Valentine & Mesout (2009) Eyewitness Testimony (Edexcel A-Level Psychology): Revision Notes
Valentine & Mesout (2009) Eyewitness Testimony
This field experiment investigated the impact of stress on eyewitness memory and identification accuracy. Tim Valentine and Jan Mesout conducted research at the London Dungeon to examine eyewitness testimony in a real-life scenario, addressing limitations of laboratory-based research. The study tested whether high stress levels would impair witnesses' ability to recall and identify a perpetrator.
Key Concepts:
- State anxiety - temporary anxiety experienced in a specific situation
- Trait anxiety - general tendency to experience anxiety across different situations
- Field experiment - research conducted in natural, real-world settings rather than laboratory conditions
Participants
56 visitors to the London Dungeon completed the study:
- 29 female participants
- 27 male participants
Participants received discounted admission in exchange for completing questionnaires and wearing heart rate monitors during their visit.
Aim
To investigate whether high levels of stress (high arousal) reduce eyewitnesses' ability to recall information and accurately identify a perpetrator.
Procedure
Setting and stress manipulation:
- Participants toured the London Dungeon whilst wearing heart rate monitors
- An actor dressed in a dark robe with pale makeup, wounds and facial scars acted as the 'scary person'
- This actor blocked participants' path during the tour, preventing them from passing
- All participants experienced the same tour route with the same actor using identical behaviour
Experimental Control: Using the same actor, same route, and same behaviour for all participants ensured that any differences in memory performance could be attributed to individual stress responses rather than variations in the experience itself.
Post-tour assessment:
- After completing the tour, researchers explained the study's purpose and obtained informed consent
- Participants could withdraw at any stage
- Questionnaires measured state anxiety (temporary anxiety during the dungeon experience) and trait anxiety (general anxiety levels)
Memory testing:
- Free recall task: Participants provided an open-ended description of the 'scary person' without prompting
- Cued recall task: Participants answered specific questions about the person's details, advised not to guess if uncertain
Identification task:
- Participants viewed a nine-person photo line-up
- The 'scary person' was placed in a randomly selected position
- Participants indicated whether the person they encountered was in the line-up
- They rated their confidence in their decision on a scale from 0-100%
Findings
Anxiety levels:
- Females reported higher state anxiety (mean = 52.8) compared to males (mean = 45.3)
- Overall mean state anxiety score was 49.0
- No difference between males and females in trait anxiety levels
Recall and identification accuracy:
- Participants with lower state anxiety recalled more accurate information about the scary person
- Only 17% of participants scoring above the median on state anxiety (52 or above) correctly identified the person from the photo line-up
- In contrast, 75% of participants scoring below the median correctly identified the 'culprit'
- Higher state anxiety was associated with decreased likelihood of correct identification
- Males made more accurate identifications than females
- Participants who correctly identified the person showed higher confidence levels in their decision
Key Finding: There was a negative correlation between state anxiety levels and identification accuracy - as anxiety increased, accuracy decreased. This represents a dramatic difference: participants with low anxiety were more than 4 times more likely to correctly identify the perpetrator compared to those with high anxiety.
Evaluation: Strengths
High ecological validity:
The study was conducted as a field experiment in a natural setting. Whilst being scared by an actor in a ghoulish outfit is not a typical daily experience, the natural environment and subsequent findings are more representative of real eyewitness situations than laboratory-based research. Participants were unaware their testimony would be tested, reducing demand characteristics.
Ecological Validity Advantage: Because participants didn't know they would need to remember the scary person, their memory encoding was natural and spontaneous, more closely mirroring how witnesses to real crimes process information. This increases the applicability of findings to real-world eyewitness situations.
Strong experimental controls:
Excellent controls were maintained throughout the study. The same actor performed in the same manner on the same tour for all participants. This standardisation provides high replicability - the study could be conducted again in a similar setting with comparable results. These controls allow the findings to be extrapolated to other situations involving eyewitness identification under stress.
Validated measurement tools:
The researchers validated the anxiety questionnaires by testing them on office workers. This validation process ensures the questionnaires reliably measure anxiety rather than other emotions, strengthening confidence in the study's findings about the relationship between anxiety and identification accuracy.
Methodological Rigour: Pre-validating the questionnaires demonstrates scientific thoroughness and ensures that the measurements truly reflect anxiety levels rather than confusion, excitement, or other emotional states that participants might experience in the dungeon.
Individual differences considered:
The research accounted for individual differences amongst participants by measuring trait anxiety. This provided a baseline against which state anxiety could be compared, allowing researchers to determine whether participants were experiencing genuine stress from the situation rather than simply being generally anxious individuals.
Evaluation: Weaknesses
Generalisation limitations due to self-selection:
All participants chose to attend the London Dungeon, suggesting they may have a preference for scary entertainment. These individuals might be affected by frightening events differently from people who do not choose such activities. Their reactions to the scary person may not reflect how the general population would respond in genuinely stressful situations, limiting the generalisability of findings to the wider population.
Self-Selection Bias: People who voluntarily visit scary attractions may possess personality traits (such as sensation-seeking or higher stress tolerance) that make them unrepresentative of typical crime witnesses. This raises questions about whether the 17% vs 75% identification accuracy difference would be the same in the general population.
Atypical nature of the experience:
Whilst the study has higher ecological validity than laboratory research, being confronted by an actor in dark robes with makeup and artificial wounds is not representative of typical crime scenarios. Real crimes involve genuine threat and danger, which may produce different physiological and psychological responses compared to the controlled 'scary' experience in the dungeon.
Potential for participant bias:
Because participants self-selected to attend a scary attraction, they may have approached the experience with different expectations and coping strategies compared to genuine crime witnesses. This pre-existing mindset could influence their memory encoding and subsequent recall, making it difficult to generalise findings to unwilling witnesses of actual crimes.
Context Matters: Participants at the London Dungeon expected to be scared and knew they were in a safe, controlled environment. In contrast, real crime witnesses experience unexpected trauma without the psychological safety net of knowing it's entertainment. This fundamental difference may affect how stress impacts memory formation and retrieval.
Remember!
Key Takeaways from Valentine & Mesout (2009):
-
High stress impairs eyewitness accuracy: Participants with higher state anxiety were significantly less likely to correctly identify the scary person (only 17% correct vs 75% for low anxiety).
-
Gender differences in anxiety and accuracy: Females experienced higher state anxiety in the dungeon (52.8 vs 45.3) and males made more correct identifications overall.
-
Field experiments enhance ecological validity: Using the London Dungeon provided a more realistic setting than laboratory studies, though self-selection limits generalisability.
-
Individual differences matter: Measuring both trait and state anxiety allowed researchers to distinguish between general anxiety levels and situation-specific stress.
-
Practical implications: The findings suggest experts should consider witnesses' emotional state when assessing the reliability of eyewitness testimony.
-
Negative correlation: There is a clear negative relationship between stress and identification accuracy - as anxiety increases, the ability to accurately recall and identify perpetrators decreases.