Key Findings

• The formal standard of proof in Sweden remains “beyond reasonable doubt.”
  This standard applies to all criminal cases and is qualitative rather than probabilistic. No formal numerical certainty threshold exists in Swedish law.

• In practice, many sexual offence cases rely heavily on credibility assessments.
  When physical or technical evidence is absent, courts must often adjudicate based primarily on testimonial evidence and narrative evaluation.

• SVA-inspired credibility reasoning is qualitative and non-falsifiable.
  Criteria such as coherence, detail, and emotional congruence rely on professional judgment and lack fixed thresholds or clear conditions for disconfirmation.

• Research indicates only moderate accuracy for structured credibility methods.
  Studies of Criteria-Based Content Analysis (CBCA), the core of SVA, frequently report classification accuracy in the ~61–70% range under controlled research conditions.

• These accuracy figures represent upper-bound estimates.
  They are typically achieved by trained researchers or forensic psychologists, not by legal practitioners applying informal or implicit credibility judgments in real cases.

• Judges, police officers, and other legal professionals perform near chance in deception detection.
  Empirical research consistently shows that legal experience does not reliably improve lie-detection accuracy beyond chance-level performance.

• This creates a functional shift in the burden of proof.
  While the prosecution formally bears the burden, credibility-centered adjudication in low-corroboration cases can require the accused to supply affirmative counter-evidence or alternative narratives to generate reasonable doubt.

• Methodological caution is therefore essential.
  Credibility assessments should not substitute for corroboration when applying proof beyond reasonable doubt in high-stakes criminal proceedings.

---

Further Reading

Methodology: Statement Validity Assessment (SVA) and Evidentiary Practice