Blog

Collecting Toolmark Impressions

For the latest NIJ grant, we need to collect a set of tool marks from screwdrivers held at one of 5 specific angles. To do this, I modeled a jig in a program called openSCAD, which is an open-souce 3d modeling program. Here is the jig in action:

Here is a closeup:

It does a nice job of creating constant pressure on heavy-duty aluminum foil:

Here is one closeup, taken with a Sony a7r3 with a macro lens and two exension rings:

Fitting Signal Detection Theory Models

At 2019 IAI in Reno, I presented a talk in which I described how models based on Signal Detection Theory could be used to help summarize the data from proficiency/error rate tests (where ground truth is known). The slides for that talk are found here:

IAI 2019 talk on SDT

As part of that talk, I created a spreadsheet that allows anyone to explore and fit Signal Detection Theory models. You can download that here:

New Blog!

This current blog is hard to maintain, so I’ve started a new blog at the link below:

https://blogs.iu.edu/buseylab/

Is the Gaussian (i.e. normal) distribution tenable?

We recently documented likelihood ratios for individual print pairs in black box studies: Busey T, & Coon, M. Not all Identification Conclusions are Equal: Quantifying the Strength of Fingerprint Decisions. In press, Forensic Science International. pdf. Also see link to Presentation

This paper is important because it demonstrates that for many casework-quality pairs (ones that a majority of examiners with make an identification on) can have strengths of evidence that do not support a phrase such as “Identification”. We often see likelihood ratios in the range of 100-1000, which is much less than is typically associated with a term like “Identification” which is often interpreted by laypersons to mean the exclusion of all others.

Tom’s Brain Surgery

In Mid-November 2019 I started waking up at 2am every morning with headaches. This happened to coincide with the time of year where Bloomington gets really dry in the winter, and so I assumed that these were sinus headaches. These got worse, and I started to have balance problems that I assumed were residual effects of the headaches.

It took me until the end of January to actually get an ENT appointment. The ENT did not think my headaches were sinus-related and ordered an MRI. The MRI revealed a golf-ball sized cyst in my right cerebellum, which was constricting the flow of cerebra-spinal fluid to the 3rd and 4th ventricles of the brain. That pressure was causing my headaches and balance problems.

Toolmark Impression Comparison Interface

For the latest NIJ grant, we need an interface that will allow examiners to compare toolmarks from screwdrivers and chisels.

We’re in the process of creating software that will allow us to do a virtual splitscreen. The examiner will move the questioned impression around on the screen, and can compare it to one of 5 different exemplar impressions taken at different angles.

Here is a mockup of the interface, although the final version will differ: