September 2, 2006
Global Warming & Bush's Reading List
The Washington Post has a very good review of James Lovelock's new book about global warming, "The Revenge of Gaia: Earth's Climate Crisis and the Fate of Humanity." [See Here] To be sure, Lovelock tends to the extreme, at least in his colorful descriptions of, and basic nihilism about, global warming. But his basic view is shared by most mainstream scientists and even if his predictions are off by a substantial margin, the prognosis remains bleak.
When it comes to science policy, three factors in particular must be considered. The first is the quality of the science that supports a particular prediction, the second is the gravity of harm that will occur if the prediction occurs, and the third is the cost associated with remedying or avoiding the harm. Global warming presents a particularly challenging issue given these three considerations. Most mainstream scientists agree that the science is solid, though it is complex and not unassailable. The consequences of global warming, especially if the most dire predictions prove true, will be catastrophic; and even many of the most modest forecasts predict very severe effects. Finally, as Lovelock believes, there is some question whether there is very much that can be done to avoid the inevitable tragedy that lies on our collective horizon. Admittedly, therefore, global warming presents no simple problem for policy makers, even those most knowledgeable about the science. One thing is for sure, however, policy makers ignorant of the science are unlikely to handle the complexities of this issue very well. Perhaps George Bush should consider adding Dr. Lovelock's book to the 60 books he claims to have already read this year. It may not be as compelling as "Pride and Prejudice," but there are valuable lessons to be learned nonetheless.
August 30, 2006
Women Clerks at THE SUPREME COURT
A subject of great interest around the blogs these days is the low number of women clerks hired by Supreme Court justices this year. It is the subject of a story by Linda Greenhouse in today's New York Times. (See here) Apparently, the number of female clerks hired this year fell by half from a year ago. Out of 37 clerkships for the new term, only 7 are women. Trying to downplay the significance of this fact, Justices Souter and Breyer, both with a history of hiring many female clerks, attributed the low number to "random variation." This proposition, of course, is testable. According to the story in the Times, 37 clerks were hired from a population of applicants that contained approximately 1/3 women. (This 1/3 number is suspect, since more than half of all graduates today are women. If the number is higher, then the expected number of clerks would be higher as well -- thus strengthening the basic conclusion I reach below.) All things being equal, therefore, we would have expected the justices to hire approximately 12 female clerks. They hired 7. The question, therefore, is what is the likelihood that this departure from what is expected would happen by "random variation." According to the Supreme Court's own precedent, Castaneda v. Partida, this can be derived by first determining the standard deviation for this binomial distribution (male/female). The standard deviation here is the square root of the product of (1) the total size of the group (37), times (2) the percentage of women (1/3), times (3) the pecentage of men (2/3). A little simple math gives us a standard deviation of 2.86. The obtained number of women this year, therefore, is almost two standard deviations from the expected number. According to the Court in Castaneda, "if the expected value and the observed number is greater than two or three standard deviations," chance fluctuations are an extremely unlikely explanation. Certain caveats must be noted. First, the sample is small here, which affects how confident we can be in drawing any conclusion. Also, the departure was less than two standard deviations, so it does not quite meet the arbitrary standard created by the Court. Finally, while the data suggest that chance is not a good explanation, they cannot tell us what alternative hypotheses might explain the disparity. So, are Justices Souter and Bryer correct? Is "random variation" a good explanation for the small number of female clerks? My answer is, probably not. But more research would be needed to say what is causing the disparity observed. I am sure the Justices will welcome further research on this question.
August 29, 2006
Breath Analyzers as Black Boxes
Breath analyzers designed to test drivers’ level of intoxication are black boxes. Breath goes in and an estimated blood alcohol concentration (BAC) reading comes out. Typically, states require periodic testing of these instruments to assess their accuracy. The testing involves using a sample with a known alcohol concentration to assess whether the instrument is properly calibrated. Defendants from time to time attempt to force the state to do even more to certify every step of the operation. In State v. Ensey, 881 A.2d 81 (N.J. 2005), the defendant moved to suppress the results of a breath test because the state failed to provide a witness who could testify concerning the preparation of the mixture used to calibrate the breath analyzer used in this case. The court rejected the defense argument that this was required to lay an adequate foundation for the admission of the test results.
A more interesting development is a flurry of decisions in Florida on the question of whether the state must turn over to the defense the software used by the breath analyzer to calculate an individual’s alcohol content. Breath analyzers must make at least three separate calculation in order to provide a BAC. See Patrick Barone, Commentary: Unlocking the Mystery Behind Breath Testing: The Right of Access to Source Codes. 4/ 17/ 06 Mich. Law. Wkly, 2006 WLNR 9002476.
First, the machine must measure the level of alcohol in the breath. This is done by measuring the resistance of an infrared light beam as it passes through a sample chamber containing the individual’s breath. The theory behind this methodology is that ethanol absorbs infrared light and, therefore, the loss of energy in the beam may be attributed to the level of alcohol in the sample.
Second, the breath alcohol level (BrAC) must be used to estimate BAC. This is done using Henry’s Law: At a constant temperature, the amount of a given gas dissolved in a given type and volume of liquid is directly proportional to the partial pressure of that gas in equilibrium with that liquid. In this context the liquid is blood and the gas is breath in the closed system of the lungs. The typically accepted ratio used in law enforcement is 2,100 to 1. For every molecule of alcohol in the breath (BrAC) there are 2,100 molecules in the blood (BAC). This, of course, is the ratio under ideal conditions and in the past there has been a fair amount of litigation over the appropriate constant to apply.
The third calculation relates to the proper application of Henry’s law. The law applies to gas in equilibrium with the liquid. This is the case only with respect to alveolar (deep lung) air. In an attempt to obtain alveolar air, officers are routinely asked the suspect to blow hard into the machine. Under normal conditions, when there is no alcohol present except that in the lungs, as one blows into the machine the breath alcohol level rises and then levels off as alveolar air passes through the machine. That is, the highest BrAC is found in alveolar air. If one does not blow hard enough alveolar air may never reach the machine, but if the only alcohol present is that evaporating into the lungs from the blood the result is a reading that underestimates one’s true BAC. Problems arise, however, if there is other alcohol present. If one were to rinse one’s mouth out with a glass of bourbon right before taking a test some of this mouth alcohol would evaporate when taking the test and thus the machine would return a reading that overestimates BAC.
In order to avoid this difficulty, many states require that an officer observe the defendant for 15 or so minutes prior to administering the breath test to be certain the individual has not put any alcohol into his mouth or done anything else such as belch that would affect the reading. Failure to observe the individual during this period occasionaly leads to the exclusion of the subsequent test. See State v. Korsakov. 34 S.W.3d 535 (Tenn. Crim. App. 2000).
The equipment itself is designed to detect this problem and to return an “invalid sample” message when some other source of alcohol seems to be present. This is accomplished through “slope detection.” The machine attempts to detect a rising “slope” of greater and greater BrAC as the individual blows into the machine and when it does not detect a slope it should return the “invalid sample” message. When calibrating an instrument, the state should test to see if the slope detector is working properly. However, there is some question as to whether this slope detection feature always works correctly. See Michael Hlastala, Wayne Lamm and James Nesci, The Slope Dectector Does not Always Detect the Presence of Mouth Alcohol, 20-MAR Champions 57 (March 2006).
In a number of recent Florida cases, defense attorney’s have asked for the computer code by which the breath analyzers make these calculations. They argue that without access to the code there is no way to know if the results of the machine would pass muster under a Daubert or Frye reliability analysis. Several judges have found that defendants are entitled to this code. See Paul Quinlan, Surprise Court Ruling threatens to Nullify Results of DUI Tests, 11/4/05 Sarasota Herald Trib. A1, 2005 WLNR 17892443. The manufacturer of the equipment used in Florida, the Intoxilyzer, has refused to produce the code which it says is a trade secret.
This litigation raises a pair interesting questions concerning all equipment used in forensic testing. First, how far back may the parties go in an attempt to deconstruct the scientific knowledge proffered by an opponent? Second, to what extent should the results of an instrument be admitted if it passes a reliability testing protocol even when we are not certain exactly why it is able to pass the protocol?
August 28, 2006
News from the Fingerprint Community
For news straight from the community of experts who specialize in fingerprint identification, you can subscribe to their weekly newsletter here. It's an informative bulletin that Kasey Wertheim sends to your email in-box each Monday morning. As their webpage explains, "The purpose of the Detail is to provide Latent Print Examiners with another avenue to receive and relay current information regarding our field." It provides lots of great information and, while it is obviously tailored to the professional community of fingerprint examiners, it does not shy away from including news and comments that are critical of the field.