Loftus and Palmer 1974 Experiment Replica
Loftus and Palmer 1974 Experiment Replica
Loftus and Palmer 1974 Experiment Replica
Tyler Rave
IB Psychology SL
02/04/11
Word Count: 2,277 (1,557) -----EDIT EDIT EDIT **************** need to cut 57
Abstract
In this experiment we are studying the effect of verb manipulation on cognitive memory retrieval. We
showed students a simple video of a car crash after explaining they would be participating in a short
term memory study. After handing out questionnaires (See appendix 4) to all the participants with three
different variations of the last question, the students filled in their observations unknowing of the
modifications of the last question. After the questionnaires were collected the students were informed
of the true aim of the experiment (See appendix 2). The study was done in a ninth grade regions history
classroom near the end of the schooldays. Our results turned out to be fully in line with the original
experiment conducted. In conclusion we’ve found that it is possible to manipulate memory retrieval
through the use of wording and form of a question due to involuntary cognition and schema.
Introduction
Cognitive psychology, or cognition, is the study of how different people involuntarily perceive,
remember, think, speak, and solve problems. Cognitive psychologists argue that the theory of the mind
can be studied, however not directly. Because the mind cannot be studied physically, it can only be
studied by taking back roads and indirect routes to observe the mind in action. To ensure laboratory
experiments on the mind are valid, deception is often used to help keep subjects behaving naturally in
the field being studied. Another argument of the perspective is that the mind is like a computer in that it
takes in information, and then proceeds to process it in its own way. Memory for example, is the
recollection of certain events and details. Elizabeth Loftus and John Palmer conducted an experiment to
show how memory can be altered by taking advantage of involuntary actions in the mind like association
and classical conditioning. Loftus and Palmer took a group of 150 people and showed a video of a staged
car accident. Multiple cars were included and the video was only shown once. Participants were then
questioned about the speed of the cars in the video using five different verbs. The results of the original
experiment showed that participants with higher verb intensity in their question were more likely to
estimate a higher speed. The aim of our study is to conduct an experiment to test the theory that
Method
Design- The independent variable in our study that changes would be the verb we used to word the
question. The dependent variable which is dependent on the independent variable would be the
estimated speed of the car in miles per hour. Different variables were used with different participants.
We used independent groups design in order to keep the legitimacy of the experiment. In order to
conduct our experiment some deception was required to acquire accurate data from the participants.
However consent forms (see appendix 1) were used and the true aim of the experiment was explained
afterwards.
Participants- In our experiment we used a ninth grade regions history class. The class was co-ed with
students aging from 13-15 years old. We randomly distributed the different questionnaires alternating
between the three after each student. We used the sampling technique of “Opportunity Sampling”.
Opportunity Sampling is a technique used by researchers that uses participants that are convenient to
use. My partner and I conducted the experiment in this class due to us both being free at this time. The
class had approximately 25 students so we simply gave questionnaires to the first 20 who handed in
Procedure- We gave the test subjects each a questionnaire (See appendix 4) with four questions on it
after they watched the video of the car accident. Prior to the video playing the subjects were instructed
to take careful notice to detail for the short-term memory experiment we were conducting. Out of the 4
questions, only one, the last one, was of any significance. The other three questions were filler to help
make our fake short-term memory experiment seem believable. The other questions referred to color of
the cars, time of day, etc. The last question however was worded in three different ways. One group of
kids received questionnaires with the fourth question reading, “Estimate the speed of the oncoming car
when it smashed the other car.” The second group had the same question but instead of “smashed” it
read “bumped”. Finally the last group, control group, got no additional verb in the question. It read,
“Estimate the speed of the oncoming car.” This group was expected to get average results around the
middle of the two other groups; we expected the “smashed” group to receive significantly higher speeds
than the “bumped” group which would have the lowest out of the three. In the co-ed freshmen class,
after we collected signed consent forms, we began our experiment. The questionnaires were given back
anonymously however we ensured all the students would try their best to give accurate and honest
results. Before handing them out, we organized the three different questionnaires so they alternated
after every subject. After the results (See appendix 3) were collected we explained that the students
weren’t really a part of a short-term memory experiment and that we really intended to see if we could
alter memory (See appendix 2). We also explained we only cared about the last question, and that
The results from our experiment(See appendix 3) depict that the more intense verb ( smashed )
got higher results in miles per hour than the weak verb ( bumped ), while the control group which didn’t
have an extra verb got results in the middle. The data in the chart includes results on Standard
Deviation and Range. The Standard Deviation of the “smashed” group was significantly higher than the
Standard Deviation of the other two groups. This difference is due to the range of 60 that “smashed” got
in comparison to the 35 and 45 ranges of the control group and “bumped”. The mean of the “smashed”
The results our experiments are fully parallel with the original study performed by Loftus and Palmer in
1974. The results (See appendix 3) suggest that the implications of a word can influence the response
given by the participant. This theory is coined as “schema”; it suggests that people process information
according to their existing mental knowledge. The participants were deceived and introduced to a
misleading question and behaved as planned. The mean estimated speed for the more intense verb,
“smashed”, were far higher than the mean speeds of the less dramatic verb, “bumped”, which received
minimal estimates . Thus we conclude that the verb involuntarily illustrated information from the
schema concerning the speed, and this interfered with the retrieval of memory and caused different
results for the two separate verbs. Our results support that the wording does affect the retrieval of
memory, which Loftus and Palmer proposed in 1974 claiming that eye-witness testimonies are invalid
due to similar findings. This was clearly demonstrated in the results of our experiment and the original
The main strength in out experiment is the ability to have control over the variables due to the
use of deception. If the participants believe they are being tested on their short-term memory, (which
in a way they are), they will not put any extra effort into paying attention to the wording. Not knowing
that random people got different variations of the question ensured the control of the variables. The
independent variable was manipulated at ease allowing different dependent variables in the three verb
groups.
However despite our perfect results, our experiment did have limitations. In particular the
aspect of our participants can be questioned. We did not have every layer of the population, nor did we
take into account the ages of our participants. The 9 th grade students we picked to be tested could have
been chosen better. We took a regions level history class with rather young students. The two main
concerns are that the participants won’t take the experiment seriously. The other being the fact that
these kid’s ages varied from 14-15 years old. In New York the driving age is 17, most of these
participants have never driven a car before, therefore lack the common second nature knowledge of
speed and acceleration. Performing the experiment alone there was a large room for error in the field of
cheating and collaboration of answers. Some kids sitting near peers could have easily realized the
difference in the last question and jeopardized the results, or inferred the aim of the experiment.
Perhaps another main limiting factor would be the ecological validity of the experiment, which ties back
to the question of age and scope of knowledge. How can 9 th grade regions history students be
generalized to full grown adults in eye-witness scenarios? This was also one of the leading points Loftus
was criticized for in his work. The experiment was performed in a classroom environment, with the
movie being projected in the front. If the participants were in a real life scenario, surely they would react
different, and perhaps take more care in analyzing what they have witnessed instead of chatting with his
Coming back to “schema” culture could be yet another factor. Both ours, and Loftus and
Palmer’s study were performed in Western countries, and research done in more Eastern cultures could
In the future, moderations would need to be made to the sample and the sample size. To avoid
unwanted communication between the students, separate rooms could be used. Lastly the sample could
vary in culture to allow schema to come even more into play. All in all our experiment was in line with
the original and supports the theory of manipulating memory retrieval through schema.
References
Bebop0090. (2010, June 1). Car Crash Modified Loftus and Palmer [Video file]. Retrieved from youtube
database.
Loftus, E. F., & Palmer, J. C. (1974). Reconstruction of Automobile Destruction : An Example of the
Interaction Between Language and Memory. JOURNAL OF VERBAL LEARNING AND VERBAL
I understand that I have the right to withdraw from the research at any time, and that any
information/data about me will remain confidential.
The research will be conducted so that I will not be demeaned in any way.
I will be debriefed at the end, and have the opportunity to find out the results at a later
date.
Name:________________________________________________________
Date:_________________________________________________________
Parent:________________________________________________________
Appendix 2 – Statement of Debriefing
“Class, now that we have collected your questionnaires we can reveal the true aim of the experiment. In
order to ensure the validity of the experiment we were forced to use a small amount of deception in our
method. The last question on the questionnaires was the only question we really needed; the others
were decoy fillers to help aid in the validity. Three different questionnaires were handed out. The last
question was altered in each. Some of you received the verb “bumped”, some received “smashed”, and
some received no verb, this was the control group. A control group is a group that we use as a basis to
compare to that doesn’t include a variable. The difference in verbs was done in order to attempt to
manipulate your memory retrieval of the video. We expected to receive higher speed estimations for
“smashed” and lower for “bumped”. If you wish to know your results you can contact us and we will
reveal the data from the experiment. Thank you for participating in our experiment today, enjoy the
weekend.”
Appendix 3- Raw Data
Appendix 4- Questionnaires
4. In miles per hour, estimate the speed of the oncoming car when it bumped the other car.
4. In miles per hour, estimate the speed of the oncoming car when it smashed the other car.