It was found that all of the participants remembered more deeply processed words that shallowly processed words. This was a significant result as shown by the Wilcoxon Matched Pairs Test.
This lead to the conclusion that deeply processed words are better recalled than shallowly processed words.
Introduction
Research into the structure of the memory, and how information is passed from short term (STM) to long-term memory (LTM), includes that suggested by Atkinson and Shiffrin (1968). They stated that there were three memory stores, the sensory, STM and LTM stores. This was called the Multi-store Model.
However, there were some problems with this model. The model suggests that all information that is transferred from the STM to LTM store is done by active rehearsal. In spite of this, people are constantly storing information into the LTM store without actively rehearsing it.
Craik and Lockhart (1972) put forward an alternative theory to the Multi-store Model. This was the Levels of Processing theory. This suggested that constant rehearsal was not necessary for storing information in the LTM store. The Levels of Processing Model proposes that the level at which information is processed determines which store it will go into. If information is deeply processed, i.e., the information is processed semantically, it will be more easily remembered. Equally, information that is processed shallowly, i.e., acoustically or visually, will be easily forgotten.
This model was tested in a study by Craik and Tulving (1975). The idea was that if information has more meaning, it will be more unforgettable, and so would be stored for longer.
Participants were shown a list of words, and then asked questions about each of them. The questions came in three categories: case (to stimulate visual processing), rhyme (for acoustic processing), and sentence (to incite semantic processing). For example, the question, ‘Is the word in lower case?’ would provoke visual processing, the question, ‘Does the word rhyme with lock?’ would stimulate acoustic processing, and ‘Is the word a type of sport?’ is an example of a word that would incite semantic processing.
The participants were split into two groups to test two different situations in which Levels of Processing would be investigated – incidental learning and intentional learning. In the incidental learning study, participants were not given any informed instructions, but were just asked to follow the orders given by the investigator. In this case, the participants were shown the list of words, but were not told that they were to try and remember them. Equally, the participants in the intentional learning study were told the purpose of their activity, and so they intentionally remembered the words.
Craik and Tulving found that words that were semantically processed words were those best recalled followed by those recalled acoustically. Also, the intentional learners generally recalled more than the incidental learners, however both groups remembered more semantically processed words than any others.
The Levels of Processing theory can also be supported by a theory by Mandler (1967). He suggested that organisation creates a lasting memory. In a study, he presented a group of participants with 52 cards, each with a different word. The participants were asked to sort them out into categories as they saw fit, individually. They were then asked to write as many of the words down as they could remember. Mandler found that those who sorted their cards into more categories recalled more than those who had less categories. Those with more categories had processed the words more deeply than those who had fewer categories, and as the Levels of Processing theory would have predicted, they recalled more words than those with less categories.
A study by Hyde and Jenkins also supported the Levels of Processing Theory. They gave different groups of participants the same list of nouns. One group was told that there would be a memory test afterwards, which instigated intentional learning, and the other group was not told. Both were tested, and it was found that both groups recalled the same amount of words. This suggested that rehearsal may not have been necessary in order to remember words, but meaning may have been more important.
Aims and Hypotheses
The aim of this investigation is to find out if information that is deeply processed is better remembered than information that is shallowly processed, as suggested by the Levels of Processing Model. The original investigation was carried out in 1975, so I aim to see if the original findings will apply twenty years on.
Alternative Hypothesis: Deeply processed words, as defined by the method of processing (semantic), will be better recalled (more remembered) than shallowly processed words, as defined by visual processing.
Null Hypothesis: There is no difference between the number of words remembered by deep processing and the number of words remembered by shallow processing.
Method
Design
I used a lab experiment method and an independent groups design. The independent variable was the type of processing, i.e., deep or shallow. The dependent variable was whether or not the participant could remember the word. All participants were given the same questionnaire and the same amount of time in which to do it. Participants were not allowed to communicate with each other while completing the questionnaire, to ensure that each questionnaire is representative of one participant, and not a collaborative effort, which may have had an effect on the results. Questions regarding the purpose of their involvement were not answered until after the completion of the task. If the participants knew too much about the experiment, they may be able to guess the aim and could try to help or hinder the experimenter. Participants were given a full debrief of their actions to deal with ethical issues, such as deception (for example, some of the participants may have believed that they were being tested on their intelligence, rather than being an example of the way in which the memory works), and informed consent (the participants were not fully informed of what they were required to do, and so could not give informed consent).
Participants
I used an opportunity sample using students from the Allerton High School Sixth Form. The sample consisted of 12 participants, aged 16-17, both male and female. The participants were approached by the investigator, who then asked them if they were willing to take part. The participants were tested during an afternoon free period, in the school library and the sixth form common room. The investigators were two 17-year-old female sixth formers.
Materials
A questionnaire was used (see appendix), made by us. They consisted of equal numbers of questions that stimulated deep processing and questions that stimulated shallow processing. Each question asked the participant to consider a particular word and then give their answer.
Example
Is the word in capital letters? KITE Yes/No
In this case, the question asks the participant to consider the appearance of the word. The participant would be stimulated to process it shallowly in order to answer the question. The participant would then circle YES or NO accordingly.
A stopwatch was used to time the participants; standardised instructions (see appendix) and a questionnaire were given to them. The standardised instructions did not give details of the tasks the participants were to perform, as in they did not say, “Circle the correct answer and then write down as many of the words as you can remember,” as this would have given the aim possibly caused a change in behaviour.
Procedures
Participants were asked to sit separately from the other participants.
Participants were given a copy of the standardised instructions and a questionnaire.
The investigator instructed the participants to answer the questions in the questionnaire.
Participants were given 2 minutes to complete the questionnaire.
Participants were then instructed to turn over the questionnaire and write down as many of the words as they could remember in 1 minute.
The investigator collected their questionnaires and gave them a full debrief of the experiment.
Each of the completed questionnaires was examined, and a total of the number of deeply processed words recalled was found, along with that of shallowly processed words. Deeply processed words were those in the questions where semantic processing was involved, for example:
Is the word a type of flower? BEACH Yes/No
This word (BEACH) is a ‘deep’ word as it instigated deep processing, asking the participant to consider the meaning of the word.
Is the word in upper case? white Yes/No
This word (white) is a ‘shallow’ word, as it instigated shallow processing, asking the participant to consider its appearance.
The following page shows which words were ‘shallow’ and which were ‘deep’.
Results
The data summary (see appendix) shows that all the participants remembered more ‘deep’ words than shallow words. For some participants there was a significant difference in the type of words remembered, and for others there was a small difference.
Inferential Statistics
The Wilcoxon Matched Pairs Test was used to analyse the results and to see whether or not the hypothesis can be accepted or rejected.
Sum of least occurring sign rank = 0. Therefore the observed value of T = 0. N = 12 (there are 12 participants). The test is one tailed, as determined by the hypothesis being directional and p < 0.05 so the critical value of T is 17. 0 is less than 17; therefore, the hypothesis that deeply processed words are better remembered than shallowly processed words can be accepted (see appendix for calculations)
I chose the Wilcoxon Test as it is the test that best supports an experiment where:
The hypothesis predicts a difference between two sets of data (in this case the data would be the number of each type of processed word remembered).
The two sets of data are pairs of scores from one person (each participant produced a pair of results, one for shallowly processed words and one for deeply processed words).
The data is ordinal (the data can be ranked).
The level of significance was p < 0.05, as this is the significance levels used in almost all experiments.
Discussion
The results showed that participants remembered more deeply processed words than shallowly processed words. They also showed that the number of deeply processed words recalled was significantly greater than the number of shallow words. The results therefore supported my hypotheses, that deeply processed words are better recalled than shallowly processed words, and that there is a significant difference between the numbers of each type of words recalled. I found the results supported these hypotheses very strongly. Every single one of the participants recalled more deep words than shallow, and there was a minimum difference of 1 of the amount remembered of each type. I think these results occurred because the of the Levels of Processing theory. Semantically processed words were deeply processed and so were easily retrievable, but visually processed words were just shallowly processed and so were not as easily retrievable.
In relation to the previous study by Craik and Tulving, I think my findings are similar to theirs. They also found that the participants remembered deeply processed words more than shallow. My findings also support Craik and Lockhart’s Levels of Processing theory. Words that are processed with meaning, i.e., semantically, are deeply processed, and so are better recalled than shallowly processed words.
The findings do not support the Multi-store Model, as active rehearsal was not used. Participants were simply asked to answer the questions relating to the key words, and then asked to write down as many as they could remember. There was no opportunity for active rehearsal, so the information could not be transferred from the STM to the LTM as the model suggests.
The suggestion that arose from the study by Hyde and Jenkins, that rehearsal may not have been necessary in order to remember words, but meaning may have been more important, could be applied to my experiment. Again, the participants had no opportunity to rehearse the words, but still managed to remember some of the words where semantic processing was required, which also suggests that meaning is perhaps more important than rehearsal.
Mandler’s theory, that categorisation helps to recall more information, may also be applied to my experiment. The participants may be considered to have categorised the words in the ‘semantic instigating’ questions. For example, when asked, “Is the word a name?” or “Is the word a type of animal?” the words they were considering was categorised into names, animals, etc. However, when it came to the ‘visual instigating’ questions, the words could only be categorised into uppercase and lowercase words. As Mandler would have suggested, the semantic words were put into more categories than the visual words, and so were better recalled due to organisation.
However, there were some limitations to my experiment. The results were obtained under artificial circumstances. It is not often that people are asked to recall seemingly pointless lists of words, so we cannot know for sure if the findings of this experiment would be depicted in a more natural situation. Also, my sample was made up of only 16-17 year olds. The results may not be representative of other age groups. Also the time at which the experiment had taken place may have had an effect on the results. The experiment took place during the last lesson of the day, when students are arguably less focused. It could be that their lack of attention had an effect on the results.
As a result of these limitations, several follow-up studies could be conducted. A study could be carried out using members of many different age groups in order to see whether age may have an effect on the validity of the Levels of Processing theory. The study could also be carried out in a more natural environment, from which the findings could be put into practice in everyday lives. The study may also be tried during different periods of the day, e.g., during the morning, before lunch, after lunch, evening, etc. Another study could be carried out to see whether organisation or deeper processing is more essential to the recollection of information, as suggested by Mandler.
The findings of the study have some application to everyday life. It can be put into practice when revising for exams – more information will be remembered if you understand it, rather than simply memorise it.
In conclusion, the study found that information that is deeply processed is better recalled than information that is shallowly processed. It also found that the amount of deeply processed information is significantly greater than the amount of shallowly processed information.
Appendix
Standardised Instructions
- Sit in an isolated area, away from other participants.
- Read the instructions on the questionnaire regarding the first task.
- 2 minutes will be given to complete the first task.
- When the first task is complete, read the instructions on the questionnaire regarding the second task.
- 1 minute will be given to complete the task.
QUESTIONAIRE:
Circle the correct answer
Is the word in upper case?
Is the word a type of flower?
Is the word in lower case?
Is the word a fruit?
Is the word in lower case?
Is the word a boy’s name?
Is the word in upper case?
Is the word a sport?
Is the word in lower case?
Is the word an animal?
Is the word a type of building?
Is the word in upper case?
Is the word in upper case?
Is it the word a colour?
Is the word a type of building?
Is the word in upper case?
Is the word a person’s name?
Is the word in upper case?
Is the word a name of a sport?
Is the word in lower case?
white
BEACH
lion
LOVE
novel
book
HAIR
hockey
phone
DUCK
CLOWN
PAPER
clock
GREEN
fork
LIGHT
SILK
coat
TENNIS
pencil
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
YES / NO
Raw Data
Wilcoxon Matched Pairs Test
Least occurring sign rank is -
0 ranks are -, therefore the sum of the least occurring sign rank, and the observed value of T, is 0.
Using the critical value table, taking into account that the test is one-tailed and p<0.05, the critical value of T is 7.
In order for the result to be accepted, the observed value of T must be less than the critical value of T. 0 is less than 7, therefore the result is accepted.
References
-
Atkinson, R.C., & Shiffrin, R.M. (1968). Human memory: A proposed system and its control processes. In K.W. Spence & J.T. Spence (Eds.) The psychology of learning and motivation, Vol.2. London: Academic Press.
-
Craik, F.I.M., & Lockhart, R.S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behaviour, 11, 671-684.
-
Craik, F.I.M., & Tulving, E. (1975). Depth of processing and the retention of words in episodic memory. Journal of Experimental Psychology, 104, 268-294.
-
Hyde, T.S., & Jenkins, J.J. (1973). Recall for words as a function of semantic, graphic, and syntactic orienting tasks. Journal of Verbal Learning and Verbal Behaviour, 12, 471-480.
-
Mandler, G. (1967). Organisation and memory. In K.W. Spence & J.T. Spence (Eds.), The psychology of learning and motivation: Advances in research and theory, Vol. 1. London: Academic Press.