|
|
Sunday, June 1, 2014
By Michael Nichols
Categories: Michael J. Nichols
Michigan attorney Mike Nichols is a leader in cutting-edge forensic evidence. Nichols spent the last 5 years of his practice dedicated to educating judges, juries and other lawyers on why a forensic analysis can only give us limited information about what happened in a case. Nichols says "uncertainty is everywhere in science and it is especially prevalent in measurements of what is in a human blood or breath sample."
There are new examples of "uncertainty" in the national media. One was revealed in the Sunday June 1st edition of the New York Times in an article that revealed the National Security Agency (NSA) collected facial feature information through the internet since 9/11. A link to the article is here:
http://www.nytimes.com/2014/06/01/us/nsa-collecting-millions-of-faces-from-web-images.html?emc=edit_th_20140601&nl=todaysheadlines&nlid=36244735&_r=1
The information was one of many new revelations from the trove of leaks by former contractor Edward Snowden. Nichols says "the conditions must be almost perfect in order to match how the feature or recognition-data was collected to the 'unknown' facial features. We know that perfect conditions almost never exist in the world." Nichols points to the acknowledgment by one of the leading developers of the facial recognition technique and an expert in the field: "Professor Megherbi put it best when she described the limitations of the technique by explaining that different angles, lighting or other variations in the condition can affect the algorithms of facial recognition software."
In fact, one of the main NSA programs called "Tundra Freeze" produced dozens of false hits among 42 different possible faces that matched just 1 unknown face in a 2011 identification project.
Another startling revelation came recently when the United States Supreme Court issued a ruling on May 27th that rejected the use of an arbitrary cut-off for Intelligent Quotients (IQ) levels for disqualifying inmates for execution. The case, Hall v Florida, :12-10882, turned in part on the analysis by the majority of justices that too many variances in the collection and analysis of IQ data can cause uncertainty in whether a person who has an IQ of 70, actually falls above or below the threshold level that may disqualify a person for execution in many states.
A link to the main article is here:
http://www.nytimes.com/2014/05/28/us/court-rules-against-florida-iq-rule-in-death-cases.html
A link to the Supreme Court opinion can be found here:
http://www.supremecourt.gov/opinions/13pdf/12-10882_4f15.pdf
Nichols says the IQ scores of one of the co-petitioners in the case is a telling example of uncertainty in measurement. "the fact that on several different times a person's IQ can be measured as disparately as 4 points or more and that is literally the difference between life and death shows why it is so vital for us to understand uncertainty and statistical applications of expressing results of measurements and analyses," says Nichols.
Nichols is an adjunct professor of forensic evidence and criminal law at the Western Michigan University/Thomas M. Cooley Law School Lansing campus. He is author of the Michigan drunk and drugged driving practice book for lawyers and also a member of the National College for DUI Defense Fornesic Evidence Committee. For the lawyer who is committed to results and a leader in complicated cases and evidence, reach out to Mike Nichols at mnichols@nicholslaw.net or call 517.432.9000.