If you use the ULW comparison tool to do your comparisons, you've undoubtedly seen the infamous 'Complex Decision' checkbox. And if you mouse over the checkbox the tooltip that defines the variables to consider are revealed. (it's in the help menu too, but who reads the manuals? amirite?)“In the future, automated systems may be able to handle the straightforward comparisons, allowing human examiners to focus on the most complex cases.”
the Black Box Study mentions:If an examiner encounters complexity due to the quality and/or quantity of features, low specificity of features, significant distortion, or disagreement among examiners, the Complex Decision checkbox should be used.
Michele Triplett's paper Complexity, Level of Association and Strength of Fingerprint Conclusions says:Latent print examination can be complex because latents are often small, unclear, distorted, smudged, or contain few features; can overlap with other prints or appear on complex backgrounds; and can contain artifacts from the collection process . Because of this complexity, experts must be trained in working with the various difficult attributes of latents.
And last but not least, Henry Swofford's presentation titled Objective Classification of Fingerprint Image Complexity draws the conclusion:The interpretation of data (Galton or ancillary features) has predominant ambiguity (the interpretation of data is questionable making it difficult to demonstrate to the satisfaction of others); or
The correlation of data is extremely limited (making it necessary to use rarity, ridge shapes, edges, pores, or features in simultaneous impressions) then the comparison is considered Complex
Seemingly then, we have a word used in multiple places in multiple ways......Shocking I know. Quick, someone form a committee to come up with a new word that does all of the work of the old word without carrying of the baggage of said word.DFIQI or other similar software algorithms can classify the relative complexity of a fingerprint comparison a priori based on analyst performance metrics at various clarity scores
We have some tension among the different definitions. Some say complexity is confined to the latent, some apply them to the comparison or tension between Examiners (A physical/Social difference). Some say disagreements among examiners count as complexity, whereas others would put a consensus judgement in which there are disagreements among examiners as a Quality Assurance measure (Complexity vs Strength of conclusion difference). Some say the use of what is traditionally known as 3rd level detail increases complexity scale whereas other's would say that third level detail has high discriminating power and therefore lowers the complexity.
Arguably, complexity matters. People seem to be attaching a risk of error to complexity and floating the idea that non complex (simple?) prints can be compared via algorithm, a la " Latent Lights Out", which OK, I'm fine with that.
The questions I'm asking specifically are:
Does your agency define complexity, and if so how is it defined?
Alternatively if your agency does not define complexity, what are the key elements of complexity as an Examiner?
Also, how does complexity correlate to risk? Uncertainty?
As always the Zero Point ID and Mayfield are good benchmarks against which to test your ideas.....