05- Methods & 06- Stature {Forensic Anthropology}

Definition of Science

Structured, testable hypotheses; never “accept” a hypothesis, only fail to reject

Continuous evaluation

Methods change as new data and critiques appear

Ask of any hypothesis

• Accuracy: does estimate correspond to reality?

• Precision: how narrow/useful is the estimated interval?

• Reproducibility: can others replicate results?

• Observer error: low intra- and inter- observer error required

• Population applicability: does reference sample match target populations?

• Practicality: time, cost, and teachability in real-world contexts

Sample

Size and composition of study material

Measures of central tendency

mean, median, mode, maximum likelihood

Dispersion

range, min/max, standard deviation (68%, 95%, 99.7% rules)

Biological profile typical components

age, sex, ancestry, & stature

Purpose of Biological Profile

Narrow population to assist identification before individualizing

Definition of stature

Total living height including soft tissue and vertebral discs

Skeletal contributors (to stature)

Feet (calcaneus, talus), legs (femur, tibia), pelvis, presacral vertebrae, skull

Stature Estimation ~common method

Regression formulas from long bone measurements (y=mx + b)

• Prefer complete long bones (femur best), incomplete bones yield larger error.

• Use population- and sex- specific equations when available

• Report estimate with standard error (range ± SE)

Sources of error and caveats in stature estimation

• Self-reported heights biased (over-reporting common)

• Living height doesn’t = stature @ death (postmortem shrinkage, age-related loss)

• Secular change: modern populations taller than past populations; update formulas

• Reference sample mismatch reduces validity

• Measurement and recording errors

What are the sex (biological) components at birth

gonads (ovaries/testies), external genitalia

What Sex (biological) components isn’t present at birth

Chromosomes is not available on bones

Skeleton shows secondary effects (sexual dimorphism) used for estimation

Scale used: 5-point ordinal (definitely male →definitely female), sometimes effectively six categories due to indeterminate cases

What are the best skeletal regions

pelvis (most reliable) & skull (second-best); long bones and other elements less diagnostic

What is the issue with juvenile sex estimation?

• Sex estimation from skeleton generally unreliable before ~ 15 years old (puberty needed for dimorphic skeletal traits)

• DNA: reliable sex determination (presence/absence of Y) but costly & subject to contamination

How much sexual dimorphism is there in humans?

• Humans have moderate sexual dimorphism; distributions of male & female traits overlap

• Overlap yields an “intermediate” zone where sex cannot be confidently assigned

• Somme populations have greater robustness or different trait distributions; methods must consider population variability

Age effects on sexual dimorphism

• Younger individuals tend to appear more gracile; older individuals can appear more robust

• Bias patterns: young males more likely misclassified as female; older females more likely misclassified as male

Measurement Approaches for Sex Estimation

• Visual (qualitative) assessment

• Metric (quantitative) assessment

Pros & Cons of visual (qualitative) assessments

• Pros: fast, can integrate subtle shapes cues, effective with experience (typical accuracy with experienced observers and appropriate bones ~ 80-90%)

• Cons: subjective, relies on observer experience: harder to quantify uncertainty

Pros and Cons of metric (quanitative) assessments

• Pros: reduces subjectivity, repeatable, provides probabilities/likelihoods

• Cons: depends on clear landmark definitions: some landmarks are ambiguous

• Methods include single-dimension measures, ratio (indices), and multivariate analyse (e.g., discriminant functions, forensic software)

Combination of visual & metric assessment

• multiple measurements usually better than single measures

• multivariant models (computerized) can handle many variables & output probabilities

Sectioning points (simple cut-offs):

Easy to use but ignore uncertainty & preform poorly in intermediate ranges. Don’t provide a probability or measure of confidence

• Example index: Ischiopubic index compares pelvic widths as a ratio; has defined male/female cut-offs & an intermediate zone

Measurement Tools

• Sliding calipers, spreading calipers, osteometric board, craniometer (craniometric frame) for standardized skull rotation

Landmark issues

• Some landmarks (femoral head, humeral head) easier & less error-prone than acetabular/ pubic points

Practicalities

• Training reduces intra - and inter - observer error; professionals outperform inexperienced groups

Biases &Contextual Factors: Age Estimation

• Prior information or contextual (e.g., monastic, cemetery, war grave) changes prior probabilities & alters cutoffs/interpretations

• Forensic samples differ from natural mortality samples (age discrimination, cause-of-death biases)

• Population-specific morphology matters; methods developed on one group may fail on another