H567/03 Section B Option 2 - Hall and Player (2008) and reducing forensic bias

Aim, participants and experimental design

The aim was to test whether an emotional context influenced fingerprint analysis and decision-making. Seventy volunteer fingerprint experts from the Metropolitan Police took part. Experience ranged from less than three months to over thirty years, with a mean of eleven years. Twelve were qualified experts working in managerial roles; the others were active examiners. These were relevant professional participants, rather than a student sample, but one organisation's volunteers do not represent every examiner or forensic discipline.

The researchers manipulated the accompanying case report. Thirty-five experts received a high-emotion account involving murder; thirty-five received a low-emotion account involving forgery. Random allocation and independent groups were used. The fingerprint material was the same across conditions, so context was the independent variable. Final comparison opinions and reports of whether context affected the expert were dependent measures.

Unlike Raine's naturally occurring groups, emotional condition could be allocated experimentally. The task took place in a familiar examination setting with professional equipment, but the case material was arranged by researchers. A familiar workplace does not make the independent variable naturally occurring.

An ambiguous fingerprint comparison

A volunteer's right forefinger impression was placed on a scanned £50 note, with detail obscured and contrast reduced to create an ambiguous mark. Experts received copies of this mark and a ten-print reference form. They were told to compare the right forefinger, reducing the need to locate an unknown finger. This resembles a verification task more closely than an unrestricted initial search.

They worked using normal aids, including a magnifier and comparator, without a time limit. Their four possible final opinions were: a match; insufficient detail for comparison; sufficient detail for comparison but insufficient for identification; or not a match. They also stated whether they would present the opinion as court evidence and completed feedback about reading the case report and feeling affected by it. Anonymous reference numbers were used.

The procedure separates the physical comparison from the subjective experience of emotion. However, not everyone read the report: 57 of 70 reported reading it. This matters because assigned condition does not guarantee exposure to the intended manipulation. The same ambiguous mark controls physical difficulty, while a single mark gives a narrow test of general performance.

Subjective effect and final opinions

The researchers reported that experts in the high-emotion condition were more likely to say the report affected them. The primary abstract states that this perceived effect did not translate into an effect on the final fingerprint opinion. This distinguishes self-report from actual task outcome. It does not establish that experts were unaffected in every cognitive process or that all contexts are harmless.

The OCR guide gives the following final-opinion counts. Each row totals 35.

The match proportions are $(6/35)\times100=17.14\%$ and $(7/35)\times100=20.00\%$. The low-emotion proportion is 2.86 percentage points higher. This small descriptive difference alone does not establish a statistically significant effect. More importantly, a final opinion is not a direct measure of all stages of attention or comparison.

The conclusion is restricted to this emotional manipulation, these experts and this task. It is possible to acknowledge emotional impact without changing a final decision. Conversely, a person may be unaware of a bias. Self-report and task performance should therefore be measured separately.