Impact of the Phonological Loop on Short Term Memories

Introduction

Memory is an aspect of psychology that has been studied for years. In this study, the cognitive approach is focused on and it is discovered how the phonological loop can affect the way in which short term memories are formed.

A theory of memory is the Working Memory Model which was developed by Baddeley and Hitch’s in 1974. This model furthered on the one proposed by Atkinson and Shiffrin which was criticised for being oversimplified. It is a branch of short term memory which temporarily stores and manipulates a limited amount of information which is required in order to perform cognitive tasks. The Working Memory Model is said to perform a number of responsibilities and involves the central executive, the visuo-spatial sketchpad, episodic buffer and the phonological loop. The central executive is the working component of the model and is responsible for the three other components. It controls the attention and divides the roles up between each, whilst also coordinating the flow of information between long term memory and the working memory model. The visuo-spatial sketchpad has a limited capacity of storage and duration of retaining information and is responsible for anything visualised and the visual location of objects in space. Much like the visuo-spatial sketchpad, the episodic buffer can only retain a limited amount of information, however, it can hold things in four ‘chunks’. It can communicate with both long term memory and the other components of working memory whilst also acting as a ‘backup’ storage system.

The main component which is being focused on in this experiment is the phonological loop. This part of memory is responsible for dealing with spoken and written material. It is active when people are required to remember something as it will replay information over and over in order for it to be memorised. It temporarily stores a limited amount of verbal speech, such as sounds or words which are known as phonemes. This verbal information is held in sound-based form (phonological form).

In 2011, Landry and Bartling conducted an experiment on a group of thirty-four psychology students to investigate whether or not articulatory suppression would influence recall of a written list of phonologically dissimilar letters in sequential recall. The experiment was an independent groups design, where one group were asked to look at a selection of phonologically dissimilar letters for five seconds and then when instructed, they were to write down as many letters on the answer sheet as accurately as possible. This was repeated ten times. In the experimental group, participants received instructions to repeatedly say the numbers 1 and 2 from the time the letters appeared on the screen to the time they were asked to write them out. This was also repeated ten times. From this experiment, the mean percent of accurate recall in the control group was 76% percent while the mean in the experimental group was 45%. It was concluded that participants who used articulatory suppression were greatly affected as it prevented rehearsal in the phonological loop due to overload.

The aim of this experiment is to investigate in a class of year one IB students whether articulatory suppression of counting aloud continuously for a period of five seconds influences the efficiency of the phonological loop in the Working Memory Model. This is a partial replication of Landry and Bartlett’s 2011 prior investigation, as a group measures design.

The null hypothesis is that there will be no significant difference between the recall of letters in serial order in students who utilised articulatory suppression compared to those that do not use articulatory suppression, and any differences will be due to random variables.

The research hypothesis is that participants using articulatory suppression are more like to recall less words in order as opposed to those who do not use articulatory suppression.

The independent variable is whether or not articulatory suppression of counting aloud, saying 1,2, 1, 2, continuously for five seconds was used or not to interfere with the phonological loop. The dependant variable is the capacity of the phonological loop as measured by the number of phonologically dissimilar letters recalled in order.

Exploration

This experiment used a repeated measures design. This design was used to help avoid any confounding variables which could arise from individual participant differences due to the same participants being used for all conditions. In addition, this meant that the same participants were used for both the controlled and experimental groups. All participants were a motivated sample of first year IB students of the ages 16 and 17. This also meant that an opportunistic sample was used.

Materials were formulated. Prior to the experiment, participants were informed of the nature of the experiment and their ethical rights and responsibilities in the consent forms (Appendix I) that were handed out, signed and collected. On the day of the experiment, participants were seated in the classroom and given standardised instructions to ensure the unambiguity in the test. By doing this, it eliminated any experimental biases. Participants then completed a practice trial where they were given different letters from the ones which were used in the actual experiment. Once completed they were given letters that were phonologically dissimilar so that the letters did not rhyme. Each slide went for five seconds and when instructed, participants were required to write down as many letters in the same order in which they were displayed. This was the same method used for the experimental group, however, participants called out numbers ‘1’ and ‘2’ from the time the letters appeared on the board, till the time it was instructed to write them down. All trials were completed in the same room with the same people.

At the end of the experiment, participants were reminded of their right to withdraw. A debriefing statement (Appendix VIII) was issued to all participants once results were finalised, reminding them of their ethical rights and the results of the study.

Analysis

In order to compare the results of the participants who used articulatory suppression and those who did not, the mean number of letters was recalled as well as the standard deviation and percentage of letters that participants got correct.

Table 1: Descriptive analysis of data

The results show that the students who used articulatory suppression were able to recall a lesser number of letters than those who did not. In addition, it can be seen by looking at the standard deviation that there seems to be a lot more variance in the non-articulatory suppression results due to the larger error bars. In the experimental group, there was a mean of 8.5, while in the controlled group there was a mean of 12.1. This shows the variance in the results.

Figure 1: Mean number of letters recalled

A Wilcoxon Matched pairs signed ranks test was used because the test used a repeated measures design, the data was interval, and converted to ordinal, with a small sample size. This was used to determine whether there was a significant difference between the controlled and experimental conditions. The test was a one-tailed test as it was predicted that the participants who used articulatory suppression would recall a lesser amount than those who did not.  The data was significant at p<0.05, meaning that there is a less than 5% chance that the results are only due to chance. It is from this data that the null hypothesis can be disregarded, and the alternative can be concluded. 

Evaluation

It can be concluded from the results that were obtained that the same conclusions were able to be drawn as those from the original experiment conducted by Landry and Bartley. There was a slight variation in the percentage of letters that participants got correct, however, this could have been due to the different experimental designs that were used. In the Landry and Bartley experiment, an independent groups design was used, while in the replicated experiment, a repeated measures design was used. Additionally, it could have come down to the participants having a varying level of memory recall from the participants involved in the original experiment.

It was expected that the experimental group who used articulatory suppression would be affected and their phonological loop would ‘overload’ causing a lesser number of letters to be recalled. This was proven and shows that the phonological loop plays a significant role in the Working Memory Model when it comes to short term memories being formed.

One of the strengths of the study was that by using the repeated measures design it controlled potential confounding variables from arising due to individual participant differences. However, this can cause the problem of the practice effect which is when the same test is taken more than once. By doing this, it can increase a participant’s overall score because they have repeated the experiment a number of times and hence, get more practice. On the contrary, the participants’ score may also be decreased over time due to fatigue or boredom. A possible way of resolving this limitation could be to counterbalance. By doing this, it could help to control order effects in the repeated measures design as both groups would still be completing the same experiment, only in a different order which reduces practice effects. The reason as to why the independent groups design was not used in this experiment was because it would be significantly more time consuming and unless the participants are matched like in the matched participants design, it causes participant variables to arise due to different ages, genders and social backgrounds. Because of this, the repeated measures design was therefore used.

This experiment used an opportunistic sample which meant that it was a quick and easy way of selecting participants for the experiment. However, it raises the limitation of the sample not being a representative sample and there being potential biases. A possible way in which this could be improved is by using random sampling instead. By doing this, it would reduce sampling biases because the sample would represent a more accurate target population. However, a disadvantage to this would be it being very difficult to achieve due to time and effort. Despite this, a strength to using the opportunity sampling design was that it was the same as the one used by Landry and Bartley which helps to eliminate any differences when the experiment was replicated.

One of the limitations in the experiment was that the PowerPoint did not include in built timings. This raises the question of whether or not each slide was played for the same amount of time – five seconds. It is believed that each slide was, however, if one slide was played for a shorter or longer period of time than another it could cause a number of uncertainties to come into place which to some extent could have affected the results. A way in which this could be resolved in future is by making sure that the PowerPoint includes inbuilt timings so that any extraneous variables can be avoided.

Another limitation was when all of the group was counting aloud ‘1’ and ‘2’, there may have been some people who did not join into this and follow the instructions that were given. This therefore could have been a reason as to why some people achieved a higher score than others because articulatory suppression was not being used and hence the phonological loop did not go into ‘overload’.  In future, a possible improvement could be to test each participant individually. This would take a longer period of time; however, it would ensure for my accurate results.

Overall the experiment was very much successful due to all controlled variables maintaining controlled throughout all trials. This allowed for any extraneous variables to be controlled as much as possible. By displaying the instructions at the beginning of each trial it also allowed for participants to be informed and reminded of what was required of them, this helped to eliminate any sort of confusions.

For a future experiment, it could be interesting to use the same experiment and instead swap the letters that needed to be remembered with numbers and vice versa for when ‘1’ and ‘2’ are called out. By doing this, both experiments can then be compared, and it can be determined whether or not there is a difference in how the phonological loop operates with letters and numbers.

From the experiment that was conducted, it can be concluded that articulatory suppression of counting aloud continuously for a period of time does in fact influence the efficiency of the phonological loop in the Working Memory Model.

References

• Episodic Buffer definition | Psychology Glossary | alleydog.com. (2019). Alleydog.com. Retrieved 13 May 2019, from https://www.alleydog.com/glossary/definition.php?term=Episodic+Buffer
• InThinking Student pages. (2019). Student.thinkib.net. Retrieved 8 May 2019, from https://www.student.thinkib.net/psychology/page/23593/working-memory-model
• McLeod, S. (2019). Working Memory | Simply Psychology. Simplypsychology.org. Retrieved 8 May 2019, from https://www.simplypsychology.org/working%20memory.html
• Wilcoxon Signed-Rank Test Calculator. (2019). Socscistatistics.com. Retrieved 21 May 2019, from https://www.socscistatistics.com/tests/signedranks/default2.aspx
• 2019). Simplypsychology.org. Retrieved 8 May 2019, from https://www.simplypsychology.org/Working%20Memory.pdf

Appendices

1. A blank copy of your letter of consent.
2. Briefing notes/standardized directions
3. Materials – PowerPoint
4. Debriefing notes
5. Raw data
6. Calculations of inferential statistics