Why Witness Confidence Does Not Equal Reliability

The Confidence-Reliability Distinction

One of the most persistent misconceptions in criminal proceedings is the assumption that confident witnesses are reliable witnesses. Research consistently demonstrates that there is little or no correlation between a witness's confidence and the accuracy of their evidence.

This matters because juries and magistrates may be naturally inclined to believe witnesses who present their evidence with certainty and conviction. The law requires tribunals of fact to resist this intuition and assess reliability through objective criteria.

Why Confidence Is a Poor Predictor

Psychological research has identified several reasons why witness confidence does not reliably indicate accuracy:

• Post-event information: Exposure to media reports, discussions with others, or police feedback can increase confidence without improving accuracy
• Rehearsal effects: Witnesses who have repeated their account multiple times often become more confident, regardless of accuracy
• Memory reconstruction: Human memory is reconstructive, not reproductive. Witnesses may fill gaps with plausible but inaccurate details and become confident in those reconstructions
• Confirmation bias: Once committed to an account, witnesses tend to interpret ambiguous information in ways that confirm their existing belief
• Social pressure: The formal court environment and desire to be helpful can increase expressed confidence

Judicial Recognition of the Problem

NSW courts have repeatedly acknowledged the danger of equating confidence with reliability. This recognition appears in both appellate guidance and trial directions.

High Court Authority

In Domican v The Queen (1992) 173 CLR 555, the High Court emphasised that juries should be warned about the danger of relying on witness confidence as an indicator of accuracy, particularly in identification cases. The Court noted that confident but mistaken identification evidence has been responsible for numerous wrongful convictions.

NSW Court of Criminal Appeal

The NSW Court of Criminal Appeal has consistently held that trial judges should direct juries that confidence is not a reliable indicator of accuracy. In appropriate cases, failure to give such a direction may constitute appealable error.

Evidence Act Recognition

Section 165 of the Evidence Act 1995 (NSW) recognises that certain types of evidence may be unreliable and empowers courts to warn juries accordingly. While not specifically addressing confidence, this provision reflects the broader legislative recognition that apparent reliability may be misleading.

Applying the Credibility Matrix

The CORE Defence Credibility Assessment Matrix provides a structured approach to assessing witness reliability that does not depend on confidence:

Identification Evidence: A Case Study

Identification evidence provides the clearest illustration of the confidence-reliability problem. Research demonstrates that eyewitness identification is significantly less reliable than both witnesses and juries typically assume.

Cross-Race Identification

The "cross-race effect" means that witnesses are generally poorer at identifying individuals of a different race. However, confidence levels are not correspondingly reduced. A witness may be highly confident in a cross-race identification that is objectively unreliable.

Stress and Weapon Focus

High-stress situations and the presence of weapons tend to reduce identification accuracy while potentially increasing confidence. The "weapon focus" effect means witnesses may have clear recall of a weapon while having poor recall of the person holding it.

Time and Distance

Brief observation periods and greater distances reduce identification accuracy. However, witnesses may not appreciate these limitations and may express unwarranted confidence.

Practical Implications for Courts

What Courts Should Consider

Rather than relying on witness confidence, courts should assess reliability through:

• Viewing conditions at the time of observation
• The witness's opportunity to observe
• The witness's state of attention at the relevant time
• Whether the witness knew the person previously
• The time elapsed between observation and identification
• Consistency between the original description and the identified person
• The identification procedure used
• Whether the witness has been exposed to contaminating information

Jury Directions

In cases involving identification evidence, trial judges should direct juries that:

• Honest witnesses can be mistaken
• Confidence is not a reliable indicator of accuracy
• Mistaken identification evidence has caused wrongful convictions
• The jury should examine the circumstances of observation and identification
• The jury should look for evidence that supports or undermines the identification

Defence Analysis

When confronting confident witness evidence, defence practitioners should:

• Examine the circumstances that may have inflated confidence (rehearsal, feedback, post-event information)
• Identify objective factors affecting reliability (viewing conditions, stress, attention)
• Highlight inconsistencies between accounts, regardless of confidence
• Consider expert evidence on memory and identification where appropriate
• Request appropriate jury directions on the confidence-reliability distinction

The goal is to shift the tribunal's focus from demeanour and confidence to objective reliability criteria. This requires careful preparation and, often, cross-examination designed to reveal the factors underlying apparently confident evidence.