How to science, by a lawyer

[Boyd Baumgartner]

https://vimeo.com/215279196

(the permissions won't let me embed)


If by 5 minutes you aren't laughing, you have no soul. Since the last time we posted videos of this guy, they were made private, I made sure to download them first.

[josher89]

"We no longer let them get away with saying stuff like this in Cook County..."

--kudos to you, Mr. Max! (insert sarcasm)

I think it's right around the 5min mark that I started to smirk...the same mother-father argument? I've never heard that one before...

[Alan C]

It requires a password. Is that just since you posted this?

[Boyd Baumgartner]

Yes, like the other videos. It's not like it's a secret anyway, you just don't have to pay $695 to hear it.

Just go here (pg 59 to start). It's literally the same content, but someone went ahead and took notes for you....

[josher89]

The ELI5 of his mother-father argument:

If we go into a room of 70 people and correctly say that none of them have the same mother and father, we are using that same logic to say that no two people have the same fingerprint.

[Steve Everist]

And a couple of minutes after the mother/father argument is the discussion on ACE-V. It's introduced by saying that we love acronyms because they're hard to figure out and the juries will just gloss over them, as if it's our goal to trick or confuse the court. It could easily be argued that the exact opposite is why acronyms are used. In the case of ACE-V, the purpose is to simplify the description of a process with the primary focus being the ability to better educate jurors and the court.

And then he goes on to minimize the comparison process as described by ACE-V as:

A: I look at the print
C: I look at the print
E: I look at the print
V: My coworker does too

This gets a few laughs, but isn't helping those in attendance. It could actually hurt them in court to assume this level of simplicity, but then have a well-educated analyst thoroughly explain and educate the court.

It's akin to describing a trial as:

Prosecution: Asks some questions
Defense: Asks some questions
Prosecution: Asks some more questions
Defense: Asks even more questions
...and then a group of people who couldn't figure out how to get out of jury duty decide the fate of a human being.

We can do better than that - from all sides.

[josher89]

If I had the time (actually, if I had the time I wouldn't do this as there's a lot of other things I'd need to do) I'd make the rebuttal video exactly as Steve described. Totally dumb down what happens in court, bring in all of the Law and Order dramatics from TV, and make an hour-long lecture on it. Then, I'd write a few articles for some minorly obscure legal journals and become an expert on legal proceedings all on my own. I'd eventually convince some think tank to give me a bunch of money to do 'research' and publish my findings.

[ER]

I've got Premiere. Write up a script, get some clips together, and let's do this.

[josher89]

Another video I just received:


youtu.be/MJYMFDBEQtk

What if there were latent print folks in the audience...would his presentation change?

He briefly mentions (around the 6:00 mark) a local lab didn't know about "basic fundamental concepts in their field" and says close non-matches and then incidental similarity. I know what a close non-match is but could only assume what incidental similarity was because of the words in the term.

Has anyone else used incidental similarity in their description of close non-matches? I have never used that term nor have I heard it used in my circles. Just curious.

Around the 12min mark, he talks about the 97% pass rate of the 2017 CTS test and likens that with 97% of juniors that take the SAT getting a 1600! Like, these analogies aren't even close! [me arriving to work on time for 280 days straight is just like my neighbor not burning his steak on the grill last night...] It's incredibly misleading and unfortunate.

[ER]

What if there were latent print folks in the audience...would his presentation change?

Listen to our interview with him. http://doublelooppodcast.com/2019/12/br ... isode-210/


To be fair:
I have some serious problems with some of the things that he says about ACE-V and the training that he's providing to other attorneys that oversimplifies complex issues. I especially have an issue with training attorneys how to mislead the jury by using connective ambiguity to dispute an ID.


However:
I can't simply dismiss his points on PTs, especially the current ones from CTS. It is a problem that we need to deal with and a component of our system that needs to improve.

Let's take the SAT analogy.
What's the point of the SAT? To distinguish and differentiate students. Not everyone can go to Harvard. Not everyone can succeed at Harvard. Harvard needs a way to tell students apart. If 97% of students scored 1600, then the SAT would be a poor tool for Harvard to use.

So, what's the point of PTs? He states (and I agree) that they should a) identify where training or improvements are needed; and b) provide an objective measurement of proficiency or competency to establish expertise in court. In my old lab we typically had about 15 people take CTS PTs each year. On average 1 person would get 1 answer wrong about once every 5 years. Is that test hard enough to identify training and improvement needs? I would say no. Depending on your agency's policy, he and his colleagues would have either passed the test (if they'd had a fully-trained verifier) or been given the opportunity to retake the test. Is that test hard enough to be used to establish expertise for court? Obviously, no. Therefore, CTS is a poor tool for labs to use to identify training and improvement needs and it's a poor tool for courts to use to establish expertise.

I'm not saying that CTS PTs are anything like the SAT, but the analogy does hold in that CTS PTs are a poor tool for what they're being used for just like an SAT with an extremely high rate of 1600's would also be a poor tool.

In other words, just because I don't like some of the other BS that I've heard, doesn't mean that I don't recognize the serious need for improvement in the PTs that we take. Thankfully, the spotlight will probably switch to DNA where they have a test even easier than ours. With DNA in the crosshairs maybe QAMs will understand that PTs should be harder and that passing a PT shouldn't just mean 100% correct conclusions. It should involve a real look at whether an examiner is PROFICIENT in the entire process. I want a test where I have to show my work AND where I get credit for showing my work AND that tells me where I need to improve AND I pass with a 90%. Is that so bad?

PS
And no, I don't think that 90% is an acceptable rate of error for our field. But if a test is so hard that competent examiners get a 90% and get to find out what to work on to get better, than that's a really good test and a great tool.

[ER]

I didn't get to see the original Vimeo link, so if you downloaded that, please PM me.

As for PT's

Tests that are too easy do not adequately serve any of these goals. So, if you're using PTs as your main QA tool and that test is calibrated to be so easy that everyone passes it, your QA program will never identify breaking points.

Couldn't agree more.