Tea Time With Jesse

Six of One, Half Dozen the Other

Ripples and Timing

Posted by middlerage on September 12, 2013

In the midst of all this sturm and drang about Syria, the ripples of Eric Snowden and his NSA leak keep expanding. Could it be his action is the most significant development of our time? On the one hand congressmen decry his treason, and on the other hand they demand answers from the NSA and ponder revamping the FISA court.

Anyhoo, in the annals of unfortunate timing, Apple yesterday announced their new iPhone 5S with fingerprint-reading gatekeeper technology. Half a year ago (and a lifetime in Eric Snowden chronology) we all would’ve said, “Cool!” Today we say, “So now the NSA is going to scoop up our fingerprints and keep them on file.” Apple is scrambling to say not to worry, the fingerprint data is only kept locally, not on servers or the cloud. To which the techno-pundits say, “Yeah, we’ll see. Let some hackers take a crack and then we’ll decide.” And to which I say, “I don’t care if it is local, if you live in California, the police have the right to break into your phone without a subpoena after a traffic stop. If your fingerprint is stored in the phone, what’s to keep them from downloading it and running it against their database of unsolved crime fingerprints?”

And finally, just to go wildy off on a tangent… What do we really know about fingerprints anyway? Here is an excerpt from a Pacific Standard article on why fingerprints may not be 100% unique:

A 2011 UCLA Law Review article, “The Need for a Research Culture in the Forensic Sciences (pdf),” whose 13 authors include Mnookin and Kaye, addressed this, raising questions such as: “How frequently might a portion of two fingerprints – or striation marks on bullets, or toolmarks, or handwriting specimens – share any given degree of similarity even if they derive from different sources?”

And another excerpt from a report to Congress by the National Academy of the Sciences:

Often in criminal prosecutions and civil litigation, forensic evidence is offered to support a claim that an evidentiary specimen is a “match” to a particular individual or other source. With the exception of nuclear DNA analysis, however, no forensic method has been rigorously shown to have the capacity to consistently, and with a high degree of certainty, demonstrate a connection between evidence and a specific individual or source. Yet, for years, the courts have been led to believe that disciplines such as fingerprinting stand on par with DNA analysis. For example, in a decision issued by the Seventh Circuit, the court reported that an FBI fingerprint expert had “testified that the error rate for fingerprint comparison is essentially zero.” In a later decision issued by the Fourth Circuit, that court cited the Seventh Circuit opinion approvingly, noting that an expert from the FBI had testified that the error rate for fingerprint comparison was “essentially zero.” The committee’s report rejects as scientifically implausible any claims that fingerprint analyses have “zero error rates.” A “zero error rate” is a myth in fingerprint analyses and in all other forensic disciplines. That is no surprise, however, because there is no such concept as a zero error rate in good scientific analysis. Of greater concern is the dearth of solid research to establish the limits and measures of performance and to address the impact of the sources of variability and potential bias in most disciplines.

Another serious concern is contextual bias. Some studies have demonstrated that identification decisions on the same fingerprint can change solely by presenting the print in a different context. In one study, for example, fingerprint examiners were asked to analyze fingerprints that, unknown to them, they had analyzed previously in their careers. Contextual biasing was introduced – that is, examiners were told that the “suspect confessed to the crime” or the “suspect was in police custody at the time of the crime.” In one-third of the examinations that included contextual manipulation, the examiners reached conclusions that were different from the results they had previously reached.


2 Responses to “Ripples and Timing”

  1. As a resident of the much-maligned Republik of California I feel I must mount a defense, tepid though it may be. My understanding is that the cops can demand your smart-phone in a traffic stop, and if you haven’t set up any sort of secure login (that looks wrong and I want to hyphenate, but this link tells me not to: http://grammarist.com/spelling/log-in-login/ ) the contents of your device are considered to be “in plain view.”

    I have set passwords on my personal electronic devices and my belief is that the cops would have to produce a subpoena to make me open it, or allow a tech to try to crack it.

    If I am wrong, please let me know.

    Of course, my personal data could already have been compromised by our out-of-control, big-brother government. Fortunately, my personal life is more boring than I can possibly describe. Need some more caffeine, NSA lackey?

  2. Dahveed said

    That last paragraph about contextual bias is pretty damning. Up to that point, I was pretty skeptical about the fingerprinting skeptics. When combined with other evidence, I probably wouldn’t require “zero error rate” to remove reasonable doubt. But a 33% error rate due to contextual biasing — that introduces some very reasonable doubt.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: