Thompson & Johnson (2008), and Clark, (2005) argued that biased instructions strongly, although indirectly, imply there should be a selection as it fails to inform the witness that the suspect may not be in the line-up, thus leading to higher false-positive identification rates in line-ups where the suspect is not present. In the court of law confidence, not accuracy is thought to be the most significant factor to convince the jury. An eyewitness may be confident and sincere in their decision, but nevertheless entirely incorrect (Cutler & Penrod, 1988). In contrast, other studies show no relationship between confidence and accuracy at all (Loftus, 1979). The literature reviewed for the present study has given rise to the research questions, does participant instructions (group A/B) make a difference on participant response (correct (C) or incorrect (IC)) and, will line-up condition (present (P) or not present (NP)) make a difference on confidence levels given by the participant in terms of Participants Response (C/IC).
Method
Design
Two 2x2 Chi-squares and two 2-way between subjects ANOVA’s were used to analyse the data. The variables examined in the two Chi-squares were participant instructions (with two conditions, group A/B) and participant response (with two conditions, correct (C)/incorrect (IC)). The first Chi-square used data only from the present (P) condition of the Line-up variable and the second used data only from the not present (NP) condition of the Line-up variable. The independent variables for the first between subjects ANOVA were Line-up (P/NP) and Participant Response (C/IC) and the dependent variable was Time (sec) for decision. The independent variables for the second between subjects ANOVA were Line-up (P/NP) and Participant Response (C/IC) and the dependent variable was Confidence (%) in their decision.
Participants
A convenience sample of 397, 187 males and 210 females, was collectively gathered by fourth year Forensic Psychology students. The vast majority of participants (122) were between the ages of 18-22 and 37 or older (101). Each student recruited 12 participants, 6 of which were assigned to the P condition of the Line-up and the other 6 were assigned to the NP condition of the Line-up. All participants chose to take part in the study after being debriefed about the experiment.
Apparatus
Two sets of participant instructions were used, one set for group A and another for group B, and a photo fit picture of the suspect. A set of seven photographs for the line-up all labelled with a three digit number, the first two digits random and the third creating numerical sequence between all seven. Photographs 1-6 were used for the P condition and photographs 1-3 and 5-7 for the NP condition. Photograph number 4 was the actual suspect, included for the P condition and removed for the NP condition. A stopwatch was used to time the participants’ decision and SPSS was used to analyse the class data.
Procedure
The class of students were split into two groups, A and B. Both groups were given instructions that differed slightly. Each student firstly assigned 6 of their 12 participants, individually, to the P condition where they showed the participant the photo fit picture of the suspect to disclose an image of the suspect and told that an eyewitness to a theft had produced this image of the person they believed to have committed the crime.
At this point group A instructed the participant to look carefully at the facial photographs 1-6 out of the seven, all of which were given in random order and labelled with a three digit number; overall the number seemed random to stop participants placing any meaning on the numbers however, the identifying number for the photographs was the last of the three digits giving the experimenter a coded number to identify which photo was chosen. Participants were then asked to look at each photograph; one by one, and decide which of the photographs they thought was the person drawn in the photo fit picture. The participant was told to take their time and go back through the photographs before making their mind up and once they had decided, tell the experimenter their answer. It was also instructed that if they believed none of the six photographs presented was the person in the photo fit, then just to say, none of these photographs. Group B gave the same instructions except they did not include that if they believed none of the six photographs presented was the person in the photo fit, then just to say, none of these photographs.
When the participant looked at the first photograph, the experimenter started the stopwatch and their decision time was recorded however, if two minutes past the participant was informed they had to then make a decision and their time was recorded as 120 sec. The participant was then asked how confident they were (0-100%) about their response and the experimenter completed a results table where details of response (C/IC) time (sec) and confidence in response (0-100%) was recorded along with other details such as age, gender and photo number chosen. The same procedure was carried out for the other 6 of the students 12 participants only this time the participants were assigned to the NP condition and they were given photographs 1-3 and 5-7 in the line up as the actual suspect photo, number 4, was removed. All 397 participants were thanked for their time, reminded that none of the people in the photographs were real suspects in any crime and asked if they had any further questions. On completion of the data collection each student input their data on SPSS where the whole class data set was collated for statistical analysis.
Results
Results of the present study were calculated using 2*2 Chi-Square, two Pearson’s Correlation tests and two, Two-Way Between Subjects ANOVA’s. The mean decision time (sec) for group A was 55.77 and their mean confidence level was 67.52. For group B the mean decision time was 62.77 and their mean confidence was 64.45. Those who chose the C response (photo 4) in the P condition had the shortest mean decision time of 46.51 and the highest mean confidence level of 74.28% in their decision
The Chi-Square was used to test the hypothesis stating that participants in Group A should be more likely than Group B, to respond with the correct answer, as a consequence of the difference between each group’s participant instructions. The 2*2 Chi-Square showed 49% of participants in group A responded C, and 51% IC. The observed count was 88 for the correct response and standardized residual was not significant in group A for both C (z = 1.2) and IC (z = -1.0) responses because both have values smaller than 1.96. The expected count tells us that as many people as was expected (78) responded C and as many as expected (102) responded IC.
In group B, 38% responded correctly and 62% respond incorrectly. The standardized residual in group B was not significant for both C (z = -1.1) and IC (z = .09) responses because both values are smaller than 1.96. The expected count (94) for correct responses and the expected count (123) tells us that as many people as was expected responded C and as many as expected responded IC. Furthermore, the chi-square showed that there was a significant association between participant instructions (Group A or B) in terms of responding correctly and making the right choice, χ 2 (1) = 4.54, p < 0.05. These findings indicate that participant instructions (A/B) had a significant impact on participants in terms of their right or wrong responses.
Those who had the highest mean decision time of 83.07 chose photo 2 and were on average 52% confident. Participants that scored the lowest average in confidence, mean = 47.56, chose photo 5 and took, on average 61.64 seconds to make the decision. From looking at the box plots it’s clear to see that the least confident age group was those under 18. The first Pearson’s Correlation displays a highly significant negative relationship between participant’s decision time, and their reported confidence, when they answer correct, r = -.365, N = 171, p < .001 (2-tailed). There was also a highly significant negative relationship between participant’s decision time and confidence reported, when they answer incorrect, r = -0.336, N = 226, p < 0.001. The results for both correct and incorrect responses show negative correlations, implicating that for both C and IC responses, the faster the participant makes their decision, the more confident they are in that decision, more so in the correct response (see figure 1 & 2).
Figure 1. Scatter plot for decision time (sec) taken and confidence in the C response.
Figure 2. Scatter plot for decision time (sec) taken and confidence in the IC response.
Discussion
The results of the present study support the stated hypothesis that line-up instructions given to an eyewitness can make a difference and influence the decisions made in terms of whether they are correct or incorrect. With regard to the literature reviewed other studies in this area do also support these findings in that instructions have an influence on the eyewitness’s behaviour in terms of inaccuracy of identification (Kassin, Ellsworth & Smith, 1989). It has been suggested that this influence may aid eyewitness’s to choose the incorrect response as because they are not instructed that the criminal may not be in the line-up they assume that it is one of them there-fore choose the person that they think best fits the memory they have of the criminal without truly judging their decision and being extremely confident about it (Wells et al, 1998). An eyewitness may be confident and sincere in their decision, but nevertheless entirely incorrect (Cutler & Penrod, 1995) and in contrast other studies show no relationship between confidence and accuracy at all ( Loftus, 1979). Further research should consider more appropriate and defined instruction for eyewitnesses directly before they see the line-up, making sure they understand clearly what has been said to them and not putting them under pressure.
References
Clark, S. E. (2005). A Re-examination of the Effects of Biased Line-up Instructions in Eyewitness Identification. Law & Human Behavior, 29, 575–604.
Cutler, B. L., & Penrod, S. D. (1988). Improving the Reliability of Eyewitness Identification: Line-up Construction and Presentation. Journal of Applied Psychology, 73, 281–290.
Kassin, S. M., Ellsworth, P. C., & Smith, V. L. (1989). The "General Acceptance" of Psychological Research on Eyewitness Testimony: A Survey of the Experts. American Psychologist, 44, 1089-1098.
Loftus, E. F. (1974). Reconstructing memory: The incredible eyewitness. Psychology Today, 8, 116–119.
Loftus, E. F. (1979). Eyewitness Testimony. Cambridge: Harvard University Press.
Thompson, W. B., & Johnson J. (2008). . The Journal of General Psychology, 135, 23–36.
Wells, G. L., Memon, A., & Penrod, S. D. (2006) “Eyewitness Evidence: Improving Its Probative Value,” Psychological Science in the Public Interest, 7, 45–75.
Wells, G. L., Small, M., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A. E. (1998). Eyewitness identification procedures: Recommendations for lineups and photo spreads. Law & Human Behavior, 22, 603–647.
