The CFA domains test two opposite working modes: producing an image from a person's memory, and producing or comparing images from physical evidence. Studying them as one generic drawing skill blurs the decisions that matter — what to ask, what to change, and what your finished image can honestly claim. Build your plan around that split: drill facial anatomy as measurable landmarks, rehearse composite interviews out loud, separate progression logic from reconstruction logic, and practice describing image comparisons in careful, non-identifying language. The scenarios and rubric below show how those decisions differ domain by domain.
Facial Anatomy: Study Landmarks and Proportions, Not Generic Features
The anatomy domain rewards treating the face as named, palpable landmarks — glabella, nasion, subnasale, pogonion, gnathion, gonion — plus the proportional relationships between them, because every other CFA domain depends on that vocabulary.
Landmarks are points you can locate on a skull, on a photograph, and on a living face, which is what makes them the connective tissue of forensic art. Nasion sits at the root of the nose between the eyes; subnasale at the base of the nose; gnathion at the lowest point of the chin; gonion at the angle of the jaw. When you study proportion, measure intervals between landmarks rather than eyeballing 'long face' or 'short chin.' A practical drill: photograph a skull cast or a face, mark each landmark, then redraw the face using only landmark coordinates and compare the redraw to the original. Differences reveal whether your proportional eye or your rendering hand is the weak link — errors become localizable instead of general.
Morphology — how the same landmark region varies — matters as much as locating it. Nasal bridge height, gonial angle, orbital shape, lip thickness, and chin projection all vary along ranges, and descriptive vocabulary for those ranges is what lets you translate a witness's words or a skull's features into an image. Practice by describing photographs aloud in anatomical terms before drawing, then check whether your drawing matches your description rather than the photo's general impression. This keeps two skills separate that otherwise merge: accurate perception and accurate rendering.
Composite Drawing: The Interview, Not the Pencil, Decides the Outcome
A composite is drawn from a witness's memory, so the governing skill is the interview — open prompts, non-leading vocabulary, and controlled feedback — with drawing used to test and refine memory, never to suggest features.
Worked scenario: a witness says the suspect's nose 'was big, like a hawk.' The interviewer-drawer responds, 'So a hooked, prominent nose like this?' and slides a partial sketch across the table after ten minutes. That is a plausible mistake: the prompt supplies the feature, and the early reveal invites recognition of your drawing rather than recall of the face. A better decision is to ask what the witness could compare the nose to, offer anatomical ranges neutrally ('bridge high or low? tip upturned or straight?'), keep the drawing hidden, and schedule one structured review near the end. Why it matters: a contaminated composite wastes investigative time even if it looks skillful.
Build the interview side deliberately. Run a full mock session with a partner who studies a photo you have never seen: free recall first, then feature-by-feature prompts, then a single late viewing with specific revision questions. Mark every moment you almost suggested an answer. Your checklist should cover opening instructions (tell the witness memory is imperfect), avoiding yes/no prompts, handling 'I don't know' without supplying options, and documenting which changes came from the witness versus from you. Reviewing that transcript teaches more than another hour of shading practice.
Age Progression and Regression: Growth Rules and Aging Rules Are Not the Same
Progression projects an image forward: craniofacial growth for children, soft-tissue and skin change for adults. Regression works backward with far weaker anatomical guidance, so each task needs different reference material and different claims about certainty.
For a child, progression is largely a craniofacial growth problem: the cranium approaches adult size early while the midface and mandible continue to grow, so proportions shift — the lower face lengthens relative to the head. A common trap is applying adult aging logic (wrinkles, sagging) to a child's face and only thinning the cheeks. The better decision is to change proportions first, using growth references appropriate to the target age, then adjust individual features. For adults, the dominant changes are soft tissue — fat redistribution, skin laxity, descent of features — applied over an essentially unchanged bony framework.
Regression inverts the problem with weaker constraints: you infer a younger face without knowing the person's earlier weight, health, or hairstyle, so reference material legitimately obtained through the requesting agency (family photographs, documented context) matters more than any formula. Practice both directions on the same face and write down which changes you could justify anatomically and which were guesses. That written split — justified versus speculative — is the habit that keeps progressed images useful and keeps you honest when the image is released to the public.
The decision table below contrasts the image-transformation tasks so you can check which anatomical basis and which decision point governs each one before you start drawing.
| Task | Starting input | Dominant anatomical basis | Key decision point |
|---|---|---|---|
| Age progression (child) | Known image | Craniofacial growth: proportions before wrinkles | Which growth references match the target age |
| Age progression (adult) | Known image | Soft-tissue aging over a fixed bony framework | How much change is justified versus speculative |
| Age regression | Known image, no earlier data | Weak constraints; family and contextual references | What is inference, and is it labeled as such |
| Skeletal reconstruction | Skull or postmortem remains | Tissue-depth data plus skull-derived features | Whether identity information has biased the base |
Skeletal and Postmortem Reconstruction: Tissue Depth and Bias Control
Reconstruction builds a face from a skull using average tissue depths placed at standard landmarks plus anatomical inference for features the skull suggests — nasal spine, dentition, brow — while controlling bias from case information.
Two-dimensional and three-dimensional reconstruction share a core method: average soft-tissue depths at standard landmarks, derived from imaging studies, are marked on the skull or its outline and connected into a facial form. The skull constrains much of the work — the nasal aperture and spine suggest nose width and projection, dentition suggests lip support, the gonial region suggests jaw shape — but eyes, ears, and hair are weakly determined, so a reconstruction is an investigative aid, not a portrait. Study by drawing the same skull twice from landmark data alone and note how much of the likeness shifts when soft-tissue guesses change.
Bias control is the discipline-specific difficulty. If you know what a proposed missing person looked like, your hand will drift toward that face regardless of the skull, and an incorrect reconstruction can feel convincing. The practical safeguard is procedural: complete the anatomical base before viewing any reference photograph of the proposed identity, and record when you first saw it. A useful drill: reconstruct from a skull cast, compare against the source face, and classify every feature as skull-driven, depth-table-driven, or guessed. Guessed features are where bias enters, so they deserve explicit notation in your case documentation.
Image Comparison: Morphological Analysis Without Overclaiming
Facial image comparison is feature-based morphological analysis: cataloguing correspondences and differences across defined regions, then expressing a conclusion in terms that stop short of claiming personal identification.
Worked scenario: you are asked whether two surveillance images show the same person. The plausible mistake is anchoring on one strong similarity — a distinctive scar — and writing 'consistent, likely the same individual.' A better decision is a structured comparison: describe each region (hairline, eyebrows, eyes, nose, lips, chin, ears, facial outline) separately in both images, note image-quality limits for each region, list correspondences and differences, and only then phrase a measured conclusion. Why it matters: a scar can be shared or change over time, and a conclusion that outruns the method can misdirect an investigation and fail under scrutiny.
Learn the distinction between holistic comparison — an overall impression of resemblance, useful for generating a hypothesis — and componential comparison, the region-by-region morphological detail that supports a written analysis. Practice classifying limits too: angle, resolution, lighting, expression, and elapsed time each change which features are reliable. A graded exercise: take two photos of yourself months apart, write the comparison, then have someone score whether your descriptions would let them find the right face among distractors. Vague descriptions ('similar nose') score poorly; specific ones ('nasion deep-set, tip bulbous, deviated slightly left') score well.
Professional Practice: Scope of Role, Documentation, and Ethics
The professional domain covers working under a requesting agency's authority, documenting your process, keeping your role limited to producing investigative aids rather than identifying individuals, and meeting the IAI's Code of Ethics.
Your product in composite and comparison work is an investigative aid: something that may generate leads, never a statement that a depicted person committed an offense or that two images show the same person. Practicing the language matters — rehearse explaining to a hypothetical officer why you will not attach a caption implying identification, and how you record witness statements, revisions, and your own decisions. Documentation discipline belongs in every certified forensic discipline because it is what makes the work reviewable after the fact.
The IAI publishes a Code of Ethics and maintains certification policies for its disciplines, including Forensic Art; read them directly and translate each obligation into a behavior you can demonstrate — impartiality, honest representation of your method's limits, proper handling of case materials, and accurate claims about your training. A short drill: write a one-page standard operating procedure for a composite case (intake, interview, rendering, review, documentation, storage) and check it against each ethical obligation you can name. Gaps in that SOP are gaps in this domain.
An Adaptable Preparation Sequence with a Self-Check Rubric
Sequence the domains in dependency order — anatomy, then composites, then progression and reconstruction, then comparison, then practice and ethics — and convert each block into an observable output you can score against a rubric.
A realistic sequence: weeks one and two, landmark labeling and proportion drills on skull casts and photographs; weeks three and four, composite interview rehearsal with recorded transcripts; week five, build age-progression and regression reference sets and practice both directions on one face; week six, reconstruct from skull images with tissue-depth markers; week seven, written two-image comparisons under self-imposed quality limits; week eight, finalize your SOP and run mixed review across all domains. Compress or extend the blocks to fit your schedule, but keep the order, because each block assumes the vocabulary of the previous one.
Score each output 0 to 2: 0 means feature-level guesses only; 1 means correct method with gaps in vocabulary or documentation; 2 means method, vocabulary, and documented limits are all present. Treat a 2 in every domain as a learning milestone — evidence your method is consistent — not a prediction of any exam result. Readiness checks: you can label landmarks on an unfamiliar skull drawing without hesitation; a mock interview transcript shows zero leading prompts; you can state which parts of a reconstruction were skull-driven versus guessed; and your comparison write-up never implies identification.
- Block 1: landmark labeling, proportion measurement, morphology vocabulary drills
- Block 2: recorded mock composite interviews with a leading-prompt audit
- Block 3: child progression, adult progression, and regression on the same face, with justified-versus-speculative notes
- Block 4: skull-based reconstruction with tissue-depth markers and bias-control log
- Block 5: componential image comparisons with region-by-region write-ups
- Block 6: personal SOP for a composite case, checked against ethical obligations
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.