An investigation to see the effect of chunking on short term memory recall.
An investigation to see the effect of chunking on short term memory recall
Contents
Introduction
Background Research
Rationale
Aims
Hypothesis including Null Hypothesis
Method & Design
Variables
Participants
Apparatus
Procedure
Controls
Results
Summary Table
Summary Table commentary
Descriptive Statistics Table
Descriptive Statistics commentary
Relationship of results to hypotheses
Discussion
Validity
Suggestions for improved validity
Reliability
Improving reliability
Implications of study
Generalisation of findings
Applications to real life
References
Introduction
Background Research
Cognition is the activity of internal mental processing. This has been the focus of many psychologists in their studies. It involves the way human minds think, recall and perceive information. Cognitive psychologists the empirical studies below, discuss the evolution of chunking.
Chunking theory, is a technique, which improves performance of recalling. If this technique actively practised or rehearsed, it will help improve retention for the subject. This is achieved by an increase in the knowledge about patterns concerning the task. These patterns are called chunks.
Chunking was first proposed as a model of human memory by a Harvard psychologist called George A. Miller in (1956). Miller's main hypotheses were 'how many digits people could be remember a few minutes after having been told them'. The answer to this hypothesis was: "The Magical Number Seven, Plus or Minus Two". This meant that a person's short-term memory (which is equivalent to working memory) could hold from five to nine units of information. Miller in 1956,quoted the following statement to support his claims:
"A human requires and organises knowledge of the environment by forming and storing expressions, called chunks. Which are structured collections of the chunks existing at the time of learning"
G. A Miller (1956) the Psychological Review
Chunking theory has a key assumption that each chunk, which is encoded in the brain, is broken into subparts and processed into meaningful units at the same time. This type of learning involves the use of short term and semantic memory. This is the learner is actively processing information. Since the 1960's, cognitive psychologists such as Newell, Shaw and Simon, 1955-60, have drawn comparisons between computers and human thought as both processes involve manipulating information.
Sperling (1960) produced a study into the capacity of short-term memory. Sperlings asked the question if people could visually see more then they could remember. Sperling used 5 participants (which were students) to conduct the experiment. Participants were shown a list of five items consisting of letters and numbers. Each item was formatted in the same font and font size but in different tones .for 50 milliseconds.
The stimulus material (the five items mixed with numbers) was deliberately chosen to prevent participants interpreting the letters as words. Sperling concluded that information is initially held in the sensory store (as did Atkinson and Shiffrin (1968)). Sperling suggested that the sensory store has a greater capacity than that for short-term memory, but the information decays very rapidly and is forgotten.
The experiment was performed in laboratory conditions. Hence the experimental conditions indicate that the experiment permits replication because of the high controls. Therefore and has good control of extraneous variables.
The results of the experiment reported that participants typically recalled the letters until were the number is present in the material. This study justifies that the learner during performance creates chunks and that it improves recall.
Simon (1974) conducted a further experiment into chunking theory. Simon established that there is a limit on effect in terms of the size of the chunk. He concluded that there is a shorter memory span for greater chunks.
Bower and Springston (1970) concluded that meaningful chunks such as (FBI) are recalled better then chunks that show no meaning such as (TZP). In order for a chunk to determine meaningfulness, the participant relies on long-term memory. The capacity of long term memory is unlimited.
The idea of phonetic sounds is important. Chase and Simon (1973) adapted Millers theory of chunking and performed a further study in relation to the game of chess. They proved that exceptional chess player would brake down each movement into chunks, which improves their performance.
Rationale
The empirical study outlined in the background research provides a basis for the investigation.
Chase and Simon (1973) findings can be linked to the investigation of the effect chunking in terms of recall. Chase and Simon (1973) found that exceptional players of chess chunked the game into a series of strategies, which improved their performance.
In relation to my experiment these findings can be of use as the learner will develop strategies to chunk the second list of letters, which consists of vowels, and consonants (which incorporate a phonetic sound that enables the letters to be chunked). These phonetic sounds will enable the learner to incorporate the letters into meaningful units, which would improve the performance giving better recall. In comparison to the first list which consists of pure consonants which cannot be chunked
I would expect the results of a participant based on Millers (1956). "The Magical Number Seven, Plus or minus two" for the participant to hold from five to nine units of information. When using the technique of chunking in the second list.
Aims
To investigate whether chunking a group of phonetically sounded syllables into meaningful units will effect short-term recall.
Hypothesis
A list of consonants with alternate vowels forming phonetically sounded syllables, can be recalled better than a list of letters consonants.
The null hypothesis
There will be no variance ...
This is a preview of the whole essay
I would expect the results of a participant based on Millers (1956). "The Magical Number Seven, Plus or minus two" for the participant to hold from five to nine units of information. When using the technique of chunking in the second list.
Aims
To investigate whether chunking a group of phonetically sounded syllables into meaningful units will effect short-term recall.
Hypothesis
A list of consonants with alternate vowels forming phonetically sounded syllables, can be recalled better than a list of letters consonants.
The null hypothesis
There will be no variance between the recall phonetically and non-phonetically sounded groups of letters.
Method & Design
To test the hypothesis, an experimental method shall be used so all conditions for the experiment can be applied to different participants to make the findings more ecologically valid.
While the experiment is being conducted, repeated measures will be implied. This method has been chosen so the same participant can do all parts of the study. The advantage of conducting the experiment this way is that we will need fewer participants. Also the same participant's will undergo the same conditions therefore there will mean fewer problems with individual differences.
There are two conditions all participants will go through. The first condition is asking participants to recall the first test consisting of sixteen consonants. The second condition is to ask the participants to recall the sixteen letters consisting of constinents with alternating vowels.
The task will take place in laboratory conditions. Therefore the conditions of the experiment can be controlled such as the temperature of the room. As the experiment is well controlled, it can be repeated, and similar results can be found. The disadvantage of laboratory settings is that it lacks realism. Therefore the experiment is ecologically invalid.
The experiment will be divided into two tests. The first test will consist of a list of random consonants as shown below:
TEST 1
N H P Z C T M Q W V L B T Z P X
The second test will consist of random consonants alternate vowels. This will in theory enables the participant to chunk the letters into small groups of words with phonetic sounds. According to the hypothesis, this should make the test easier to process and therefore higher recall. The test is shown below:
TEST 2
M E R A L I S OR E N U R L U K
Both written horizontally in bold "New Roman Style" font at size 30. This was done to make the formatting clearer, hence easier to read. Each participant will be given the same test for the same duration of 20 seconds to look at the stimulus material and have a duration of 16 seconds to write there recall of letters down. The diagram below shows an example of the answer sheet, which is given to all participants before the experiment is conducted.
The results are to be written by the participants so they list the recalled letters wherever it to be allocated.
Name .................................. (m/f)
The experimental conditions need to be reliable and easily replicated for further studies. The experiment is going to be set in a classroom where the environment can be controlled. For a replicable experiment, a good control of variables is needed. The experiment will be designed in a way so that the participants will find it easy to carry out the experiment. This will allow the experiment to be carried out more effectively. There will be twenty participants conducting the experiments, in the same ratio for each gender. This will make the test fair and better to analyse.
The test will be conducted similarly, that of other cognitive psychologists such as Sperling in 1960. This is because the designs are similar for both experiments.
Variables
The dependent variable is the variable we are measuring. It is the numerical amount of how many letters each participant can remember.
The independent variable is whether a list of letters can be categorised into (chunks) meaningful units using the technique of chunking. By grouping phonetically sounded syllables (consonants and vowels list) into chunks and recalling.
Participants
The participants will be of equal ratios in terms of gender, The number of participants used will be twenty (ten male) (ten female). This will make the test fair and better to analyse.
They will all be within a twelve-month age band as they are in the same year eleven group.
The targeted population is of 15-16 year olds. The type of sampling used in the experiment is Opportunity sampling. This method was chosen because it's quick and more ethical. Also the year eleven group were the only participants available.
In order to make my experiment ethically correct I will follow the guidelines enlisted below.
) Each participant will be told the objectives of the investigation at the same time and be informed any procedures which can influence there decision.
2) All participants will be debriefed at the same time before and after the experiment is conducted. Also have consent of the participants.
3) There will be no deception in the experiment, as all information will be informed to the participants.
4) When analysing the results no names will be given in order to prevent experimental bias and follow the ethical guideline of confidentiality.
I will follow the 1990 Ethical Principles for conducting Research with Human Participants.
Apparatus
The following apparatus will be used in the experiment:
. An overhead projector will be used to display three acetate sheets. Making the test fairer in terms of the time limit on looking at the stimuli. The first sheet, the title of the experiment will be written. Prior testing will be done to see whether the information can be viewed clearly and precisely.
2. Stop watch to time how long participants can write and look at the stimulus material (i.e. the letters)
3. Pencil for the participants.
4. Paper for the participants with pre-printed allocated boxes for the letters for each test.
Procedure
. Debrief the experiment to the participants. Consent must be given from all participants before the experiment begins in order to make it ethically correct.
2. Set up the Overhead projector.
3. Ensure all participants are seated appropriately.
4. Switch the overhead projector on, placing the first acetate "psychology experiment" in place, making sure everyone has view, and switch off.
5. Pass out the papers with printed boxes for each letter.
6. Provide all participants with writing material.
7. Place test 1 with only the constinents on the overhead for 20 seconds
8. Allow 16 seconds for participants to write. Repeat the same procedure for the second test.
9. When the test is finished, ask the participants to briefly write behind their answer slip which memory technique they used.
0. Collect the slips.
1. Before dismissing the participants, explain the reasons (i.e. to find out if he technique of chunking works) for the experiment to follow ethical guidelines.
Controls
The experiment will be set in laboratory conditions this is an ideal form for the experiment as there is good control of the extraneous variables. The extraneous variable is important as it can influence the dependent variable. This variable is a matter, which is unintentionally likely to influence the experiment, by the participant. Examples of extraneous variables are emotions and stress, as the timings of the tests are close to exam times, it is important to note that these extraneous variables are likely to exist.
Results
Summary Table
The following table shows the results of the first test in our experiment:
A table to show the number of consonants scored by male and female participants in Test 1.
The number of consonants scored by female Participants.
The number of consonants scored by male Participants.
6
3
3
3
2
7
5
7
5
8
2
2
4
2
2
3
2
3
0
A table to show the number of consonants and vowels scored by male and female participants in test 2.
Female Consonants & vowels
Male consonants & vowels
3
8
8
7
6
6
3
4
6
5
3
1
0
0
1
6
5
9
0
0
Summary Table commentary
There is a valid trend in the data. At first sight, the male participants generally have a greater level of memory in both tests. In the first test of randomly selected consonants, the males had a greater recall level. The highest level was 8 consonants where the lowest was 0. We can comment that this could have been due to a wide range of abilities when choosing the participants. In comparison the highest number of consonants for females was 6 and the lowest is 2. We can now see that the highest consonants of the females are lower than the highest for the males. The lowest of the females is higher than the lowest of the males.
In the light of test two, in terms of males, the highest level was 16 consonants where the lowest was 7. This shows that the number of consonants for males was twice as much than in the first test. This showed that they remembered the consonants and vowels better in the second test. In comparison the highest number of consonants for females was 16 and the lowest is 5. We can now see that the highest consonants of the females are equal to the highest for the males. The lowest of the females is lower than the highest of the males.
A Chart to showing how many the male and female participants scored in Test 1. (Consonants)
The following chart illustrates the results of test 2 between male and female participants consisting of consonants with alternating vowels.
Descriptive statistics
Descriptive statistic
Female consonants
Male consonants
Female consonants + vowels
Male consonants + vowels
Standard Deviation
.36
2.66
3.26
4.03
Range
3.0
8.0
1.0
9.0
Mean
3.5
3.5
1.5
1.6
Median
3.0
2.5
0.5
0.5
Mode
2.5
2.0
3.0
3.0
Descriptive Statistics commentary
It can be shown from the table that the in the first test, the standard deviation is greater for the males. The deviation is higher in the second test. The mean shows that there is a large difference in the two tests. The mean is a lot higher in the second test than the first test. According to the median, in the second test, both males and females had the same level of results.
This shows that the participants have recalled more consonants in the second test, which coincides of random consonants alternate vowels. This shows that the participants had chunked the letters into small groups of words with phonetic sounds. This technique was used as after both experiments were conducted we asked the participants to write down any memory technique used. Most participants had chunked the phonetically sounded syllables down. Which made the text easier to process and recall.
Relationship of results to hypotheses
The number of participants tested was of 20 people. Consisting of 10 male and 10 female. The hypothesis states that a list of consonants with alternate vowels forming phonetically sounded syllables, can be recalled better than a list of letters consonants can be accepted as the statistics show that this is true in the results for both genders.
This shows that the null hypothesis can be rejected.
Discussion
Validity
Generally the study was a accurate measure of chunking the technique of chunking. It illustrates that chunking information down, influences the effect of recall. Which in turn helps to improve performance of recall which can be supported by the results.
The internal validity of the experiment set in a class room therefore if we had extra resources and finance we would set the experiment in a real laboratory settings, which would provide us a ideal and more reliable set of quantitative data. The external validity of the experiment concerns the ecological validity of the experiment. I feel that because the experiment was set in laboratory conditions, as it would lacks mundane realism therefore the experiment has a low external validity.
The extent of the internal validity (extraneous variables) was maximised as during the deigning of the experiment we took into account any experimental bias that could occur (therefore we didn't include any names when analysing the quantitative data produced). We also took into account the demand characteristics. The experimental situation (the classroom) prompted the participants to conduct the experiment in a more serious manner. This can act as an extraneous variable.
A further study could have been done into using different ages and different cultural backgrounds to discuss whether the concept of chunking can be applied, and make it more generalised to life.
An additional study could be conducted into recalling a list of letters mixed in with digits and test the recall of it.
The problems faced by this is that it will be difficult to find the sample participants.
Sperling (1960) produced set his experiment in laboratory conditions. In effect the data produced would therefore lack ecological and external validity, as this would not be done in a real life situation. For this reason a field or naturalistic experiment where the environment of the settings would be more natural for the participant so the participant would behave "naturally". Therefore this situation, would have an increase in ecological validity and therefore be more generalised to life.
Suggestions for improved validity
The validity of the experiment could have been improved by increasing the sample number of participants. Which would in turn make a more generalised conclusion to a larger population. The experiment could have been further valid by using participant from different cultural background and compare the results. The results would then indicate that the use of memory may not be effected by cultural backgrounds. By setting the experiment in more natural settings for the participants in order to make the experiment more ecologically valid.
The methodology of the experiment could have been improved by counterbalancing, to control the order of effects. Giving half the participants could have done TEST 1 first while the other half do TEST 2. This prevents improvements due to practise. Or poorer performance due to boredom.
Reliability
The reliability of the experiment refers to dependability of the experiment. If the experiment was replicated. Exactly the same results may not be replicated as the extraneous variables such as the stress levels of the participants due to GCSE exam pressure is a factor difficult to control. As the experiment was set in laboratory conditions there was a high level of controls, such as the time restriction for looking and writing the stimulus material. Therefore similar results can be concluded. The experiment could have received constant errors due to noise. Therefore it was more reliable for the experiment to be conducted in silence.
The procedure of the experiment was planned well and easy to carry out. The problems incurred in the methodology of the experiment were that all participants were pre seated before we arrived therefore we couldn't seat individuals preferably on there own or in a boy girl situation.
Participants was opportunity sampled which introduces a biased perspective to the findings as all participants live in the same county and were all in the same region of age (12 month gap). Therefore the results cannot be implied to real life.
Participants should have been more randomly selected; therefore the study would have been more ecologically valid.
The apparatus;- The stop clock used was needed to be more accurate. The methodology of timing the recall and looking at the stimulus material can influence the study .as the participants could have seen the stimulus material for longer then the required time. This endangers the procedure as participants can then have extra time for rehearsing the material.
Improving reliability
Mistakes were made due to human error .The timing situation could be more accurately. This could be achieved by using a projector and stop clock, which switched off itself. Using a more varied chose of participants in terms of age and nationality would be more ecologically valid.
An alternative methodological technique can be used to improve the reliability. This can be achieved by using a more varied sample of participants. Using participants from various different cultural backgrounds which will make the results more ecologically valid.
Also by using participants of different ages. The effect on the results will be that the older end may have a poor recall due to age. In comparison to the younger people who will have a higher recall rate.
Implications of study
The Magical Number Seven, Plus or minus two". Conducted by G.A Miller (1956) means that a person's short-term memory (which is equivalent to working memory) can hold from five to nine units of information. Millers' finding supports my study's research into chunking. The trend of the results show that all participants recalled at least five to nine units of information in the consonants test. As these findings supports my hypothesis.
In comparison to the second test consisting of vowels and constinents making a phonetically sounds four participants recalled fully all the letters. The study supports that list of consonants with alternate vowels forming phonetically sounded syllables forming meaningful units. Bower and Springston (1970) concluded similar results that meaningful chunks such as (FBI) are recalled better then chunks that show no meaning such as (TZP)) can be recalled better.
Generalisation of findings
The problems which may incur in generalising the results to populations outside the targeted population of fifteen to sixteen year olds is that, the sample of participants was undersized to generalise the findings. Also the participants used all came from the same county (Yorkshire) and most of the participants used were of the same cultural background. Therefore the results cannot be generalised as they will not apply to all people of those ages.
Applications to real life.
The results of the study focusing on the technique chunking can be applied practically to real life situations. As, chunking can be thought as a learning mechanism . It can be applied to learn a telephone number by breaking the digits down
E.g. 0800308408 can be chunked down into 0800 30 84 08.
The technique of chunking can also be applied in teaching. In order to help children find it easier to learn and therefore improve there performance.
References