common, he had worked with some (not all) in the distant past.  In specialized fields, this kind of thing is often unavoidable and does not mean that the reviews will necessarily be biased in the author’s favor.  (Sometimes it works the other way!)  In any case, the reviewers recommended publication and the paper was published.

Then the deluge began.  A journalist and self-described “minorities anti-defamation professional” whose pseudonym is Scott Rose wrote a letter to the University of Texas administration alleging that Regnerus’s paper falsified data.  This is the most serious professional charge that anyone can level against a scientist, comparable to a malpractice charge against a doctor.  The first wrongdoing (as I pointed out in a letter published in the Austin American-Statesman) was for UT Austin to act on such complaints from a person who was not in a competent professional position to make such assessments.  Scott Rose is not a sociologist.  Rose has since published the full “evidence” he plans to present to UT Austin, and it consists of two kinds of arguments.  One kind comprises disputes over methods and definitions that Regnerus used.  If Rose had been selected as a reviewer of Regnerus’s paper, these arguments might have played a role at that point.  But Rose, not being a qualified sociologist, has no professional standing to make them, and they must be assessed on their merits by other professional sociologists.  The other kind of argument consists of various ad hominem attacks on Regnerus’s funding sources, which include organizations such as the Witherspoon Institute that favor conservative causes.  While taking funding from organizations with a political agenda is certainly a possible source of bias, in the field of sociology it is hard to avoid.  Even the federal government has a political agenda, and one’s source of funding cannot be construed as prima facie evidence of research falsification.

Rose also cites the other outrage against the peer-review process:  a special audit report written by a member of the Journal of Social Science Research’s editorial board on the question of whether the peer-review process that led to publication was flawed.  The member, Darren Sherkat, found essentially nothing wrong with the peer-review process.  Instead, he took the opportunity in the audit to review the paper himself, and used terms (“bulls---“) that in my opinion have no place even in a conversation about another colleague’s work, let alone a report on the integrity of the review process.

I have not even mentioned the press coverage with derogatory headlines, the letter signed by over a hundred sociologists objecting to Regnerus’s conclusions, and the politically motivated letter-mobbing of the journal’s editor, James Wright, which pressured him to request the review audit.  Releasing a draft audit to the media, as Wright did, was clearly a craven attempt to deflect hostile politically motivated attacks from himself.  It showed no respect or regard for Regnerus, and probably did not even achieve its intended purpose.

In an opinion piece published in the Chronicle of Higher Education, sociologist Christian Smith takes Regnerus’s side and expresses better than I can, the point that the scientific integrity of the field of sociology is at stake here.  I will ask a question.  In the 1930s, many prominent scientists and engineers in Germany lost their reputations, their jobs, and some eventually their lives because of a non-scientific reason:  they happened to be Jews, or outspoken Christians, or simply opposed to some political aim of the government.  Everyone now agrees that this was a grievous violation of human rights, an early warning sign of the greater wrongs the German government would do in World War II.  While that situation differs from the one Regnerus finds himself in by degree, does it differ in kind from what Jewish scientists suffered in Germany in the 1930s?  Regnerus has reached scientific conclusions that oppose the prevailing political winds.  Though his punishment has come from activists rather than official government sources, it is no less politically motivated and no less unjust.  Smith thinks the integrity of the social-science research process is threatened by the “public smearing and vigilante media attacks” mounted against Regnerus.  If such attacks are successful, we have taken a long step away from scientific integrity and a long step toward the encouragement of a political atmosphere that is totalitarian in its effects.

Sources:  Among the many articles published on this controversy in the last few weeks, I have used the following.  The Austin American-Statesman published a description of UT’s inquiry on July 11 at  The Chronicle of Higher Education has published “The Regnerus Affair at UT Austin” by Peter Wood on July 15 at and “An Academic Auto-da-Fé” by Christian Smith on July 23 at

Monday, July 23, 2012

The Defibrillator That Fails May Be the One You Need

Everybody who has watched medical shows on TV has sooner or later witnessed a simulated attempt to start a heart going again with a defibrillator.  The doctor in charge tells everybody else to get out of the way—he places electrodes on the patient’s chest—then bang!—the body arches upward and, depending on what the dramatic needs of the moment are, either starts breathing again or gets covered up for the last time.  Used properly in real life, automated external defibrillators (AEDs, for short) can be lifesavers.

For a variety of reasons including circulatory problems and electrical shock, a person’s heart can go into an ineffective kind of twitching known as ventricular fibrillation, and blood basically ceases to flow.  This is called sudden cardiac arrest.  Invariably the person becomes unconscious and has no manually detectible pulse.  It used to be the case that unless properly equipped emergency workers arrived with an AED within four to six minutes of sudden cardiac arrest, it meant curtains.  Then it occurred to AED manufacturers to make their devices simple enough so even a sixth-grader could use one, as has been demonstrated in practice tests.  For the last fifteen years or so, easy-to-use AEDs have been showing up in public places such as airports, bus terminals, universities, and malls, and people have been rescued by quick-witted bystanders who grabbed an AED and used it in the right circumstances.  But as reported in IEEE Spectrum last spring, a disturbing number of AEDs out there fail to do their job, or would fail if called upon to work.

About 300,000 people in the U. S. alone die from sudden cardiac arrest each year.  Some  of these folks would not benefit from application of an AED, but many of them—possibly as many as 40,000 a year—could be saved by a properly applied working AED.  In Seattle, Washington, city authorities undertook to develop a registry of the location of every publicly accessible AED in town, and then promoted a “citizen defibrillation program” with advertisements, public information publications, and training.  The result is that as many as 45 percent of witnessed cardiac arrest cases (situations where the person is not alone) survive.  Contrast this to the U. S. average of 4 percent, or the even more dismal figure of 0.5 percent for Detroit.  The last thing anybody trying to use an AED expects is a little message on the machine’s display saying something like “BATTERY LOW” or “SELF-TEST ERROR 17.”  But it happens.

AEDs are medical devices, aren’t they?  And so they must go through the same rigorous Food and Drug Administration qualification and inspection tests as other medical devices, mustn’t they?  Well, not quite, it turns out.  AEDs are in a kind of legal gray area that allows manufacturers simply to say that their product is “substantially equivalent” to other AEDs, and then they can bypass the usual medical-equipment tests and qualifications.  So it’s up to the manufacturers to ensure that the batteries will stay charged and the unit will be operational even after years of total neglect, and perhaps environmentally harsh conditions of high and low temperatures and humidity in outdoor locations.

This would be a hard trial for any piece of electronics, but for a unit that someone’s life may eventually depend on, it’s doubly difficult.  And the Spectrum  report shows that an FDA investigation found over 90 percent of AED failures were not investigated sufficiently to identify the cause.  Most of these failures showed up during routine tests, but 750 of the reports of failure between 2005 and 2009 followed a death in which the AED was involved.  And at least one manufacturer maintained a “fix-on-fail” policy.  That is, when the same design problem began to show up in a number of calls for repair of AEDs, you would think the firm would act like most auto manufacturers do and issue a recall to all owners of that model device.  No—this outfit simply waited for the next failure to occur instead of notifying the owners of all potentially defective AEDs.

So what’s the answer?  Changing the law to make AEDs qualify through the same rigorous process as other medical devices is one alternative.  But the manufacturers claim, with some justification, that this will send prices (already in the $2000 per unit range) through the roof.  Ideally, an AED would be as cheap and reliable as a fire extinguisher so that people (for example, heart patients) could afford one at home.  But this isn’t going to happen if prices are north of two kilobucks a pop.

My libertarian streak makes me reluctant to say this, but the way we got fire extinguishers in every public building was by means of fire codes:  laws that compel building owners to have so many fire extinguishers for a given square footage of space.  And it’s the owners’ responsibility to make sure those extinguishers are operational too.  Maybe the only way to make sure AEDs work and are widely accessible is to pass similar codes requiring AEDs, at least in places where the demographics indicate it would be helpful.  Because people younger than 20 rarely go into ventricular fibrillation, for example, K-12 schools might not need more than one in a large building.  But rest homes, for example, could use more.

This issue strikes close to home for me, because as a 58-year-old male I’m in the prime demographic of those who might need an AED some day.  Both my grandfathers died of circulatory problems, and while I try to eat right and exercise, there’s only so much you can do.  Let’s hope the next time anyone you know needs a defibrillator, that one will be handy—and it will work, too.

Sources:  The article “A Shocking Truth” by Mark Harris appeared in the March 2012 issue of IEEE Spectrum, the general-interest publication of the Institute of Electrical and Electronics Engineers, on pp. 30-34 and 57-59.  I also consulted the Wikipedia article on defibrillators.

Monday, July 16, 2012

Hacking for Solidarity: Broadcast Engineers and the Polish Resistance in the 1980s

In 2008 it was my privilege to visit a Polish friend of mine at his home in Warsaw.  We had become first acquainted back in the 1980s when it was his privilege, as a university professor of electrical engineering, to leave what was then a Communist country for a one-year sabbatical accompanied by his wife and two sons at the University of Massachusetts Amherst, where I was also teaching.  Some people would have defected and just stayed in the U. S., but Andrezj (pronounced “Andrey”) loved his native Poland enough to return and work for a brighter future, which as it turned out came to pass in 1989, just before the fall of the old Soviet Union.  His sons are both thriving young professionals in Poland now, and he does not regret his decision to return. 

I was reminded of my friendship with Andrezj and his family when I read in a recent historical journal about a little-known aspect of the Solidarity movement in Poland.  Officially a trade union, Solidarity formed around Lech Walesa, a shipyard electrician who led the movement and became modern Poland’s first freely elected President in 1990.  During the 1980s, the Communist government of Poland imposed martial law and rigidly suppressed and controlled speech.  All media outlets, including radio and TV, were under government supervision.  As the Solidarity movement grew, however, a group of radio and television engineers joined with other technical types (mainly university professors) to do (what we would term today) hacking of the official radio and TV networks.  Pretty soon after that, TV-watching Poles began to see images of things like the words of the national anthem superimposed on the video feed of dull official programs.  Now and then, the audio of the TV channel would give way to music of the national anthem, a joke, or some popular song that had nothing to do with the official program.   At other times, Poles listening to their radios began to receive signals from “pirate” radio stations broadcasting information that the government did not want them to hear.

The content of the messages sent by what came to be called “Radio Solidarity” was not as significant as the mere fact that somebody, somewhere, was messing with the government’s system, and could get away with it.  These activities were not without risk.  While routine protests such as marches were usually punished merely by fines, the Polish government pursued the Radio Solidarity hackers more vigorously and put many into prison.  But as the movement grew, more volunteers arose to replace those who were arrested among technicians and scientists who were technically proficient enough to tap into the broadcasting network, or build and hide radio transmitters in houses of sympathetic citizens.  At one point, an underground flyer called for people to wear radio resistors pinned to their lapels in memory of two brutal police actions against Polish workers in 1970 and 1981.  Evidently, large numbers of Poles did so, and once the authorities caught on to this form of “resistance,” you could go to jail for it.

Opinions differ as to how effective Radio Solidarity was in aiding the movement toward democracy.  One industrial worker happened to be high on a crane one evening in Gdansk as a Radio Solidarity broadcaster asked his listeners to turn out their lights for a minute if they were listening.  According to the witness, he saw half the city go dark.  A precise assessment at this point is impossible, but the emotional encouragement that such activities brought to people who might otherwise have thought they were struggling alone against the system could have been invaluable.

The story of Radio Solidarity brings to mind the more recent social-media-powered actions of those in the Middle East who have attempted to free themselves from oppressive regimes.  Unlike the largely successful Solidarity movement in Poland, which not only brought about regime change in that country but may have contributed to the downfall of the Soviet Union itself, the record of recent opposition to governments in Middle Eastern regions is more mixed so far.  Perhaps it is fair to say that technical means of communication in a political movement is necessary these days, but not sufficient by itself to insure a smooth transition from oppression to freedom.  If citizens do not have a clear, united vision of what a democracy should be like and what goals they should strive for, all the communications facilities in the world are not going to make much difference.  Revolutions are always a last resort, and while some can turn out peacefully, the result is always unpredictable.

Ironically, the limited nature of network systems in the 1980s may actually have helped Radio Solidarity reach more people than hackers today can easily manage to do.  If the Polish national TV network back then resembled those of most other totalitarian countries, there was basically one channel and one program.  That meant if you hacked into the network feed, you had the entire country as an audience automatically.  Today, of course, with the multitudes of various communications media—TV, radio, the Internet, mobile apps, social media, texting, and so on and so forth—it is much harder to reach an audience that is not already primed to hunt for a particular website or participate in a particular meeting, and so it can be harder now to reach large numbers of people.  But this problem can be overcome, especially in times when young media-savvy people are eager to help out.

We can be grateful to the Radio Solidarity engineers and professors who did their part in overthrowing the oppressive Polish regime and catalyzing the downfall of the Soviet Union.  And we can hope their example can be learned by those today who are dealing with similar problems in their own countries.  It seems, though, that freedom of communication is not a magic bullet that keeps away the real bullets a hostile government can use against its citizens.  Syria is a current bad example of this.  Despite repeated internal and international calls for the Syrian government to relinquish its stranglehold, its rulers persist in shooting thousands of protesters.  I don’t know whether anyone in Syria reads this blog, but I would consider it an honor to be in trouble with the government of Syria, and hope that those fighting for true freedom in that country can use all the means at their disposal, including electronic ones, to find a good way out.

Sources:  The article “Dissident Visions through Technological Use:  Radio and Television Solidarity in Poland, 1982-1989” by Carmen Krol appeared in the Autumn 2011 issue of “Antenna,” a newsletter published by the Mercurians, a special-interest group of the Society for the History of Technology.

Monday, July 09, 2012

Higgs 1, Neutrinos 0

The world of high-energy physics breathed a collective sigh of relief last Wednesday when a team of researchers at CERN, the European physics lab, announced results from their Large Hadron Collider (LHC) that make it “99 percent certain” there is indeed a thing called the Higgs boson.  Further work is needed before the discovery makes it to the point that there’s only a one-in-a-million chance that it’s not true, which is allegedly the standard that physicists now maintain before a new particle is generally accepted.  But judging by the signs of exultation and emotion, such as the tears of Peter Higgs (the theorist who first predicted the particle in the 1960s), such acceptance is only a matter of time.

The mood at CERN contrasts with the emotions that a much less publicized announcement last February probably inspired.  Late last fall, some CERN researchers working with an Italian physics lab known (appropriately) as OPERA announced that they had clocked a subatomic particle called a neutrino (Italian for “little neutron”) as going just a hair faster than light.  I blogged on this last October 9 and remarked that if it turned out to be false, you wouldn’t see any headlines about it.  Well, I was wrong, but the headlines were a lot smaller than the ones about the discovery of the Higgs boson.

Turns out that the OPERA people had overlooked a bad connection in their fiber-cable system that messed up their timing by just enough to make it look like the neutrinos were going faster than light, which they probably weren’t.  It’s all rather fuzzy with overlapping error bars, but the problem was significant enough to account for their faster-than-light result, which they have now admitted may have been in error.  Several other similar experiments have not turned up any results indicating that neutrinos can exceed the universe’s speed limit either.

In these days of limited public resources (when were they ever unlimited?), what justification is there to spend many billions of dollars on things such as the Large Hadron Collider or the OPERA experiment?  Judging by the fact that the torch of leadership in high-energy physics has clearly now passed from the U. S. to Europe, the American public seems to have answered, “None.”  Some readers may remember the proposed Superconducting Supercollider planned for Texas, which was cancelled in mid-project in 1993 and never revived.  If it had been completed, it would have easily surpassed the LHC in energy capability, and so we might have been hearing about the Higgs boson discovery from Waxahachie rather than Geneva.

The fact is, no single nation can claim that its citizens were responsible for the discovery.  Thousands of Ph. D. physicists worked on the plans and experiments that made it possible, including hundreds from the U. S.  High-energy physics today is a thoroughly international enterprise, and to do anything significant experimentally means to become a member of a huge team that resembles a multinational corporation more than a small university-style laboratory.  This takes nothing away from the honors due to those who found the Higgs, but it does raise a question of responsibility.

CERN as an organization lent its name both to the Higgs-boson discovery and the announcement that neutrinos possibly traveled faster than light.  Some would say that, hey, this is the way science works, and further checking indeed revealed that the faster-than-light discovery was in error.  While I have no realistic doubt that the Higgs-boson discovery is legitimate, checking it may take anywhere from years to forever if you want to do it with another accelerator, because the LHC is the only show in town, that is, the world.  Of course there are many different experiments that can be done with the LHC to demonstrate the Higgs, so it’s not like we had only one instrument reading, for example.  But the point here is that CERN is becoming something of a monopoly in the high-energy-physics business.  And monopolies may try not to act like they’re monopolies, but it’s not always easy.

When scientists such as those at CERN are asked to justify their arcane projects, they generally reply that physics, like many other scientific pursuits, is ultimately a cultural activity, and has to be its own justification.  Some people paint beautiful paintings, others discover beautiful particles.  From the viewpoint of engineering, I have to agree, because it has been a long time since a discovery in the realm of high-energy physics has led to any major practical applications.  Probably the last time this happened was in the 1930s, when Lise Meitner and her colleagues discovered nuclear fission.  That led to the development of atomic weapons and nuclear energy, both of which had consequences that are still with us today.  Beyond that, however, I can’t think offhand of any consumer or industrial products that intrinsically rely on muons or neutrinos or quarks or any of the other members of the subatomic zoo discovered since then.

This doesn’t mean that practical applications will never come along.  But they appear to have slowed down, certainly, and so I have to agree that experimental high-energy physics is just like observational astronomy:  done for the beauty of the thing, and not for any practical uses of the knowledge that may result.

Still, a billion dollars is a lot of money, and I hope that CERN and organizations like it remember that they exist only by permission of a tolerant public.  Incidents like the mistaken announcement of faster-than-light neutrinos do nothing to enhance their reputation, and so I hope that further checking takes place the next time a surprising result like that shows up, and also that the physics worker bees pursue their goal of confirming the Higgs boson with one-in-a-million certainty.  That’s only several thousand dollars per chance, if you want to think of it that way.

Sources:  The announcement that errors were found in the faster-than-light OPERA experiment was carried by the online version of the journal Nature at  Numerous news outlets reported on the Higgs-boson discovery, and I used the report from National Geographic magazine at 

Monday, July 02, 2012

Now We Know: Stuxnet R US

Almost a year ago in this space, I wrote about a sophisticated new computer virus called Stuxnet which had apparently done considerable physical damage to almost a thousand uranium-enrichment centrifuges in Iran in 2010.  At the time, it wasn’t clear who designed Stuxnet, although guesses were that either Israel or the U. S. was responsible.

Well, on June 1 of this year, the New York Times published excerpts from an upcoming book that confirms those suspicions, and goes into a lot more details.  It turns out that Stuxnet was only one of several cyberattacks that originated with a project called “Olympic Games” that began during the presidency of George W. Bush, who encouraged his successor Obama to continue it.  Obama took Bush’s advice and persisted with the program even after the Stuxnet virus escaped “into the wild,” which is how the computer-security community learned about it a year ago.

There are two related ethical concerns that these latest revelations highlight.  One has to do with U. S. participation in cyberwarfare generally.  And the other has to do with the fact that someone in the current administration spilled so many beans about what we were doing.

Cyberattacks are following a well-trodden path down which earlier forms of militarily useful technology passed decades or even centuries ago:  telegraphy, radio, aviation, and nuclear weapons, to name a few.  The trend is from discovery to initial, usually rather amateurish, experimentation, and then to serious funding and adoption by all sides in a conflict.  With regard to cyberwarfare, we are now beyond the amateurish-experimentation phase and well into serious adoption by at least one side:  the U. S. and Israel, which turns out to have collaborated closely with the U. S. in the Stuxnet project for both technical and diplomatic reasons.  If history is any guide, we can now anticipate a cyberwarfare counterattack by one or more of our enemies sooner or later.

This is made especially likely because cyberattacks turn out to be pretty cost-efficient.  Software experts examining the Stuxnet virus at the time it was first found estimated that it was fairly cheap to develop, under a million dollars.  The latest revelations in the Times show that this may have been an underestimate, because the CIA went to the trouble of building a working model of part of Iran’s nuclear facility using the identical machines that were the target of the attack, in order to be sure it would work.  Still, it was cheap compared to a full-scale airstrike with cruise missiles, for example.

Cheapness cuts both ways.  The U. S. isn’t the only country with sharp computer whizzes willing to develop evil viruses to mess up critical infrastructure.  Stuxnet was highly specialized to do a specific kind of damage to only one facility, but virus-writers worldwide have highjacked its innards to do other malicious things since then.  It is not beyond the realm of possibility to imagine someone taking the basic Stuxnet format and designing a virus to, say, whack out an industrial controller commonly used to regulate the speed of steam turbines in power plants.  I’m not knowledgeable about the degree of sophistication of power-plant software or the tightness of their security measures, but I’m sure it varies from place to place, and while some on-the-ground collaboration was needed for the attack on Iran, that might not be necessary for some forms of virus attack.  The point here is that with its vast array of computer-dependent infrastructure, the U. S. is very vulnerable to just the kind of cyberattack we mounted against Iran.

Which brings up the second ethical concern:  did we have to go so public with all the details of what our responsibility was in Stuxnet?  Critical information about decryption technology used in World War II was kept in the dark for decades.  I would expect the kind of details we read in the Times to come to light some day, but less than two years after the attack?  Perhaps this is a deliberate ploy to warn counterattackers that yes, we can do this and you’d better watch out.  But because cyberattacks rely on lapsed vigilance and poor security measures (the Stuxnet actually got into the target network through a carelessly used flash drive someone carried into the secure facility), it seems like telling our enemies all the details of our attacks and responsibility for them, will just make them all the more cautious and less likely to fall for such things in the future.  In other words, if you’re going to fight a war with secret stuff, blowing the secret doesn’t seem like a good idea.

At the risk of sounding excessively political, one could speculate that the publicity about Stuxnet was another attempt to show the present administration in a “tough-guy” mode, consistent with recent revelations about how the President himself personally authorizes every drone attack on targets that have included a U. S. citizen in at least one instance.  These drone attacks have drawn criticism from the President’s own party, notably former President Jimmy Carter.  Like cyberattacks, drone strikes are, in the short term, a “no-risk” mode of warfare that carries no domestic downside in terms of U. S. casualties incurred during the attacks.  But an older code of conduct in the battlefield would view the kind of button-pushing fight we are presently engaged in as morally suspect, if not downright cowardly.

It was close to inevitable that cyberwarfare would take its place along more conventional means of fighting a military conflict.  But now that we have told the world we’re doing it, we should not cry foul if some fine day our own computer systems fall victim to a low-budget, focused attack that could do even more damage than ours did to the Iranian uranium facility.


Sources:  The New York Times report on the Olympic Games efforts appeared online on June 1, 2012 at  President Carter’s criticism of drone strikes as being violations of human rights appeared in the same publication on June 24, 2012 at  My blog “Stuxnet and the Future of Cyberwarfare” was posted July 24, 2011.