An Experiment To Investigate The Effects Of Semantic Context On Recall 

Introduction

The aim of the experiment was to investigate the effects of semantic context on recall. We chose to investigate schema theory as memory is a fundamental process that influences us on a daily basis. The notion that humans categorise and recall information to improve mental efficiency and capacity can aid research into the accuracy of memory and the potential influence external and internal factors can have on recall. As a group of students, memory and recall are vital to our success in our examinations. Investigating possible memory enhancing factors is an interesting and valuable topic to study.

Schema theory states that all knowledge is organised into units. Within these units of knowledge, or schemata, the stored information is associated or linked with each other in order to improve the efficiency of the encoding, storage, and retrieval of knowledge. According to this theory, schemata represent cognitive structures about concepts: objects, relationships between objects, situations, sequences, actions, language, imagery all loosely related in some manner. Bransford and Johnson (1972) investigated the efficiency of knowledge retrieval when a schema was introduced at various points. Participants in the experiment were read a series of sentences that were all related to a specific topic/schema and were then asked to recall them. In three conditions, the participants were either given the schema before the sentences were read out, before the recall, or not at all. This was to mimic the suggested 3 stages of memory systems; encoding, storage, retrieval. The first condition was the most successful; the participants correctly recalled the most sentences accurately, because they were able to encode the sentences in relation to their existing schema on that topic.

The cognitive process of fitting new information into existing schemas, perceptions and understanding is called assimilation; this is the process that the first group used to recall the most amount of sentences. The other process linked with schema theory is accommodation, which is the cognitive process by which you memory internalises new information by rearranging and assorting current existing schema in order to allow the new information to be incorporated. The results of this experiment suggested that initially using a topic header will increase the chance of accurate and better recall. Students can use this information to achieve better results in exams. If students header their work with a related word, the information they are encoding will automatically assimilate with that schema, when recalling, the topic word will help them find all associated information more efficiently. The experiment we carried out was a partial replication of Bransford and Johnson (1972) where we used two conditions, one in which the participants were given the semantic context prior to being read the statements and the second group of participants were not be given the semantic context. In this experiment, the Independent Variable was whether the participants were given the semantic context of ‘kite’ or not. The Dependent Variable was the number of correctly recalled statements out of 14. The one-tailed research hypothesis is as follows:‘Participants who are given the semantic topic ‘kite’ prior to listening to the recording will correctly recall ore statements out of 14 than those who are not given the topic at all’The null hypothesis is as follows:‘There will be no significant difference in correctly recalled statements out of 14 between those who are given the semantic context ‘kite’ prior to the test and those who are not given the semantic context ‘kite’ prior to the test. ’

Exploration

The research design we used is called independent measure. This is when each condition has its own set of unique participants in order to avoid demand characteristics and order effects. In our experiment, this was crucial as if we used the same participants they would already have heard the semantic context and the statements and would have had an advantage when recalling them. We used an opportunistic sample of 20 16-18-year-old students at Varndean College. We chose these people out of convenience. This was important as it meant participants were not expecting to be involved in a test and therefore could not prepare. Also, as the participants were all students, the results will be more accurate in suggesting ways in which students can improve their memory recall for examinations. They were then divided into two groups by a number system where each participant was labelled either one or two as they entered the experiment room and were then assigned a seat based on their number. This avoided any bias in pre-organising groups. In order to ensure the least amount of extraneous variables between the two groups, the statements were recorded by an automated voice and were put onto a timed recording instead of spoken or written. This meant that both groups heard the same voice speaking at the same pace and excluded the potential differences in emphasis and interpretation of facial expression that a live reading of the passage could employ. This also avoided any variations in reading ability/speed that could affect a participants willingness to continue with the experiment or ability to perform representatively of their condition. All participants were also exposed to the same environment, for example, the room visuals, temperature, and noise levels were maintained.

Furthermore, the instructions given were standardised and written not spoken. This meant that the instructions could not offer any clues by an audible emphasis on what was important or would help the participants gain a higher score. This is important as participants might guess the aim of the experiment by the tone or body language of the person explaining the task and therefore their results would not be representative of the condition. The statement we used were taken from the original Bransford and Johnson (1972) experiment. They were 14 statements based on the semantic context of the word ‘kite’. We used these as we knew they had already been used in an experiment that had significant results and therefore were reliable as a material. However, they also were useful as each sentence was meaningless without the context, this allowed us to see the real impact of having the schema to assimilate the information to. Procedure20 Participants are numbered and allocated a seat based on their number. Participants are asked to read the standardised instructions on the board and to read and complete the standardised consent form. 10 of the participants are then given the ‘semantic context’ by a card passed around with the word ‘kite’ on it. The 20 participants are then played a recording of 14 statements read allowed with 3-second intervals. When the recording is finished, the instructions ask them to write down the statements they can remember from the recording as accurately as they can. The participants are then given a ANALYSIS

Descriptive Statistics

Condition A: Participants are given the semantic context of ‘kite’ Condition B: Participants are not given the semantic context of ‘kite’MeanMedianSD σCondition A5. 352. 1931712199461Condition B3. 231. 6613247725836The mean number of correctly recalled statements tells us the average ability of either the participants who were given the semantic context or the participants who were not. This allows us to compare all the data collected. As there were no anomalies, the mean will be an accurate representation of the central tendency. The standard deviation is important as it tells us the spread of the ability of each condition and it uses all the data collected. This allows us to determine how strong a central tendency the mean demonstrates. As you can see, the results of our experiment suggest that Condition A fared much better than Condition B. We have supplied evidence to suggest that our research hypothesis is correct in predicting that the participants who were given the semantic context of ‘kite’ prior to hearing the recording would recall more sentences than the participants who were not exposed to the topic. Inferential statisticsIn order to clarify whether the difference between Condition A and Condition B is statistically significant, we chose to do an inferential statistics test. We used the Mann Whitney ‘U’ test because the data is at least ordinal and the design was the independent measures degree. Statistical findings are interpreted with regard to the data and linked to the hypothesisThe calculated value 21. 5 is lower than the critical value 27, the null hypothesis is rejected (p<0. 05), there was a significant difference in correctly recalled statements out of 14 between those who are given the semantic context ‘kite’ prior to the test and those who are not given the semantic context ‘kite’ prior to the test.

Evaluation

Findings discussed - refer to model/theory(relevant)The calculated value 21. 5 is lower than the critical value 27, the null hypothesis is rejected (p<0. 05), there was a significant difference in correctly recalled statements out of 14 between those who are given the semantic context ‘kite’ prior to the test and those who are not given the semantic context ‘kite’ prior to the test. strength/limitations of design - stated and explained(relevant)strengths / limitations of sample - stated explained(relevant)strengths/limitations of procedure - stated and explained(relevant) - in same space knew they were missing something.