An Experiment On Memory & The Recall Of Words When Using Word Lists
An experiment on memory & the recall of words when using word lists.
NULL HYPOTHESIS 4
INDEPENDENT VARIABLE 4
DEPENDENT VARIABLE 4
EXTRANEOUS VARIABLES + CONTROLS 4
APPARATUS & MATERIALS 5
SUMMARY OF RESULTS 6
RESULTS – TABLE 1 8
RESULTS – 2 9
STANDARDISED INSTRUCTIONS -311
DEBRIEFING SHEET -312
WRITTEN WORD LIST -4 13
SAMPLE OF PARTICIPANTS RECALL OR WORDS (I.) GROUP A(1.1) -6 14
NUMERICAL CALCULATIONS -7 16
It was hypothesised in Craik & Lockhart’s levels of processing theory that the quality of recall of serial written material is a function of the depth of cognitive processing. The aim of this study is to test the levels of processing theory to see if the level at which information is processed has an effect on recall. The hypothesis being tested was is that the participants who process the semantic list will have a greater word recall rate than those who process the structural level list.
One hundred further education students (68 females and 32 males, mean age 17.4 years) were selected using opportunity sampling for this independent measures experiment & were tested for recall of one of 2 lists of 15 words, each list presented in written form with one minute to study the list. Half of the participants were given the list that required a semantic level of processing & the other half the list that required only a structural level of processing. The results of the experiment showed that those who processed the semantic word list on average recalled more correct words (mean 8.94) than those in the structural group (mean 4.6). The results confirmed the hypothesis, & it was concluded that they support the levels of processing theory of memory
Obtaining an understanding of the structure & functioning of human memory is a major pursuit in modern psychology, because it is necessary for the understanding of learning, cognition & the major conscious & unconscious processes. One ongoing discussion is whether memory can be divided by structure & function into different stores, or it should be considered a continuum of levels of better recall, corresponding to the depth of levels of processing of the information. The present paper presents a brief overview of the two opposing theories, & provides an account of an experiment that was designed to prove that the depth of processing is positively and causally correlated to the quality of word list recall.
In 1968 Atkinson & Shiffrin proposed a multistore model of memory that consisted of sensory, short-term & long-term stores with different functions, capacity & persistence of stored information. Other researchers elaborated on this model, identifying additional structural components to each of the stores. Badderly (1992) suggests that the short-term consists of components he calls the phonological loop, the visual-spatial sketchpad & the central executive. Baddley & Vallar (1984) found neurological evidence for the existence of the phonological store by comparing the verbal functioning of a patient with an infarction in the left hemisphere with that of healthy control individuals.
The above authors emphasize on the importance of rehearsal of audio-verbal material in short-term memory (more specifically the phonological loop) before it can be copied into the permanent long-term store. Unrehearsed material is lost.
Craik & Lockhart (1972) proposed that the processing of a memory trace begins as simply perpetual and becomes deeper & more abstract. There are no structural blocks of memory corresponding to this process but memory can be described as a continuum of deepening levels of processing. The persistence of a memory trace is a function of the depth of processing. Lewandowski & Hockley (1987) found that deeper processing caused better recall independent of the encoded material, that is considered to be an important sort of interference by the multistore models.
The aim of this study is to recreate Craik & Lockhart’s level of processing theory experiment to see if the level at which information is processed has an effect on recall.
The hypothesis is that the participants who process the semantic list will have a greater word recall rate than those who process the structural level list.
There will be no significant difference between the number of successfully recalled words from the semantic & structural lists & any difference is due purely to chance.
The hypothesis was tested by carrying out an experiment. Independent measures was the experimental design used to carry out the research. The Independent variable for this experiment was the word lists given to the participants & the dependent variable was the number of words correctly recalled.
The independent variable is the lists given to the participants. I.e. whether the processing is on a semantic or structural level.
The dependent variable is the number of words recalled in three minutes by both group A (Structural list) & group B (semantic list)
Extraneous variables + controls
An EV is the possibility of noise form adjacent classrooms thus distracting the participants. This is controlled by scheduling the experiment at 4.30PM so the other rooms were empty.
An issue is that since the participants are not given the full information at the start of the experiment it is essential that a full debriefing is given later. To resolve this the students were debriefed by the researcher reading out the debriefing sheet & also answering up any follow-up questions that were asked (See appendix)
The participants must always be informed at the outset of the experiment at they do have the right to withdraw at any point (even after). Also all information about the participant must be kept confidential indefinitely unless the participant has agreed otherwise in advance.
The target population was all full time students at a centre of further education. 100 participants (68 female/32 male with an age range between 16-25 years) were selected by opportunity sampling. This method was used since it was the least time consuming method available.
Apparatus & materials
The stimulus material was two lists of words & instruction cards of two types. Each word list contained 15 words of general use that should not cause uncertainty in comprehension. Instructions: (a) "Answering on the sheet is each word written in capitals or not", (b) Answering on the sheet do the two word on each line mean the same thing". (a) was considered to require shallow processing & (b) to require semantic processing.
The following was also used:
Ø Response sheets – a blank piece of paper
Ø Standardised instruction sheet (See appendix)
Ø Debriefing sheet (See appendix)
The participants were randomly divided into two groups. The first 50 were assigned to group A (structural), the other 50 assigned to group B (semantic). For both groups the experimental conditions were kept constant.
(1) for group A & (2) for group B. The participants were given the word list facing down on the desk in front of them & were told not to turn it over until instructed to do so. They were told to turn over the paper at the same time and answer the questions presented on the sheet (See appendix 5 & 6). They were given sixty seconds to do this and then told to turn the paper face down again. Once this was done all the sheets were collected back in & a blank sheet handed out, the participants were then told to write down the words from the list as best they could remember in any order.
The lists of words produced by the participants were compared with the original lists, and the numbers of correctly recalled words were recorded.
A table to show the mean, median, mode, range & standard deviation of the two groups
Structural test Semantic test
Mean 4.6 8.94
Median 4 9
Mode 5 8
Range 10 8
St. Dev. 1.95 2.27
The complete numerical data is given in table 1 & the means are represented graphically in figure 1
Summary of results
Both the mean and the median were higher in group B (semantic). The mean for group B was 8.94 compared with a mean of 4.6 for group A. The median for group B was 9 compared with that of group A which was 4. The highest score recorded during the experiment was a perfect 15, this was achieved by a member of group B with the semantic list & the lowest score of 2 was in group a with the structural list. On average group B recalled more than group A which suggests that levels of processing has an effect on word recall. The mode (the most frequently occurring number) was also higher in group B than A being 8 & 5 respectively.
The results do support the experimental hypothesis: - participants who process the semantic list will have a greater word recall rate than those who process the structural level list.
The experimental results show that the recall of words was better for participants who processed the words on a semantic level rather than on a more shallow or structural level. The results agree with the findings of Craik & Lockhart and confirm the initial hypothesis that the level of processing determines the persistence of memory.
The results also conform with other research. Craik and Tulving (1975) found that people were about three times as likely to recall a word if they had originally answered questions about its meaning than if they had originally answered questions about the word’s physical appearance. Similarly, Parkin (1984) discovered that people who made semantic judgments about a word’s category or its synonym performed much better on a surprise recall test than did people who made non-semantic judgements (for example, about the number of vowels contained in a word or whether it had been printed only in capital letters).
There were several limitations to this experiment, firstly only students were used who are not representative of the population as a whole. The time of day may have also had an effect on the participants (4.30pm), at the end of the day the participants may have been tired & had their minds on other things i.e. what’s for tea? It would have been better if the experiment had been carried out at several different points during the day.
Further research could be to test the differences in responses for males & females, the effect that the time of day has on recall (if any) & obtaining a wider sample of the population to make the findings more representative of the population as a whole.
The findings of this experiment alone are not sufficient to prove the levels of processing theory but the study overall supports the ideas of Craik & Lockhart.
It can be concluded from this study that the ability to recall words did seem to improve when the information was processed semantically rather than structurally and therefore supports the hypothesis. The obtained results show that group B (semantic) did recall more words than group a (structural). It is also clear when looking at the histograms that group b recalled more than group A.
The deeper & more abstract the information is processed the more easily it can be recalled.
Psychology: the Science of Mind and Behaviour
Hodder & Stoughton 2001
· Atkinson & Shiffrin (1968)
· Badderly (1992)
· Baddley & Vallar (1984)
· Craik & Lockhart (1972)
· Craik & Tulving (1975)
· Lewandowski & Hockley (1987)
· Parkin (1984)
Results – table 1
Participant no. Structural test Semantic test Participant no. Structural test Semantic test
1 5 6 26 6 10
2 4 7 27 4 15
3 6 8 28 5 8
4 4 10 29 2 9
5 11 9 30 7 12
6 8 6 31 5 9
7 3 4 32 6 6
8 5 7 33 1 13
9 3 7 34 3 9
10 4 13 35 4 8
11 5 7 36 6 10
12 3 8 37 2 7
13 8 9 38 5 9
14 4 6 39 8 13
15 5 9 40 2 10
16 2 8 41 5 8
17 6 10 42 4 9
18 3 11 43 4 8
19 3 13 44 6 8
20 3 10 45 5 10
21 5 14 46 2 7
22 4 8 47 3 7
23 5 9 48 3 8
24 4 9 49 7 11
25 4 7 50 8 8
Caption: table of results showing number of words recalled per participant.
Results – 2
Standardised instructions -3
Thank you for taking part in this experiment.
As with any experiment you have the right to withdraw at any time before during or after the experiment if you wish.
You will be given a sheet of paper face down in front of you, please do not turn it over until instructed to do so. When you turn the sheet over you will be given sixty seconds to complete the task specified at the top of the sheet. You may write your answers on the sheet & when you are finished or time is up please turn your paper back over and they will be collected in.
*After this has been done*
Now you have been being given a blank sheet of paper, please write down as many of the words as you can remember in any order. Your responses will be collected by the experimenter and will remain completely anonymous at all times.
Debriefing sheet -3
Thank you for taking part in this experiment.
The aim of the experiment was to test the memory and recall of words. The people who agreed to take part in the experiment were split into two groups, group a & group b. Group a were given a list of words that would be processed by the brain on only a low level & group b given a list that required the brain to process the words more deeply.
The aim of the experiment was to test if group b with the deeper of semantic word list recalled more words than group a.
Your data will be collated along with the data of the other participants and will remain anonymous & confidential at all times.
Please do not hesitate to ask if you require any further information or have any questions or comments.
Written word list -4
Sample of participants recall or words (i.) Group A(1.1) -6
1. Is Hello spelled with capital letters? _____
2. Is DRIVE spelled with capital letters? _____
3. Is row spelled with capital letters? _____
4. Is LEARNING spelled with capital letters? _____
5. Is make spelled with capital letters? _____
6. Is SHOPPING spelled with capital letters? _____
7. Is WORKING spelled with capital letters? _____
8. Is maths spelled with capital letters? _____
9. Is WATER spelled with capital letters? _____
10. Is OCEAN spelled with capital letters? _____
11. Is relax spelled with capital letters? _____
12. Is drunk spelled with capital letters? _____
13.Is restless spelled with capital letters? _____
14. Is ANGRY spelled with capital letters? _____
15. Is back Spelled with capital letters? _____
(ii.) Group B(1.3)-6
1. Does hello mean the same as hi? _____
2. Does drive mean the same as ride? _____
3. Does row mean the same as fly? _____
4. Does learning mean the same as knowing? _____
5. Does make mean the same as create? _____
6. Does shopping mean the same as drinking? _____
7. Does working mean the same as playing? _____
8. Does maths mean the same as science? _____
9. Does water mean the same as H2O? _____
10. Does ocean mean the same as sea? _____
11. Does relax mean the same as rest? _____
12. Does drunk mean the same as intoxicated? _____
13. Does restless mean the same as tired? _____
14. Does angry mean the same as upset? _____
15.Does back mean the same as front? _____
Numerical calculations -7
Mean Group A Group B
230/50= 4.6 447/50= 8.94