Studies in Detail
- 4. Studies in Detail
(a) Describe and Evaluate Loftus & Palmer’s (1974) study of the effect of leading questions on estimates of speed and one other study. This must be selected from the following:
- Yuille & Cutshall (1986) real-life case study of leading questions and eye witnesses
- Charlton et al (2000) naturalistic experiment in St Helena
- Gesch et al (2003) Influence of supplementary vitamins, minerals and essential fatty acids on the anti-social behaviour of young adult prisoners
Study In Detail 1 = Loftus & Palmer (1974) A Laboratory Experiment.
Loftus & Palmer’s research consisted of two separate experiments – experiment one and experiment two.
Experiment one
The AIM of this study was to see if the phrasing of a question would affect estimates of speed, applying these findings to the idea of leading questions in court. A leading question is one that suggests to the witness what answer is desired or leads him or her to the desired answer.
The PROCEDURE involved 45 students being put into 5 different groups. Seven films were shown, each involving a traffic accident, with each film lasting between 5 and 30 seconds. After every film the participants had a questionnaire to fill in. First they were required to give an account of the accident, and then had to answer specific questions. The critical question was the one asking about the speed of the vehicles. The first group of nine participants were asked ‘about how fast were the cars going when they smashed into each other?’ and the other groups were asked the same critical question but with the word smashed replaced by collided, bumped, hit or contacted. The same procedure was followed for each film. This study is an example of a laboratory experiment because the procedure is controlled for all participants and the IV (which is the verb in the critical question) is deliberately manipulated whilst trying to keep all other variables constant. The effect of the IV is then measured on the DV which in this case is the estimate of speed in mph given by participants.
The RESULTSshowed that the verb ‘smashed’ produced the highest mean estimate of speed and that the verb ‘contacted’ produced the lowest mean estimate.
Table to show estimated speeds according to which verb was used in the question.
| Verb used | Mean speed estimated (mph) |
| Smashed | 40.8 |
| Collided | 39.3 |
| Bumped | 38.1 |
| Hit | 34.0 |
| Contacted | 31.8 |
Collided, bumped and hit had speed estimates that were fairly well spaced and decreased according to the perceived severity of the verb.
The CONCLUSION was that the wording of a question can affect a witness’s answer. Loftus and Palmer gave two interpretations/explanations for the findings of their 1st experiment.
1. Firstly, they argued that the results could be due to a distortion in the memory of the participant. The memory of how fast the cars were travelling could have been distorted by the verbal label which had been used to characterise the intensity of the crash.
2. Secondly, they argue that the results could be due to response-bias factors, in which case the participant is not sure of the exact speed and therefore adjusts his or her estimate to fit in with the expectations of the questioner. (This is also an example of a demand characteristic)
Evaluation of Loftus & Palmer’s Study.
Strengths
- Reliability – this study was well controlled meaning that the study is replicable which increases the reliability of the findings. For example, all participants watched the same film and were asked identical questions except for one word change (the IV).
- Objective data – by using estimates of speed, the researchers gathered quantitative data and therefore no interpretation of the data was needed making their study more objective. It also allows the researchers to carry out statistical analysis of the data to determine if the results are significant and furthermore, comparisons can be made by drawing up charts and graphs of the results.
- Less demand characteristics – the aim of the study was disguised by ‘hiding’ the leading question amongst others. This meant the participants would be less likely to guess the aim of the study therefore helping to reduce demand characteristics (behaving in a way that attempts to meet the demands of the situation).
Weaknesses
- Ecological validity – this study is low in ecological validity because the experiment used a situation which was not like real life. The participants would not experience the same emotional strain whilst watching a film clip as they would if they had been witness to a real car accident. Therefore the results may not be generalisable to real life eye witness accounts. However, although the task lacked mundane realism in many ways; by having week long gap between seeing the ‘incident’ and recalling it made the situation more realistic, as witnesses are unlikely to record what they have seen immediately.
- Sample – using students as participants it might be said that the findings are not generalizable to the whole population. Students’ memories may be different to other age groups and people with different occupations. Furthermore, the sample tested was relatively small so again may not be representative of all eye witnesses.
- Validity – the validity of the situation the participants were put into is questionable. The experimenters claimed to be measuring the effect of leading questions on witness recall however the task used to test this was memory of a film clip. Eyewitnesses to actual crimes do not see them on a video but as an on-going part of day-to-day events. Participants in the experiment may pay more attention than a witness would and have fewer distractions. Also, participants in the experiment are likely to be less motivated and aroused than eyewitnesses seeing a real car accident and their reactions may not be representative, especially as emotional state is known to affect recall.
Experiment two
The AIM of this experiment was again to determine whether leading questions would influence the response of an eyewitness.
A similar PROCEDURE to the first experiment was used whereby 150 student participants viewed a short (one minute) film which contained a 4 second scene of a multiple car accident, and were then questioned about it. There were three conditions and the independent variable was manipulated by the wording of the question.
50 of the participants were asked ‘how fast were the cars going when they hit each other?’
50 of the participants were asked ‘How fast were the cars going when they smashed into each other?’
50 of the participants were not interrogated about the speed of the vehicles.
One week later, the participants returned and, without viewing the film again, they answered a series of questions about the accident. The critical question was ‘Did you see any broken glass?’ The critical question was part of a longer series of questions and was placed in a random position on each participants question paper. There was in fact no broken glass in the film.
The RESULTS collected in response to the question ‘Did you see any broken glass?’ are noted in the table below:
| Response | Smashed | Hit | Control |
| Yes | 16 | 7 | 6 |
| No | 34 | 43 | 44 |
These results show a significant effect of the verb in the question on the mis-perception of glass in the film. Those participants that heard the word smashed were more than twice as likely to recall seeing broken glass.
The CONCLUSION outlined tries to account for the results of the second experiment. Loftus and Palmer developed the following explanation called the reconstructive hypothesis:
They argue that two kinds of information go into a person’s memory of an event. The first is the information obtained from perceiving an event (e.g. witnessing a video of a car accident), and the second is the other information supplied to us after the event (e.g. the question containing hit or smashed). Over time, the information from these two sources may be integrated in such a way that we are unable to tell from which source some specific detail is recalled. All we have is one ‘memory’.
For example in Loftus and Palmer’s 2nd experiment, the participants first form some memory of the video they have witnessed. The experimenter then, while asking, “About how fast were the cars going when they smashed into each other?” supplies a piece of external information, namely, that the cars did indeed smash into each other. When these two pieces of information are integrated, the participant has a memory of an accident that was more severe than in fact it was. Since broken glass corresponds to a severe accident, the participant is more likely to think that broken glass was present.
The EVALUATION of this second experiment would be similar to experiment one.
Study In Detail 2 = Yuille & Cutshall (1986) A Field Study/Case Study.
A gun shop in Canada was robbed in broad daylight. As the thief was escaping, the shop owner followed him on to the street. Both had guns. The thief fired twice and the shop owner fired 6 times, the thief was killed.
There were many witnesses to the event and all had a clear view so the police were able to get a detailed account because the witnesses could corroborate each other’s accounts which could then be linked to forensic evidence about the event.
The AIM of the study was to look at the problems of laboratory research in studying eyewitness testimony, to look at the accuracy of eyewitness accounts and to compare eyewitness accounts taken straight after an incident with those taken four to five months after. Another aim was to see how eyewitness memory could be affected by leading questions.
The PROCEDURE involved asking 21 witnesses to a gun shooting which had taken place four to five months before to take part in research. 13 agreed to take part in the research. Directly after the gun shooting, all 21 witnesses were interviewed by police. Four to five months after the initial interviews, the 13 participants were interviewed by researchers. The interviews followed the same style as the police interviews: the participants were asked to describe the event and were then asked questions about it. Two leading questions were added to the interviews, one asking about the thief’s car headlight (“did you see the broken headlight?”) and another asking about the car quarter panel (“did you see the yellow quarter panel?”). The headlight was not broken, and the quarter panel was blue. Half of the participants were asked these leading questions, and the other half were asked non-leading variants (“did you see a broken headlight” and “did you see a yellow quarter panel?”). The participants were asked about their emotional state and stress levels before and after the incident, and were also asked whether they had any problems afterwards. These results were recorded on a 7 point scale – 1 being perfectly calm and 7 being extremely anxious.
The scoring procedure was very precise, so the police and research interviews could be compared. The content of the witnesses’ reports were scored for the number of accurate action, person and object descriptions.
Action descriptions related to things such as movements or gunshots. Person descriptions related to things such as age, height, hair colour and clothing. Object descriptions related to items in the crime such as money or the gun.
Table showing the accuracy of person descriptions.
| Detail | Police interview | Research interview |
| Hair colour | 77 | 80 |
| Clothing colour | 66 | 59 |
| Weight, height, age | 50 | 51 |
| Details excluding weight, height, age | 82 | 80 |
The RESULTS showed that the researchers recorded more total details than the police: the police recorded 649 total details and the researchers recorded 1056 total details. The researchers recorded half action, half descriptive details, whereas the police recorded more action details than descriptive details. This was largely due to the fact that the researchers asked more about descriptive details than the police.
The witnesses were split into peripheral and central witnesses, and this accounted for the variation in what was reported. 7 of the witnesses were central, and 6 were peripheral. Both of these groups were equally accurate in their accounts, with 84% of central witnesses were accurate, and 79% of peripheral witnesses were accurate. Leading questions did not affect the witness’s memories of the incident.
The CONCLUSION was that eyewitness statement is remarkably accurate – more accurate than lab experiments would suggest. Central witnesses were more accurate in their accounts, although both groups were fairly equal. This is something else a lab experiment cannot capture.
Evaluation of Yuille & Cutshall’s Study
Strengths
- Ecological validity – this study possesses a very high level of ecological validity. This is because it followed a field study design, and assessed the testimony of real witnesses to a real crime, instead of assessing participants in an artificial situation.
- Reliability – the findings are reliable. This is because of the precise scoring procedure. If this procedure were to be repeated, then the study is likely to produce the same results. Also, the data produced were quantitative, and so researcher bias is limited and is unlikely to affect the findings.
- Challenge to laboratory findings – the accuracy of Yuille & Cutshall’s witnesses and the lack of effect that the leading questions had upon recall supports the criticism that laboratory studies lack ecological validity since the findings of such studies appear not to generalise to witnesses of real crimes.
Weaknesses
- Sample – it may not be generalisable, and it would be unfair to criticise lab experiments with the evidence from this study. This is because the study assessed real witnesses from a very small sample of people. The researchers themselves said they were investigating flashbulb memory, which is completely different from what is studied in the lab.
- Subjectivity – the scoring system was sometimes inaccurate. This means that the results may be inaccurate. For example, the statement “He looks as if he was in his early twenties” was recorded as incorrect as the thief was actually 35 – regardless of whether the thief did look in his twenties. The type of scoring used converted qualitative data into quantitative data, and so a degree of subjectivity may also be present.
- Validity of the task – although the researchers aimed to measure eyewitness memory, they acknowledge that they may in fact have been testing flashbulb memory instead. Flashbulb memory is particularly vivid and accurate memories associated with key emotional events.
Study in Detail 3 = Charlton et al (2000) – The St. Helena study
A recent study was in St. Helena, a British Colony in the South
Atlantic Ocean, which received television for the first time in
1995.
Despite expectations that the introduction of television would produce an increase in anti-social behaviour, the researchers concluded that very little changed following television’s arrival.
The AIM of this study was to investigate the effects of television on children’s behaviour, in particular whether TV affected pro-social and anti-social behaviour.
The PROCEDURE involved studying the playground behaviour of 3-8 yr. old children before and after the introduction of TV. This study is a quasi/naturalistic experiment because there is a clear IV and DV but the IV (Before TV and after TV) occurred naturally as a result of the island introducing TV. The IV was not deliberately manipulated by the researchers, it occurred naturally. To collect data video recorders were set up in 2 schools 4 months prior to the introduction of TV in Mar 95 and again 5 years later. The children’s free play during break times was recorded for a 2 week period both in 1995 and again 5 years later. Findings were compared to establish if behaviour had changed.
Eight playground behaviours were categorised as follows:
|
Pro Social |
Anti-social |
|
|
Independent researchers in the UK watched the video footage and tallied how many times the children or a group of children displayed these behaviours. Two researchers watched the same video footage and only agreed tallied behaviours were recorded in the results. This is called inter-rater reliability.
The RESULTS were based upon 64 comparisons made between the behaviour of children in 1995 and 2000. Only 9 of the results were statistically significant:
- 2 Showed decreases in anti-social behaviour amongst boys
- 5 showed increases in pro-social behaviour in boys and girls
- 2 showed decreases in pro-social behaviour in boys
- Boys displayed less hitting and pushing after TV was introduced
- Boys less willing to help or show affection
- Both boys and girls showed significant increases in pro-social behaviour overall.
In CONCLUSION the introduction of TV had no negative effect on children’s behaviour (quite the opposite) this contradicts most laboratory research.
Evaluation of Charlton’s Study
Strengths
- Reliability – the use of videos means recorded behaviours can be checked and verified by independent observers and eliminates subjective interpretation and bias. This makes the research more objective and scientific. However, it is difficult to say if a change in behaviour was an effect of watching TV due to a lack of control over extraneous variables.
- Ecological Validity – very good because the study looks at how actual TV programmes have an effect on real culture which has never experienced TV before. Children were observed in their natural setting of a playground where they would normally engage in social behaviour and are not constrained by rules.
- Ethics – ethically this was a suitable way to investigate the role of the media on young children’s behaviour since it did not involve any direct manipulation by the researchers since the variables were naturally occurring.
Weaknesses
- Generalisability - St Helens is a close-knit community with a culture of close supervision and parental control meaning it is more difficult to generalise the results to more urbanised and modern cultures that we live in since the children are not a representative sample of all cultures.
- Validity – was low because the television viewed was not similar to mainland TV programmes. Popular children’s programmes involving aggression such as Teenage Mutant Ninja Turtles and Power Rangers were not broadcast.