Phoebe Crampton
P2500572
PSYC2013_1819_520
A study to examine the effects of false memory using a within groups design on university students
A study to examine the effects of false memory using a within groups design on university students.
Abstract
False memory is a well-known, common phenomenon which has spiralled stacks of research. The idea of retrieving something that isn’t there tended to be more of a key research focus in the late 20th century in comparison to now, the 21st century, therefore this report focuses on trying to replicate findings of mature study. 30 university students participated in a similar procedure to the DRM (Roediger and McDermott, 1995), where distractor words and critical lures were used to attempt implant false memories into participants. The literature discusses the findings to this study and evaluates the current and previous findings to try and increase worldwide knowledge of the topic as a whole.
Introduction
Memories are frequently presumed to be entirely accurate to what an individual perceives to have happened in a certain event. However, memories are not just stored and retrieved; information has to be encoded first. This therefore means memories are very vulnerable to interference which can easily be manipulated, misplaced or even created. If these false memories can be elicited and retrieved, without people being deceitful about events, then surely there has to be a distortion of scenarios in the brain. Maybe this is interference with past events, fantasy or even events from dreams that have somehow changed into memories. Bahrick, H. P., Hall, L. K., & Berger, S. A. (1996). piloted a study which displayed that people tend to actually remember the good grades they achieve in high school, but forget the bad ones; and then over estimate them! This evidence tries to argue that well documented events are typically correct, unless they reflect poorly on the individual – therefore relating back to the idea that fantasy and imagination can have an effect on memory.
Another psychologist, Deese (1959), contributed a huge amount into the phenomenon of false memories. He initiated a procedure about how false memories can be elicited called the DRM paradigm, which was later made popular after further research by Roediger and McDermott (1995). Participants in original studies were given an oral presentation of associated words to a prototype (critical lure), for example; bed, rest, dream: which would all be associated to sleep. They were told to try to remember these words for recall – however the prototype was not involved in this list, only the associates to this activity. Once the presentation was over, participants were asked to recall as many words as they could. They found that the prototype would typically be recalled in this list by the subjects, thus creating a paradigm that it is very easy to create a false memory. Furthermore, research by Roediger and Mcdermott (1995), conducted additional tests which implemented a recognition task. After seeing the word lists, participants were shown the critical lure, and asked if they can either remember hearing that word in the word list, or whether they just find the word familiar. They found a 72% increase in participants who said that they remember the word being said in the baseline study. This replication provided evidence of robustness of the false memory effect and therefore intrigued more researchers to dig further for more information.
Much research focuses on stimulating false memories, however research conducted by Hyman, I. E., Husband, T. H., & Billings, F. J. (1995) tried to investigate actually implanting a memory to the participant to see it evoked memories. Researchers interviewed participants over a number of days, with previous knowledge about their childhood, and asked them to recall 3-5 incidents which they remember from their childhood. These events were also joined by a false situation. It was found that every fourth participant claimed to remember the fabricated event and added an extra detail to it. A more recent example of false reports of childhood events supports this finding, meaning it has temporal reliability (Ost, J., Foster, S., Costall, A., & Bull, R. 2005). This is supported further by other studies, such as Lindsay, D.S., Hagen, L., Read, J.D., Wade, K.A., & Garry, M. (2004). who also say this can increase again by 60% if a false picture is used in this event.
The research above has demonstrated extremely important for our knowledge of memories and how they can occasionally be false. With saying this, it would be interesting to see whether todays level of social media and technology has taken an effect on false memories and how they can be coded. Therefore, this experiment will try to indicate a replication of the baseline studies, however, if it doesn’t have the same findings, it can hopefully provide a fresh understanding about the subject. Following the research conducted previously, it is hypothesized that 1: there will be a significant difference between critical lures and distractor words, 2: there will be a difference between critical lures and original words and 3: there will be a difference between original words and distractor words.
Method
Participant’s
Participants in this study were all 2nd year Psychology students at the De Montfort University, Leicester, England. All participants were recruited using opportunity sampling due to them all being around at the time of the experiment. the contrast between male and female was 5 men (16.7%) and 25 females (83.3%), meaning that there was a total of 30 participants. The youngest participant was 19, involving 50% of the sample, and the oldest participant was 41. This wide age range prevents age related bias. The mean age of this study was 21.40% (SD=5.21).
Design
In this study, a within groups design was used, so there could be a comparison made between conditions. This is an experimental design. The independent variable was the type of word used in the experiment which had 3 levels – original words, unrelated distractors and the critical lure. The dependant variable was the percentage recall of each condition.
Materials
The materials used in this experiment are as follows; a computer for the participants to take the task on, the computer used for this task was A HP AIO EliteOne 800 G3 1KB38EA#ABU Core i5-75007 8GB 1TB DVDRW 23.8Touch BT CAM Win 10 Pro. A quiet room would be needed for the participants to focus, and chairs and desks need to be provided for the comfort of participants. An A4 briefing sheet and consent form was given to participants before the experiment, therefore a printer with paper would be needed to print many of these out for the participants to fill out before the experiment took place. This document had size 12 and had Ariel font, with black ink (see appendix A). Pens for the participants to use would also be needed. For the experiment, Superlab version 5 was used to create the word lists, where the words for the encoding process were in the middle of the screen in font style Tahoma, size 36. In the recall phase of this experiment, where the grid was shown, the font style ‘Calibri’ was used in size 18. A further sheet of paper with the grid for participants to record their answers down would be needed, this was also written in ‘Calibri’ in size 12 font (see appendix B)
Procedure
Participants entered the room and were asked to first read a briefing and consent form, which they then would sign and hand one copy back to the researcher. Participants were given two of these, one to keep and one to hand back in. They then sat down on a chair in front of a computer screen, which had the experiment ready on the screen. Participants were allowed to sit where they pleased. Participants read the instructions on the first slide and had a trial run to make sure that they understood the process. This trial run was the same format as the real experiment, however with different words. Participants instructions guided them to try to remember as many words off of the given word lists as they can and when the grid appears after the word list was shown, participants were told to write down the words that they remember, as well as the ones on the grid which they think was included in the list too (see appendix B). They did this for all 3 conditions. Each trail had a mixture of all 3 conditions in the grid after each trial, but only original words in the trial list. Once the whole experiment had been completed, participants were asked to look back at their results. They were shown a copy of all of the words that had been shown in the grid at the end of each trial. Participants were asked to count up how many original words, unrelated distractor words and critical lures they had after each trial. They then added each trial together, for example all of the original words they got correct across every trial, and then written this into a grid which was used to calculate the finishing results (see appendix C). Participants swapped their answers with the person sitting next to them to prevent participants from changing their answers due to demand characteristics and also to prevent embarrassment of the participants once they were asked to shout out their answers. Once participants told the researcher their answers, the participants were debriefed and then was allowed to leave the room.
Results
The Mauchley’s test that was taken out was conducted revealed that the assumption of Sphericity had been violated, therefore the degrees of freedom were corrected using the Greenhouse-Geisser estimates of Sphercity ( =.69) . A repeated measures ANOVA indicated that there was a difference between the conditions F(1.38,40.22) = 131.97, P< .00, ηρ2 =.82. Post hoc analyses (Bonferroni adjusted) revealed that there was a significant difference between recall original words and unrelated distractor words, (p<.00, 95% CI [60.00, 74.53]) and recall critical lures and unrelated distractor words (p<.00, 95% CI [-73.12,-49.00]), however there wasn’t a significant difference between original words and the critical lure words(p =.84, 95% CI [-8.20, 20.61]). (see appendix D for SPSS graphs) Figure 1 shows these results below.
Figure 1
Discussion
The results show that two-tailed hypothesis 1 and 3 have been accepted and the results were in fact significant. It is therefore concluded that false memories can be implanted and that somewhere along the line of perception and output, there is something which can trigger a false response. This means that the paradigm created by Deese (1959) has great temporal validity and the findings are still true to this day. However, hypothesis 2 was not supported and was not significant which suggests that critical lures are not important when studying false memory.
Research which was discussed earlier, (Roediger and McDermott, 1995), had conflicting results to the results in this experiment. although hypothesis 1 and 3 have been accepted, and supports the previous research, hypothesis 2 conflicts this and could show that our rationale for this experiment has proven correct; that the previous research has become dated, and therefore inaccurate. To make these statements more accurate, an exact replication of the study would need to be taken out, as this experiment tested the concept, however not in the same was as Roediger and McDermott (1995).
Using a majorly female heavy sample could have affected the results in this experiment. Shelton and Kumar (2010) concluded the same results in their findings. Of course, in the particular experiment, only 16.7% of the sample were male, meaning that this is gynogenic research, and therefore heavily bias toward this gender. This means that for the results to be more meaningful, gender must be equal to see a more rounded result. Future research may wish to focus on balancing the gender sample to prevent this bias, and maybe find more significant results which can we more appropriately applied to both genders to have a more effective result for application to real life therefore increasing the ecological validity for this study.
One weakness of this study would be that the results could be untrue. Participants were asked to calculate each other’s results; therefore, it cannot be accurate to say that they were recorded down correct. Many things such as misreading answers, participants not paying attention and also demand characteristics can all have an effect on the results. This means that the study may lack reliability. However, a more accurate way, although more time consuming, to collate this data may have been for the researcher to do this themselves, so that the data is definitely correct. This may be something that could be considered in further research.
Overall, it seems that false memories can be implanted successfully. This suggests that although previous studies were conducted many years ago, their findings still stand, and are still correct to use in todays world. This follow up study was successful and can hopefully provide more information about when false memories can occur.