Methodology

Analytical Methods

Forensic Disciplines

Established analytical methods applied to the documentary record. Each method produces documented, citable findings with full methodology transparency.

Cross-Document Contradiction Analysis

Systematic comparison of claims, figures, and assertions across pleadings, correspondence, contracts, and witness statements. Each assertion indexed and cross-referenced against the remainder of the documentary record.

Timeline Forensics

Chronological reconstruction of documented events, communications, and claimed occurrences. Identifies sequence inconsistencies, temporal conflicts, and periods where the documentary record is silent.

Evidence Gap Analysis

Identification of documentation that would be expected to exist if a given claim were accurate. Maps the relationship between assertions made and the documentary material produced in support.

Statistical Analysis

Quantitative examination of financial data applying Benford's Law digit frequency analysis, outlier detection, and distribution testing. Produces conformity classifications with statistical significance measures.

Forensic Stylometry

Computational authorship analysis using Burrows' Delta and function word frequency distributions. Produces quantitative similarity scores with documented methodology and known error rates.

Document Metadata Examination

Analysis of embedded file metadata including creation dates, modification history, authorship fields, and software identifiers. Compares metadata positions against the dates and authorship represented on the face of each document.

Email Header Forensics

Examination of email transmission headers, routing data, and server timestamps. Verifies the authenticity and transmission path of email correspondence against the positions advanced by the parties.

Image Forensics

Error Level Analysis, metadata extraction, and compression artefact examination of photographic and documentary image evidence. Identifies regions of an image that are inconsistent with a single-capture origin.

Claim Evolution Analysis

Tracks how specific assertions change across successive statements, pleadings, and correspondence. Documents the progression of each claim from its first appearance to its final form.

Beneficiary Analysis

Maps the financial and positional consequences of each disputed claim. Identifies which party benefits from each variance, omission, or inconsistency in the documentary record.

Financial Pattern Analysis

Examination of transaction patterns, rounding behaviour, threshold clustering, and duplicate entries across financial records. Each pattern tested against documented forensic accounting standards.

Cross-Reference Verification

Multi-source corroboration testing across the documentary record. Examines whether assertions in one document are supported, contradicted, or unaddressed by the remaining material.

Evidence Classification

The Ranking System

Findings classified by evidential weight. Each level reflects the degree of documentary support and the availability of alternative explanation.

Level	Classification	Criteria	Example
4	Binary	Direct contradiction established by documentary evidence. No alternative explanation advanced or available on the material reviewed.	Claimant states under oath event occurred 10 June. Email from claimant dated 8 June references event as "last week."
3	Material	Inconsistency for which an alternative explanation exists but is not supported by the documentary record. Material to the issues in dispute.	Same item claimed at materially different amounts across multiple documents with no explanation for variance.
2	Strong	An alternative explanation is available but is not fully consistent with the surrounding documentary record. Supports wider patterns when considered alongside other findings.	Invoice dated after claimed payment date. Possible explanation: invoice date error. Supports larger timeline analysis.
1	Supporting	Contributes to the broader documentary position. An alternative explanation is available. Relevant when considered in conjunction with findings at higher levels.	Peripheral inconsistency for which an explanation may exist. Documented for completeness and for the weight it may add to the overall evidential picture.

Classification in Practice

Exhaustive analysis of a documentary record produces volume. Classification produces focus. Findings are separated by evidential weight so that those instructing may direct attention to the material of greatest significance.

The case summary contains Level 4 and Level 3 findings only. The full report documents the complete evidential position across all levels.

Quantitative Analysis

Statistical Methods

Court-admissible quantitative methods producing documented confidence levels with academic citations and full methodology transparency.

Benford's Law Analysis Suite

Digit frequency analysis applying established forensic accounting methodology. Naturally occurring financial data follows predictable mathematical distributions first documented by Newcomb (1881) and formalised by Benford (1938). Data that has been estimated, rounded, or constructed from sources other than genuine transactions tends to deviate from these distributions.

First Digit Test (D1) - Primary conformity indicator; expected distribution: 30.1% begin with 1, decreasing logarithmically to 4.6% for 9
Second Digit Test (D2) - Identifies subtler deviations; expected range 11.97% (digit 0) to 8.50% (digit 9)
First-Two Digits Test (D1D2) - 90 possible combinations; higher sensitivity to constructed data; identifies threshold avoidance patterns
First-Three Digits Test (D1D2D3) - 899 combinations; isolates specific anomalous values in large datasets
Last-Two Digits Test - Uniform distribution expected; identifies rounding patterns and digit preference

Statistical Significance Testing

Independent statistical tests quantify the probability that observed digit distributions arose from natural transaction data. Results expressed as confidence levels with full methodology documentation.

Chi-Square Test - Goodness-of-fit test comparing observed against expected frequencies; p-value quantifies deviation significance
Z-Test (Per-Digit) - Tests individual digit deviation; identifies which specific values depart from expected frequency
Mean Absolute Deviation (MAD) - Nigrini's preferred metric; thresholds: Close (<0.006), Acceptable (<0.012), Marginal (<0.015), Nonconformity (>0.015)
Distortion Factor - Measures systematic directional bias; positive values indicate upward tendency, negative values indicate downward tendency
Kolmogorov-Smirnov Test - Non-parametric distribution comparison; detects subtle deviations across entire dataset
Anderson-Darling Test - Enhanced sensitivity at distribution tails; identifies anomalous patterns in extreme values

Conformity Classification

Results classified per Dr. Mark Nigrini's established forensic accounting standards. Each dataset receives a conformity rating with supporting statistical evidence.

Close Conformity - MAD < 0.006: Data consistent with natural occurrence; no statistical indicators of deviation
Acceptable Conformity - MAD 0.006-0.012: Minor deviations within normal variation; insufficient for adverse inference
Marginal Conformity - MAD 0.012-0.015: Borderline results; warrants further investigation of flagged transactions
Nonconformity - MAD > 0.015: Statistically significant deviation from expected distribution; warrants further examination of the underlying data

Outlier Detection Methods

Multi-method anomaly identification using complementary statistical approaches. Flagged values cross-referenced against the documentary record.

Z-Score Analysis - Standard deviation method; flags values exceeding 2.5-3.0 standard deviations from mean
Modified Z-Score (MAD-Based) - Robust to outliers in source data; uses median absolute deviation for baseline
Interquartile Range (IQR) - Distribution-agnostic method; flags values below Q1-1.5*IQR or above Q3+1.5*IQR
Composite Anomaly Scoring - Weighted combination of methods; produces per-transaction anomaly score for further examination

Anomaly Pattern Detection

Analysis for common anomaly patterns documented in forensic accounting literature. Each pattern tested independently with statistical significance assessment.

Round Number Bias - Clustering at round values (00, 50, 000); consistent with estimation rather than documented transactions
Threshold Avoidance - Clustering below approval limits, reporting thresholds, or tax boundaries (e.g., values at £9,900–£9,999 where a £10,000 threshold applies)
Duplicate Detection - Exact and near-duplicate identification across dates, amounts, and descriptions within the documentary record
Sequence Analysis - Invoice/reference number gaps, arithmetic progressions, and pattern detection in sequential identifiers
VAT Calculation Verification - Mathematical testing of claimed tax calculations; identifies computational errors and false tax figures
Temporal Pattern Analysis - Transaction timing distribution; flags unusual clustering in dates, times, or reporting periods

Output: Quantified Results

Each statistical analysis produces documented results with full methodology transparency, suitable for use in proceedings.

Conformity Score - Overall dataset rating with confidence interval and supporting test results
Flagged Transactions - Itemised list of anomalous entries with individual anomaly scores and reasons for inclusion
Visualisation Package - Digit distribution charts, deviation graphs, and comparison exhibits prepared for use in proceedings
Methodology Documentation - Complete description of tests applied, thresholds used, and academic citations for each method
Limitation Statement - Clear disclosure of sample size constraints, test applicability, and confidence boundaries

Quality Standards

Documentation Standards

Evidentiary standards applied to all analysis and deliverables.

Source Citation

Every finding includes complete source citations: document name, page reference, paragraph number, and date. All assertions traceable to original documentation.

Methodology Documentation

Analytical methods documented in sufficient detail for independent verification. Each finding includes the method by which it was reached.

Confidence Assessment

Findings classified by evidential weight based on documentary corroboration, source reliability, and the availability of alternative explanation. Distinction maintained between documented facts and analytical observations.

Limitation Disclosure

Analysis scope and limitations documented. Gaps in the documentary record identified. Conclusions qualified where evidence is incomplete or ambiguous.

Output

Deliverables

Evidence packages structured for use in proceedings, prepared to the standard required by those instructing.

Standard Deliverables

Executive Summary - Overview of principal findings with evidential classification and areas warranting further inquiry
Detailed Analysis Report - Complete findings with methodology documentation, source citations, and supporting analysis
Inconsistency Schedule - Indexed catalogue of identified inconsistencies with citations, classification levels, and cross-references
Timeline Documentation - Chronological reconstruction with source documentation and identified sequence issues
Statistical Analysis Report - Quantitative findings with methodology documentation, significance testing, and conformity classification
Questions Arising from the Evidential Record - Documented inconsistencies organised by subject matter with source citations
Exhibit Package - Visual materials including timelines, comparison charts, and summary tables prepared for use in proceedings
Source Index - Complete index of documents examined with citation key for reference throughout proceedings