Hypotheses
The experimental hypothesis is the leading question will make the participant recall an item that was not there, more often than the participants who were asked the non-leading question.
The null hypothesis is that the leading question will have no effect on the recollection of the picture, and the question will be answered correctly i.e. the object was not present in the picture.
Method
Design
The experiment was conducted as a field experiment due to the fact that Loftus’ study lacked ecological validity as it took place in laboratory settings. Conducting a field experiment for the study therefore increased its ecological validity. The more natural surroundings increased confidence in results and allowed the results to be generalised. Also, cause and effect could be identified because the independent variable (leading/non-leading question) was still manipulated and the dependant variable (whether participant was led or not) measured.
However, extraneous variables were harder to control and therefore it is more difficult to be sure that the answer given for the leading question was due to the question and no other factor. Noise had to be controlled so as not to distract the participant, therefore the participant was asked to work individually and not as part of a group. Knowledge of the subject may also affect the participants’ response. The participant may not know what the question was asking or what the picture was about. Size of the picture may also have affected the response i.e. the size did not allow the participant to observe the picture very carefully.
The experimental design that was used was independent measures, due to the fact that the same questionnaire could not be given to a person twice with the only change being one question i.e. the critical leading question. Using a repeated measures design would have created order effects because the participants would have answered the questions before and therefore have a better chance of answering correctly. Also, the leading question would have no effect on them, and the purpose of the study is to find the leading questions’ effect.
However, participant variables could have affected the results. To minimize this and prevent the participant variables turning into confounding variables, the participants used were all students from St Mary’s College in which they were studying and were between the ages of 16 and 18. Also, they would have needed a certain number of GCSE’s to gain a place at the college.
There were two conditions: the control group were given the questionnaire without the leading question and the experimental group was given the questionnaire with the leading question. The independent variable was the leading/non-leading question and the dependant variable was whether the participants chose the correct answer and was the item that did not exist ‘recalled’.
Due to the fact that it was a field experiment, extraneous variables could have affected the results and therefore distorted them. If the variables had not been controlled then a cause and effect relationship could not be established and the results would not be considered reliable enough to draw a conclusion from. Noise could have prevented the participant from concentrating on memorising what is in the picture. This would have led the participant to answering the questions differently than they would in the absence of the noise. To control this variable, the experiment was conducted in the library on a Wednesday afternoon when most students are outside to minimise disturbance.
Psychology students may have had knowledge about the experiment being conducted. This would have led them to distort the results with demand characteristics because they would act to support the hypothesis or go against the hypothesis intentionally. This variable was controlled by asking the participant after they had agreed to participate in the experiment, whether they were a psychology student. If they were psychology students they were thanked for their time and were told they could not participate.
Working with others could have affected the participants’ answers in the questions due to the variation of opinions and recollections. This may have caused conflict in deciding which answer to choose and therefore the experiment would no longer be leading/non-leading. This was resolved by asking people to look at the picture and complete the questionnaire by themselves and not work in groups.
The experiment used 20 participants. They all read and signed a consent form and read a sheet explaining to them what they had to do. They all looked at a picture of a bowl of fruit for 15 seconds and then completed a leading or a non-leading questionnaire. They were all then given a debriefing sheet to read and given the right to withdraw.
Ethics
Ethical issues had to be dealt with whilst conducting the experiment. For the participant to be able to take part in the experiment, they were given a consent form to read and sign. The consent form gave information about the participants’ right to withdraw from the experiment, and ensured them about the confidentiality of their results. This meant that informed consent was gained where it was possible (consent could not be gained for the fact that the participant may be led by the critical question in the leading questionnaire, because this would mean the hypothesis could not be tested), the participants were told of their right to withdraw at any time from the investigation should they feel the wish to do so, and the participants were told that their results would be unidentifiable.
To be able to test the hypotheses of the experiment and find out whether leading questions can affect the recollection of a memory the participants had to be deceived i.e. by the leading question. Therefore deception in this investigation was unavoidable. To help the participants understand the experiment and its purpose, and to inform them of the presence of unavoidable deception and its necessity, each participant was debriefed after completing the questionnaire. The debrief contained exactly how they may have been led and by which question. They were again given the right to withdraw if they did not agree with the deception. The participants were reminded about the confidentiality of their results, and how the results would be unidentifiable.
Pilot study
A pilot study was conducted to find any problems with the questionnaire and the clarity of the picture. Five people were asked to complete the questionnaire after being the shown the picture. The wording of the leading question was changed from ‘There were kiwis in the fruit bowl. Yes or No’ to ‘How many kiwis were in the fruit bowl? Three or None’. The reason for this change was that the first version of the question was a statement and not a question; therefore it did not allow the participant a choice.
Participants
The target population was students at St Marys College, Shear Brow, Blackburn. This was because the students were the people available at the time and because opportunity sampling was used, they were the target population.
A sample of 20 students was used with the age range of 16 to 18 and the gender ratio was 10 males and 10 females. 10 of these students formed the control group who were asked the non-leading question and the other 10 formed the experimental group who were asked the leading question. They were asked to choose from the two questionnaires (one containing the leading question and the other containing the non-leading question) hence they became part of either the control or experiment group.
The sample was obtained using opportunity sampling. Opportunity sampling is when the participants that are selected are the ones available at that time. This type of sampling was used due to the fact that it was quick and convenient. However, it was harder to get a representative sample and bias may have occurred to choose the most helpful looking participants. This was controlled by choosing one person from each table in the library to participate. This would ensure that a variety of people participated and not people from the same table because they could have similarities which cause bias. Variables of the sample that were controlled were the age group; the gender; the ethnicity due to the fact that St Marys is a multi-ethnic college therefore participants with different ethnic backgrounds were used and finally, the students and their subjects. The students who studied psychology would be able to guess the aim of the investigation and answer accordingly, therefore they could not participate.
Materials and Apparatus
A consent form (Appendix 1) was needed for ethical reasons of getting an informed consent of the participant; give them the right to withdraw at any time and to ensure confidentiality of their results. Paper-based standardised instructions (Appendix 2) related to the aim of the investigation, were used to ensure all the participants received the same instructions, this would not have been possible if the instructions were conveyed to the participants verbally and would have become an extraneous variable. A picture of a fruit bowl (Appendix 3) was used to test the hypotheses, because the participant would need to remember it and recall what it looked like and contained for the experiment. A stopwatch was made use of to time the 15 seconds during which the participant looked at the picture, therefore its use was related to the hypotheses Questionnaires (Appendix 4 and 5) were used containing four questions and one critical leading/non-leading question (e.g. Were there any kiwis in the fruit bowl? Yes or No), for the purpose of testing the hypotheses. The debriefing sheet (Appendix 6) used explained how they might have been deceived in the experiment, how they had the right to withdraw and ensured about confidentiality of their results due to ethical reasons.
Procedure
The students that participated were approached in the library on a Wednesday afternoon. They were asked whether they would like to partake in a psychological investigation. If they agreed to this, they were asked if they studied psychology. If they did study psychology they were thanked for their time and told that they would not be able to participate in the investigation. However, if they did not study psychology, they were a given a consent form to read and sign. After signing the consent form, they received a sheet containing standardised instructions for the experiment. After they had completely read and understood the instructions, they were shown the picture of the fruit bowl for 15 seconds. Then they were given a questionnaire to complete which they chose themselves from a choice of two, therefore allocating them to the control/experimental group. Whether the questionnaire contained the leading or non-leading question depended in which group (control or experimental) the participant was in. Following the completion of the questionnaire by the participant, they were given a debriefing sheet and asked whether they would like to withdraw their results due to the deception that may have occurred to gain the results needed. If the participant withdrew their results were terminated. If they chose to remain participated in the investigation, they were ensured about confidentiality of their results. The results of the questionnaire were recorded using a tally chart. The participants were then thanked for their time and were given the opportunity to ask any questions. This procedure was repeated until the sample was full.
Results
The aim of the experiment was to find out the effect of leading questions on a person’s memory, and whether the question inserted an object into the memory that was not originally there.
It was expected to find that the leading question made the participants recall some kiwis that did not exist, from a picture of a plate of fruit that they were shown, and the non-leading question would be worded so as not to encourage any particular answer.
A table to show the percentage of participants who were either led by recalling some kiwis or un-led recognising there were no kiwis, in the leading question group and non-leading question group.
The table shows that out of all the participants in the investigation, only 10% were led to believe the existence of some kiwis and gave the incorrect answer. These participants were part of the control group who were given the non-leading question.
The pie chart shows that the 90% of the participants answered as anticipated by not recalling any kiwis, therefore answering the question correctly. However, 10% gave the incorrect answer despite the question being non-leading.
The pie chart shows that all of the participants answered the question correctly and did not recall any kiwis. This was not anticipated due to the fact that they were asked the leading question.
The results from the investigation show that the null hypothesis (the leading question will have no effect on the recollection of the plate of fruit, and the question will be answered correctly) is correct and therefore is the hypothesis that is accepted. The experimental hypothesis (the leading question will make the participant recall some kiwis that are not there, more often than the participants who were asked the non-leading question) has to be rejected because no participants who were asked the leading question, were led by it.
Discussion
Due to the investigation, it was found that the leading question that was supposed to make the participant recall some non-existent kiwis, had no effect on the memory recall of the picture shown to the participant. Therefore the null hypothesis which stated that the question would have no effect on memory recall was accepted, and the experimental hypothesis which stated that the style of question would affect the participants’ answer, by making them answer incorrectly, was rejected.
Previous experiments such as the one carried out by Loftus and Palmer (1974) contradict the findings of this investigation. This is because in the study carried out by Loftus and Palmer, the participants were led by the wording of the question. Bartlett’s theory on reconstructive memory (1932) was also found to be inapplicable to the experiment because the participants’ memory was not reconstructed. The participants in this experiment were not affected by the wording of the question and did not recall any kiwis, as was shown by the percentages: only 10% of the participants were led and this 10% was part of the non-leading question group. The fact that all of the leading question group answered correctly and did not recall any kiwis, shows that despite the question being leading due to its wording, it did not have an effect on the answers given.
The findings of the experiment can be generalised to conclude that the students at St Marys College are not affected by the wording of questions. These results can be applied to real life situations such as criminal investigation because they show that the wordings of questions do not affect the memory and how it is answered. Therefore any type of questions can be asked in the participant will not be forced to recall something that did not happen.
However, the experiment did have limitations. The intelligence of the participants varied and due to this uncontrollable factor, each participant’s recollection of the picture was different. This problem can be overcome by using a random and large variety of participants, to provide a larger range of intelligence that can give a more general explanation.
Lack of motivation could have been caused due to the simplicity of the task and its unimportance, leading to lack of interest in the task. This would mean the participant did not answer the questions properly and gave the task their full concentration, which could have led to a distortion in the answers given. Motivation could be increased by issuing a reward on proper completion of the task or making the task more difficult to achieve, therefore the participant would require more concentration.
Extraneous variables may also have affected the participants’ answers. Noise and interference from others could have disturbed and distracted the participants from the task, causing errors in the answers given and making them less reliable.
Replication of the experiment is possible but it would be difficult, due to the amount of noise and distractions that took place. Although the location i.e. the library could be replicated, the level of noise and distraction cannot be replicated. This can be overcome by making the experiment take place in laboratory settings; however this would result in lack of ecological validity. The experiment carried out is ecologically valid because the experiment took place in the library, which was a natural environment of the participants.
The wording of the questions may also have affected the answers given. ‘How many kiwis were in the fruit bowl? Three or None’, due to the fact that the question is implying that there were kiwis in the bowl, and then the choice of answers includes ‘None’, the question is contradicting itself which may have caused confusion to the participant. A solution to this could be that open-ended questions could be used, and therefore not limiting the participants’ answers and giving them the freedom to answer the question as they see fit.
The question asked can be deemed as unusual because some people may not know what a kiwi is or what it looks like. Therefore knowledge of the subject would have affected the answers given. Also, participants may not have paid the picture a lot of attention because of there lacks of interest, therefore they would be indifferent to whether there were any kiwis in the picture or not. A different question could be given in which a more noticeable fruit is mentioned rather than a kiwi.
Although care was taken to ensure sampling bias does not occur, participants who looked more helpful may have been chosen. This could be overcome by using random sampling; however this method is very rare and time-consuming.
The small sample of 20 participants is hard to generalise the findings from. A larger sample than 20 could be used to attain a more general explanation of the results, which show that leading questions do not affect the answers of students at St Marys College. This is because the college contains above 1300 students and 20 participants are not a representative sample.
The task is too artificial in the sense that it is not very serious; resulting in participants less likely to give attention to the answers they gave. This is because they know that the consequences of incorrect answers will not be as serious as it would be in a real life situation.
In a repetition of the experiment, a larger sample could be used obtained via random sampling. There would be a larger range of individuals with different intelligence levels. This would help achieve results that are more reliable and therefore generalised more easily. The experiment would take place in laboratory conditions so that a cause and effect can be established definitively, and extraneous variables such as a noise and distractions from other people would no longer be problems. The two questionnaires would contain all leading or all non-leading questions, so that the results do not depend on the answer of one critical question. Open-ended questions could be used to give the participant more freedom to answer. Also, the subject of the task could be changed to something more serious such as the scene of a car crash, making it more alike to realistic situations, and giving the participant more motivation to give the task more attention, due to the seriousness of the subject.
Further research stimulated from this investigation, could be investigating the differences gender can cause in answering leading and non-leading questions i.e. is one gender more easily led than another? The experiment could be conducted in the same manner, however there would be two groups who are asked the leading questions (one consisting only of males and the other of females) and two groups who are asked the non-leading questions (again one of males and the other of females). The results for the leading questions group for each gender would then be compared, and so would the results for the non-leading questions group. They would be compared using tables and bar charts, comparing the mean number of correct answers for each gender. There would either be no relationship between the two genders, or one gender would be more easily led than the other. This would be apparent from the results.
This investigation was carried out to find the effect of leading and non-leading questions on a person’s recollection of a memory. It was found that there was no relationship between the independent (style of question: leading/non-leading) and dependant (whether participant was led or not) variables and kiwis were not inserted into the memory of a fruit plate, for those people who were asked the leading question.
References
Books:
AS Level Textbook – find out more
The huge one form library
Websites:
www.wikipedia.org
www.skepdic.com/memory
eyewitness.utep.edu/bibliographies
Appendix
Appendix 1 (Consent Form)…………………………………………………………………………………….
Appendix 2 (Instructions)………………………………………………………………………………………
Appendix 3 (Picture)…………………………………………………………………………………………………
Appendix 4 (Non-leading Questionnaire)…………………………………………………………….
Appendix 5 (Leading Questionnaire)…………………………………………………………………….
Appendix 6 (Debrief)……………………………………………………………………………………………….
Appendix 7 (Calculations)………………………………………………………………………………………..
Consent Form
Dear participant,
Please sign this consent form to confirm your participation in this investigation, concerning eyewitness testimony. You have the right to withdraw from the investigation at any time and your information will be kept confidential at all times.
Thank you.
Instructions
- You will receive a picture to study for a 15 second period.
- After the 15 second period, the picture will be removed from sight and you will be asked to choose a questionnaire from a choice of two.
- You will then receive five minutes to complete the five questions in the questionnaire.
- On completion the questionnaire, you will receive a debriefing sheet which will contain more information about the investigation.
- You will be able to ask any questions concerning the investigation.
Thank you.
Questionnaire
Please answer all of the questions.
Please circle your choice of answer.
- Were there leaves for decoration under the fruit?
Yes No
- Were there any bananas?
Yes No
- How many peaches were there?
Two Three
- What colour were the apples?
Red Green
- Were there any kiwis in the fruit bowl?
Yes No
Questionnaire
Please answer all of the questions.
Please circle your choice of answer.
1. Were there leaves for decoration under the fruit?
Yes No
- Were there any bananas?
Yes No
- How many peaches were there?
Two Three
- What colour were the apples?
Red Green
- How many kiwis were in the fruit bowl?
Three None
Debrief
Thank you for participating in the investigation. Unfortunately, you have been deceived slightly, due to the nature of the aim. The investigation was carried out to find the effect of leading and non-leading questions. Due to this, question five about the kiwis may have led you to think there were kiwis in the picture, when in fact there were none. This technique is used in criminal investigation.
You have the right to withdraw your participation in this investigation which will lead to your data being destroyed. If you do not wish to withdraw, your data will be kept confidential and it will not be identifiable.
Please do no hesitate to ask any questions concerning the investigation.
Thank you
Tally chart to show number of people led/un-led in the leading/non-leading group:
Percentage of participants led in the control group:
% = Number of incorrect answers x 100
Total number of answers
= 1 x 100
10
= 10%
Percentage of participants un-led in the control group:
% = Number of correct answers x 100
Total number of answers
= 9 x 100
10
= 90%
Percentage of participants led in the experimental group:
% = Number of incorrect answers x 100
Total number of answers
= 0 x 100
10
= 0%
Percentage of participants un-led in the experimental group:
% = Number of correct answers x 100
Total number of answers
= 10 x 100
10
= 100%