Unit 5: Forensic Science: Research and Applications

Forensic science is the application of scientific principles and analytical methods to matters of law, converting physical traces from a crime scene into evidence admissible in court. As a thrust research area it sits at the intersection of chemistry, biology, physics, computing and law, and its research agenda is driven by the twin demands of the justice system: individualisation (linking a trace to a unique source) and evidential reliability (defensibility under cross-examination). This unit orients a dissertation towards the current, fundable and publishable directions within the field.

I. Orientation: The Field and Its Governing Principles

Forensic science (formalised through Locard's work, early 20th century) rests on the premise that every contact leaves a trace and that traces can be recovered, characterised and interpreted.

• Locard's Exchange Principle: whenever two objects come into contact, material is transferred both ways — the foundation of trace evidence recovery (fibres, glass, soil, gunshot residue).
• Individualisation vs. classification: evidence is either class-level (blood group, shoe size, ink type) narrowing to a population, or individualising (fingerprint minutiae, STR DNA profile) attributing to one source.
• Chain of custody: the documented, unbroken record of possession from collection to court; a break renders evidence inadmissible regardless of analytical quality.
• Admissibility standards: the Daubert criteria (testability, known error rate, peer review, general acceptance) govern whether a method's output is accepted as expert evidence.
• The ACE-V paradigm: Analysis, Comparison, Evaluation, Verification — the reasoning framework underlying pattern-based disciplines.
• Research drivers: reducing subjectivity, quantifying error rates, and building statistical weight-of-evidence models (likelihood ratios) rather than categorical "match/no-match" claims.

II. Forensic DNA Analysis

Molecular individualisation and its research frontier

DNA typing is the most statistically robust individualising discipline, and its research thrust is towards smaller, degraded and mixed samples.

A. Foundational method

• STR profiling: amplification of Short Tandem Repeat loci by PCR, sized by capillary electrophoresis; the standard CODIS core comprises 20 loci giving random-match probabilities below 1 in a trillion.
• Quantitation and thresholds: a profile is scored above an analytical threshold (peak height in RFU) to distinguish true alleles from baseline noise.

B. Thrust research directions

• Probabilistic genotyping: software (e.g. likelihood-ratio engines) deconvolutes DNA mixtures from multiple contributors, replacing analyst judgement with a computed LR — a leading dissertation topic.
• Touch / trace DNA: recovery from a few shed epithelial cells raises research questions on secondary transfer (DNA deposited by someone who never touched the object), central to interpretation validity.
• Massively Parallel Sequencing (MPS): next-generation sequencing reads sequence variation within repeats, enabling forensic phenotyping — predicting eye colour, biogeographic ancestry and approximate age (via methylation).
• Forensic Investigative Genetic Genealogy (FIGG): matching crime-scene profiles to consumer genealogy databases to identify distant relatives; a hot area with unresolved privacy and consent research questions.

III. Fingerprint and Pattern Evidence

From categorical opinion to measurable error rates

Pattern disciplines are under the strongest pressure to demonstrate empirical validity, making error-rate research the dominant thrust.

A. Basis of comparison

• Levels of detail: Level 1 (ridge flow pattern — loop, whorl, arch), Level 2 (minutiae — ridge endings and bifurcations), Level 3 (pores, ridge edges).
• AFIS: Automated Fingerprint Identification Systems rank candidates by minutiae topology; the examiner still performs the final ACE-V comparison.

B. Thrust research directions

• Quantifying examiner error: black-box studies measuring false-positive and false-negative rates to satisfy the Daubert "known error rate" requirement.
• Latent print chemistry: research into enhancing weak latents on difficult substrates (cyanoacrylate fuming, physical developer for wet paper) and dating deposits from decay of amino acids and lipids.
• Statistical fingerprint models: developing likelihood ratios for ridge configurations to replace absolute identification claims.

IV. Forensic Chemistry and Toxicology

Identifying substances and reconstructing exposure

This branch answers what a substance is and what it did to the body, with instrumentation and interpretation both active research fronts.

A. Analytical workhorses

• GC-MS: gas chromatography separates a mixture; mass spectrometry fragments each component to a characteristic spectrum — the confirmatory "gold standard" for drug identification.
• LC-MS/MS: liquid chromatography with tandem MS handles thermally labile and polar analytes (many drugs of abuse and their metabolites in blood).

B. Thrust research directions

• Novel Psychoactive Substances (NPS): designer drugs whose structures change faster than reference libraries; research focuses on non-targeted screening and high-resolution mass spectrometry to flag unknowns.
• Post-mortem redistribution: drug concentrations shift after death, complicating cause-of-death interpretation — a toxicology research priority.
• Alternative matrices: hair, oral fluid and dried blood spots for detecting window of exposure; hair segments record months of drug history.

V. Digital and Cyber Forensics

Evidence in a volatile, encrypted world

The fastest-growing thrust area, driven by ubiquitous devices and the fact that most crime now leaves a digital footprint.

A. Core principles

• Order of volatility: capture the most transient data first (RAM, cache) before disk, because powering down destroys volatile evidence.
• Forensic imaging and hashing: create a bit-for-bit copy and verify integrity with a cryptographic hash (e.g. SHA-256); a matching hash proves the copy is unaltered.

acquire  -> write-blocked bit-image of source
hash     -> H(image) == H(source)   # integrity proof
analyse  -> work only on verified copy

B. Thrust research directions

• Anti-forensics and encryption: research into lawful recovery from encrypted volumes and detection of steganography and data hiding.
• Cloud and IoT forensics: evidence spread across jurisdictions and ephemeral servers; the research problem is acquisition when data is not physically local.
• Deepfake and media authentication: detecting AI-generated or manipulated audio/video, an urgent frontier as synthetic media enters courtrooms.
• Mobile and app artefacts: reconstructing user activity from application databases and geolocation logs.

VI. Forensic Biology and Anthropology

The body as evidence

These disciplines identify individuals and reconstruct events from biological remains, with quantification again the research thrust.

A. Scope and methods

• Serology: presumptive and confirmatory tests for body fluids (Kastle-Meyer for blood, acid phosphatase for semen) preceding DNA work.
• Osteology: the skeleton yields a biological profile — sex from the pelvis, age from epiphyseal fusion and dental development, stature from long-bone regression equations.

B. Thrust research directions

• Forensic entomology: insect succession on remains estimates the post-mortem interval (PMI); research refines developmental data across temperature and geography.
• Microbiome as a clock: succession of the thanatomicrobiome offers a data-driven PMI estimator less sensitive to environment than temperature alone.
• Body-fluid identification by mRNA/methylation: tissue-specific gene expression distinguishes menstrual from peripheral blood — impossible by classical serology.

VII. Crime Scene Reconstruction and Impression Evidence

Reassembling the sequence of events

Reconstruction integrates the physical evidence above into a coherent, testable account of what happened.

A. Principles and evidence types

• Bloodstain Pattern Analysis (BPA): droplet shape and directionality reconstruct the mechanism; area of origin is triangulated from impact-angle trigonometry.

sin θ = width / length     # θ = impact angle of a bloodstain

where width and length are the minor and major axes of the elliptical stain.

• Impression evidence: tool marks, footwear and tyre impressions compared for class characteristics (make, size) and individualising accidental damage.

B. Thrust research directions

• 3D scanning and photogrammetry: laser scanners capture scenes as measurable point clouds, enabling virtual re-walkthroughs and courtroom visualisation.
• Ballistics and IBIS: correlating breech-face and firing-pin marks on cartridge cases through image databases; research targets objective 3D surface metrics over subjective microscopy.
• Cognitive bias research: studying how contextual information biases examiner conclusions, motivating sequential unmasking and blind verification protocols.

VIII. Cross-Cutting Research Themes

What a dissertation must engage with regardless of specialism

Certain concerns run through every forensic sub-discipline and define the modern research agenda.

• Validity and error rates: the post-2009 NAS report and 2016 PCAST report demand empirically established accuracy for every method; measuring this is itself a research output.
• Likelihood ratio framework: expressing evidential weight as LR = P(evidence | prosecution) / P(evidence | defence), moving away from categorical opinions across DNA, glass, fibres and marks.
• Standardisation and accreditation: ISO/IEC 17025 laboratory accreditation and validated Standard Operating Procedures underpin admissibility and reproducibility.
• Emerging technologies: machine learning for pattern classification, portable spectroscopy (Raman, FTIR) for field screening, and rapid on-site DNA — each raising validation and interpretability questions suitable for original research.
• Ethics and privacy: genetic genealogy, biometric databases and phenotype prediction pose consent and civil-liberty problems that responsible research must address alongside technical performance.