Also, to eliminate any ethical problems I will give each participant a full debrief after the experiment. This will tell the participants which word list they had, the title of the experiment and what it entails.
Abstract
I aimed to replicate the methodology that Bower et al. used in their experiment and to discover if the organisation of words did increase recall. There were two conditions in my experiment, one set of participants had identical words arranged in an organised manner, and another group had the same words arranged in a disorganised manner. This meant that my independent variable was the organisation of words and my dependent variable was the number of words recalled.
The participants I used were selected using opportunity sampling at a college open day. The thirty participants I used were therefore high school students, in their final year so aged fifteen or sixteen years. As a result, my sample was not representative of the general population.
The results I collected were analysed using the Mann Whitney U statistical test. From this I calculated that my results were significant to a level of P<0.005 which is a much higher level of significance than that expected which was P<0.05. Therefore, I was able to conclude that my experimental hypothesis, organisation of words increases recall, is a true statement.
Introduction
When investigating memory in psychology, organisation is a key factor to consider. Without organisation of a memory, the speed at which information is recalled or accessed from the memory would be considerably slower. There are many different models and theories that suggest that memory is organised or structured.
Craik and Lockhart (1972) developed the ‘Levels of Processing Theory.’ They argued that deep processing needs to occur for information to be transferred from the short-term memory (STM) to the long-term memory (LTM). Organisation is a form of deep processing. It creates a lasting memory, so conscious processing or rehearsal does not need to occur.
Mandler (1967) proved Craik and Lockhart’s theory by giving participants a pack of 52 cards with a word on each of them. The participants were then asked to organise these into two to seven categories. When given a free recall test, it was found that those who used two categories recalled the least amount of words, but those who used seven categories recalled the highest amount of words. Mandler concluded that those who used the most categories had imposed the most organisation, thus improving recall.
Collins and Quillan (1969) also developed a model based on organisation in memory. This is known as the ‘Hierarchical Network Model’ (HNM). They thought that information was stored in a hierarchical fashion in memory. All of the major concepts are known as ‘nodes’ and each one of these has a number of different properties associated with it. These nodes are all inter-linked. Retrieval of information was suggested to occur by searching through the semantic hierarchy. Collins and Quillan tested the HNM by giving participants sentences containing concepts and properties that they then had to state whether were true or false. The more levels of hierarchy the sentence contained, the longer it took the participants to verify the sentence.
Bousfield (1953) also provided support for the HNM. He gave participants a list of sixty words. These words had all been derived form four categories, but were shown to the participants randomly. He found that in a recall test, the participants tended to group the words together into categories. This suggested that the participants had tried to organise the words, hence showing organisation in memory. Rubin and Olsen (1980) found similar findings in a more naturalistic setting. Students were asked to recall staff members’ names and it was found they tended to do so by department.
Tulving (1968) describes this form of organisation (when the participant uses organisation to aid recall) subjective organisation, as opposed to experimenter organisation (organisation imposed by the experimenter). Bower et al. (1969) used experimenter organisation in their work. They asked participants to learn a list of words, which were arranged into conceptual hierarchies. For one of the groups, the words were arranged into hierarchies, whereas for the other group they were arranged at random. The findings were pretty clear cut. The group with the list of organised words recalled, on average, 65 per cent of the words, whereas the group that were shown the words at random recalled an average of only 19 per cent of the words.
The aim of my investigation will be to discover if organisation affects recall, so I, like Bower et al.’s work, will use experimenter organisation. Therefore, my experiment will follow along very similar lines to Bower et al.’s. However, I will use different hierarchies. This is because I am going to use categories rather than loosely linked hierarchies that Bower used. Also, I am going to replicate Mandler’s work in terms of the number of categories used, seven. I am going to use thirty-five words – five in each category. I got this value from Bousfield’s work, who used sixty words, but I figured that would take too much time. Finally, I will give my participants one-minute to recall as many words as possible. This is based on Bower’s experiment.
From the research I have conducted, it is quite clear that the organisation that I will impose will increase the recall, so a directional hypothesis will be used. Therefore, my experimental hypothesis is:
‘Organisation of words will increase the number recalled’
My null hypothesis is:
‘Organisation of words will have no effect on the number recalled’
Method
Design
I used an independent measures experimental design. The independent variable of the experiment was the organisation of words and my dependent variable is the number of words recalled by my participants.
Participants
I used forty participants, twenty of each gender. All participants were aged between sixteen and eighteen years of age and attending Dereham Sixth Form College. I used opportunity sampling, by going into the common room at the sixth form and randomly selecting participants who showed curiosity or interest regarding my experiment. I selected two participants at a time and gave one the organised word list and one the disorganised word list.
Apparatus/materials
I used two word lists, one of which contained a list of organised words and one where the words were arranged at random (see appendix 1 and 2). I designed these words myself by first choosing a word and then associating four words that associated with it. This continued in a hierarchical fashion until seven ‘categories’ had been arranged. This gave me thirty-five words in total. All of these words were words that would be very familiar to the participants.
Standardised procedures
I used standardised instructions (see appendix 3) to explain to each participant what is required of them. This also assured the participant that they are free to leave my experiment at any time. I then gave them one minute to try and learn as many words as possible from the list they were given. The list of words was then taken away and a piece of paper was given to them to recall as many words as possible within ninety seconds.
Controls and Ethics
Firstly, one ethical consideration that I had to deal with was informed consent. To achieve this for each one of my participants I provided them with a list of standardised procedures (appendix 3). This was provided to ensure all participants were on level terms regarding what they knew about the experiment and also to inform them of the nature of the experiment and that they can leave at any time.
Noise was a problem in my experiment as too much noise could have affected recall. I attempted to overcome this by conducting my experiment in a classroom where very little background noise was occurring. This is because in general people who entered the classroom were very considerate and understood that experiments were occurring so were quiet.
Finally, a control, which had to occur, was the control of gender in my experiment to achieve an ideal representative sample. I did this my recording the gender of each participant and keeping this under control by trying to even out the male/female ratio. This was done by enticing the minority gender to partake in my experiment. However, this could have been a problem within itself in that these people may not have wanted to take part but I deliberately selected people who were interested.
Results
By analysing my raw data (see appendix 4) it is clear that the organisation of words did have effect on how many were recalled. Most of my participants performed better on the recall test when they had learned the organised wordlist than those who had learned the disorganised word list. I illustrated this comparison in a line graph (appendix 5). This is made clear from the mean, which I calculated for each condition.
The nature of my experiment was an independent measures design. I was testing the difference between the amount of words recalled by those who learned an organised list to those who learned the disorganised list. This provided me with ordinal data, which enabled me to carry out the Mann Whitney U statistical test. (See appendix 6)
From carrying out this statistical test, the value I got for U’ was 0. When put into the table of critical values I found that my experiment was highly significant to the level of P<0.005. This value is lower than my expected level of significance which was P<0.05, hence I will retain my experimental hypothesis and reject my null hypothesis.
Discussion
F1: Explanation of findings
From the Mann Whitney U test that I have conducted on my results, I found that the level of significance of my experiments was P<0.005. This was a higher level of significance than I predicted which was P<0.05. This means that I can reject my null hypothesis and have proven my experimental hypothesis, therefore, I have found overall that the organisation of words increases recall.
F2: Relationship to background research
My experiment was based on the work of Bower et al. (1969). The evidence that I produced is in support of their findings, organisation does increase recall. My results are not quite as conclusive as Bower et al.’s as they found a very significant difference between the two groups (65% of words for the organised condition compared to 19% in the disorganised condition). However, I still found a significant difference. A possible reason for the difference between the strength of conclusion of my experiment and Bower’s experiment is that the participants in the disorganised condition may have used subjective organisation when learning the word list.
This leads to my next point because the use of subjective organisation was evident in the work carried out by Bousfield, so my work is also in support of his work. This was clear because when the participants recalled the words that learned the disorganised list, some tended to place them in an organised manner, so they had therefore spontaneously organised the words themselves. Hence, although those in the organised condition had imposed organisation on the list, those in the disorganised condition possibly also had subjective organisation on their list.
F3: Limitations and modifications
The fact outlined above concerning the participants in the disorganised condition being able to place organisation on their list poses a first problem with my experiment, as this suggests that both conditions had organisation. However, there is still a difference between the recall in the two conditions so it is clear that not all the participants were able to organise the disorganised word list. A possible solution to this problem could be to establish a different word list that is more difficult to place subjective organisation onto.
Another issue concerning my experiment was that on occasion I had two participants at once. This posed a problem because on recall I found that participants may say a word out loud when they remember it so the other participant would have heard and written it down, as if they had also recalled that particular word. A good solution to this problem though would be to not experiment on two participants at once, or I could have asked the participants to try not to say the words out loud, although the latter may not have been very reliable.
Finally, a modification I would make in the future would be to use a more representative sample of the whole population, rather than students because students memories may be better than the general population because of high intelligence and to good methods of remembering things due to revision for examinations. Hence, I cannot generalise across the whole population that organisation causes and increase in word recall.
F4: Implications and suggestions for further research
A first implication of my research is that organisation does increase recall, implying that the more organised an individual is the better that particular individual will be able to remember things. An example of this is, in student terms, the more organised a students notes are presented (such as notes being organised into clear and distinct topics) for revision, then they will perform better in the examination.
The other implication is of course the opposite of this, the more disorganised a person is, the less likely they are to remember things.
Further research that I could carry out would be to consider the difference that organisation has on the recall of males and females. I could carry out exactly the same experiment design and record the gender of each participant (as I did in this experiment) but actually explore the difference between males and females in terms of the number of words recalled.
I addition to this point, I could investigate whether organisation affects the recall of different aged individuals. This leads back to a modification that I would make regarding the use of a representative sample. I could use participants from all different age groups, such as adolescent, adult, OAP and so on. By doing this I could produce evidence to show that organisation has greater effect on recall of younger or older people.
References
- Bousfield (1953) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p43)
- Collin and Quillan (1969) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p44)
- Bower et al. (1969) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p44)
- Craik and Lockhart (1972) in Michael W. Eysenck and Cara Flanagan, 2000, Psychology for AS level, Psychology Press, Taylor and Francis Group (p46)
- Rubin and Olsen (1980) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p44)
- Mandler (1967) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p44)
- Tulving (1968) in Richard Gross and Rob McIlveen, 1999, Aspects of Psychology: Memory, Hodder and Stoughton (p44)
- Mandler (1967) in Michael W. Eysenck and Cara Flanagan, 2000, Psychology for AS level, Psychology Press, Taylor and Francis Group (p48)
Appendix 1
Word list A – Organised
Appendix 2
Word list B - Disorganised
Appendix 3
Standardised Instructions
I am conducting an experiment as part of my ‘A’ level psychology coursework. The experiment involves the memory aspect of psychology. I will give you a list of 35 words to learn in sixty seconds and then you will have 90 seconds to recall them.
Do not feel pressured by this experiment, you may leave at any time you wish.
Thank-you for your co-operation
Appendix 4
Results
Appendix 5
Comparison Graph
Appendix 6
Statistical Test
The Mann Whitney U Test
U = NANB + NA (NA + 1) – T
2
= 15 * 15 + 15(15+1) – 120
2
= 225 + 120 – 120
= 225
U’ = NANB – U
U’ = 15 * 15 – 225
U’ = 0
The value used = 0