PS1010 Lab 3: Investigating the Serial Position Effect

PS1010 Lab 3 Handout V1 THREADGOLD 2018
PS1010 Lab 3: Investigating the Serial Position Effect
MODULE LEADER: Dr Victoria J. Willan VJwillan@uclan.ac.uk
LAB MAIN LEAD: Dr Emma Threadgold – EThreadgold1@uclan.ac.uk
Models of Memory
A consistent feature of almost all models of human memory is the idea that human memory can
be subdivided into several sub-processes that differ in terms of their capacity, duration and the
encoding of the information they contain. Typically such models make provision for three memory
processes:
• Sensory Memory
• Short Term Memory (STM)
• Long Term Memory (LTM)
Sensory memory, short term memory, and long term memory differ in their capacity (the amount
of information they can hold) their duration (the time they can hold information for) and also how
information might be lost from the store (known as forgetting).
Table 1: Capacity, Duration and forgetting in memory sub-processes.
These three memory processes were often referred to as memory stages. It was argued that
information is transferred from sensory memory to short-term memory and finally to long-term
memory. This serial processing model is often referred to as the modal model of memory (see
Figure 1 for the modal model of memory.

! 1
LONG
TERM MEMORY
(LTM)
SHORT
TERM MEMORY
(STM)
SENSORY
MEMORY

Figure 1: The Modal Model of Memory
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Case Studies of Amnesia – ‘HM’ and ‘KF’
Strong support for the modal model of memory can be found in case studies of amnesia,
particularly HM (see Wickens, 2005, for a general review of this syndrome). HM had a normal
short-term memory (STM) but a deficient long-term memory (LTM). Another man, KF (investigated by Shallice & Warrington,1972), had the opposite problem, his LTM was quite normal but his
STM was very poor. These two cases provide an example of what is known as a double
dissociation.
Double dissociation means that brain damage to two different areas affects memory in different
ways. In other words cognitive abilities which are related (different sub-types of memory) are
found to be located in different areas of the brain. Specifically damage to the temporal lobes,
including the hippocampus, impairs LTM (in the case of HM) whereas damage to the speech areas
impairs STM (in the case of KF). The KF case implies that information can flow directly from sensory memory to LTM without passing through short term memory. So Figure 2 shows that the modal
model might equally be represented as follows:
Be that as it may, both models give rise to an interesting question:
When participants recall information in short term memory tasks, how can we be sure from which
memory system the recalled information is derived?
The Serial Position Effect
The Serial Position Effect is a phenomenon characteristic of certain types of short term
memory task. The effect is most easily observed when participants are presented with a supra
span list (i.e. more than 9 items) of common words and are immediately asked to recall the
information in any order. You will recall that according to Miller (1956) short term memory has a
capacity of 7 +/- 2 items), so 9 items is over the typical short term memory capacity. In these
circumstances recall is never 100 per cent, however the probability of recalling a word is a function
of the serial position (1st, 2nd, 3rd, 4th etc) in which it was presented. If the experiment is
repeated across a large number of participants, and the percentage recall is calculated for each
serial position, the outcome, in this case for recall of a 15 word list, looks something like this:
! 2
SENSORY
MEMORY

SHORT
TERM MEMORY
LONG
TERM MEMORY
Figure 2: The Modal Model – An Alternative Representation
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Exactly the same effect is observed if the list consisted of 20 items, or 30 items. It is believed that
the serial position effect occurs because more than one memory system is involved. The
enhanced recall of items presented first (serial positions 1 to 5 in our example) is known as the
primacy effect and is thought to reflect the contribution of the long term memory system. The
enhanced recall of the items presented last (serial positions 11 to 15) is known as the recency effect and is thought to reflect the contribution of the sensory and short term memory systems. The
explanation for this is as follows:-
Primacy Effect: Participants are thought to adopt an incremental rehearsal strategy as the list of
words is presented to them. Suppose the lists consists of the items
……………alpha, beta, gamma, delta, epsilon, phi, iota, lambda,……….
The participants build up the list as it is presented to them by adding the new item. Thus one
could imagine that the participant’s sub-vocalisations are as follows:
“alpha”
“alpha, beta”
“alpha, beta, gamma”
“alpha, beta, gamma, delta”
“alpha, beta, gamma, delta, epsilon”
This works well as long as the participant has time to rehearse the list before the next item is presented. Eventually this is no longer possible and the participant has to abandon the strategy.
However, it does mean that in our example the first word, alpha, is heard once and repeated subvocally five times, the second word, beta, is heard once and repeated four times and so on. Now
! 3
Percentage Recall
0
25
50
75
100
Serial Position
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Figure 3: A line graph of the serial position effect.
PS1010 Lab 3 Handout V1 THREADGOLD 2018
the argument goes that the sub-vocal repetition of an item leads to the transfer of information to
long term memory.
The Recency Effect: The last items are present in short term memory and the very last may also
be in sensory memory. One theory of forgetting in short term memory is that the memory traces
of items decay quite rapidly within a few seconds unless they are refreshed, therefore the last
items have less time to decay before they can be recalled. In addition because the capacity of
short term memory is believed to be between 5 and 9 items, the last items are likely to displace
those that came earlier in the list. It is not hard to appreciate that items in the middle serial
positions may be poorly recalled.
There is, however, a counter argument to the one that has just been developed. This argument
hypothesises that all the items might be present in only one memory system (typically short term
memory). However they differ in terms of the strength of their memory traces. Thus the first
items have strong memory traces because they were subjected to varying degrees of maintenance rehearsal and the last items are also strong because they have had less chance to decay.
Now all we need to do is to assume that the probability of recall is a function of memory trace
strength and the U-shaped serial position curve will occur. It is a fair point and the more
parsimonious explanation of the two.
Distinguishing between the two hypotheses:
1. The single memory hypothesis argues that the primacy and recency effects are due to the same
underlying cause. So we would predict that if we could discover a manipulation that affected
memory trace strength we should observe its consequences in both the primacy and recency
effects.
2. If the dual (triple) memory hypothesis is correct then we might be able to manipulate the primacy and recency effects independently. The strongest support for the dual memory model
would come if we could find two manipulations – one of which reduced the recency effect whilst
leaving the primacy effect intact – and another which reduced the primacy effect and left the
recency effect unaltered. We would call this a double dissociation – there is evidence to support this and this is the path we are going to take.
Reducing the recency effect: We know that if items are left in short term memory and not subjected to maintenance rehearsal they are rapidly forgotten. This phenomenon is known as the
Brown-Peterson effect after the researchers who independently demonstrated the phenomenon.
Both gave their participants sub-span lists of letters (in the case of Brown) and nonsense syllables
(in the case of the Peterson’s) and then imposed a variable delay of between 0 and 30 seconds
before the participant was asked to recall. Both experimenters prevented their participants from
rehearsing during the delay interval by giving them an interpolated task (counting backwards in
threes from a random 3 digit number). According to the dual memory model delaying recall will
mean that items stored in short term memory will be rapidly lost whereas items in long term
memory will not.
! 4
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Imposing a delay between presentation and recall – what would we expect to see?
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………..
Reducing the primacy effect: Our rationale suggested that the incremental rehearsal strategy
(repeating words over) was responsible for items at the beginning of the list transferring into long
term memory. So if we can stop the participants from adopting the incremental rehearsal strategy
we should prevent this transfer and abolish the primacy effect. There are two possible strategies.
The first is to speed up the rate of presentation of the list, thereby leaving less time for incremental rehearsal. The other is to ensure that each item in the list receives equal attention. This can be
done by a fiendish manipulation known as the repeat n-1 technique. The participant is required to
respond to each item by repeating the item that occurred immediately before the item that has
been presented. Refer to Table 2 for the n-1 technique procedure:
Table 2. The N-1 technique participant and experimenter procedure.
! 5

PS1010 Lab 3 Handout V1 THREADGOLD 2018
Requiring the participant to repeat the previously presented item during presentation N–1.
What would we expect to see?
…………………………………………………………………………………………………………………
………………………………………………………………………………………………………………….
…………………………………………………………………………………………………………………..
Following both manipulations – one reducing the recency effect and leaving the primacy effect
intact (delay) and one that reduced the primacy effect and left the recency effect intact (N-1) if
successful, this is what we would see in relation to the control condition (free recall).
Key Terms – you should understand and be able to explain what is meant by each of the
following terms. Each of these is contained within your handout. If you are unsure of any of
these, please take another look through your handout!
Sensory memory
Short-term memory
Long-term memory
Double dissociation
The single memory hypothesis
The dual or triple memory hypothesis
Serial position effect
Primacy effect
Recency effect
N – 1 manipulation
Delay
! 6
Percentage Recall
0
25
50
75
100
Serial Position
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Figure 4: A line graph of the serial position effect with delay and N-1 manipulations.
PS1010 Lab 3 Handout V1 THREADGOLD 2018
PS1010 Lab 3: Investigating the Serial Position Effect – Questions
Experimental Design
What experimental design was used? …………………………………………………………………….
The independent variable was?……………………………………………………………………………
How many levels? 3 levels? …………………………………………………………………………………
The dependent variable was? ………………………………………………………………………………
Task Materials:
How many/what kind of words? ………………………………………………………………………….
Other wordlist detail ………………………………………………………………………………………
………………………………………………………………………………………………………………..
………………………………………………………………………………………………………………..
How were words presented? ………………………………………………………………………………
…………………………………………………………………………………………………………………
What was the rate of presentation? ……………………………………………………………………….
………………………………………………………………………………………………………………….
Mode of recall: how were words recalled? ……………………………………………………………….
Recording the data: What did you do? ………………………………………………………………….
Experimental Conditions
The conditions are listed below – but what did they involve?
1.Control condition. ………………………………………………………………………………………….
………………………………………………………………………………………………………………….
2. Delay condition ……………………………………………………………………………………………
………………………………………………………………………………………………………………….
! 7
PS1010 Lab 3 Handout V1 THREADGOLD 2018
3. N-1 condition.
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
Complete counterbalancing – With three conditions participants can perform the task in six
different orders:
1, 2, 3 – control, delay, N-1
1, 3, 2 – control, N-1, delay
2, 1, 3 – delay, control, N-1
2, 3, 1 – delay, N-1, control
3, 1, 2 – N-1, control, delay
3, 2, 1 – N-1, delay, control
There are 2 sets of 90 words – set 1 is used by experimenter 1 and set 2 by experimenter 2. Each
set of words is split into 6 word lists and these are presented in one of the above condition orders.
Raw data table: This will contain the scores for each participant for each serial position for each
condition. The data will be summed across participants for each serial position to give the total
words recalled in each serial position. This is done separately for each of the 3 conditions. This
will be presented in a serial position curve Graph: The total words recalled will be plotted against
serial position, separately for each condition (there will be three lines on your graph, one per condition).
The Serial Position Graph will be available on Blackboard in the Excel sheet – please use the
graph that has been created with YOUR group data. It will be labelled for your group (A1, B2
etc). and available after the lab session.
Copy and paste this graph it into your results section. Remember to appropriately label the axe,
give the graph a sensible and informative title, and refer to it in the text.
Statistical analysis
What is the primacy effect? ……………………………………………………………………………….
What is the recency effect? …………………………………………………………………………………
We have calculated the primacy and recency score for each participant in each of the 3 conditions (6 scores for each participant). Each participant will have a score in the following categories – primacy – control / primacy – delay / primacy – N-1 / recency – control / recency – delay / recency – N-1. This data has been entered into an Excel file for your group and will be
available on Blackboard, again please use the data for your group. This data can be copied
and pasted into an SPSS file. You need to set this up in the appropriate manner. Use your
notes from the ANOVA workshop to do this. Please note you will need to perform two separate ANOVA’s – one for your primacy scores (control / delay / N-1) and one for your recency
scores (control / delay / N-1). See Figure 5.
! 8
Figure 5: Performing the ANOVA analyses.
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Whilst predictions based on the single memory model would be that both portions of the curve
would be significantly reduced, if the dual memory hypothesis is correct then we might be able to
manipulate the primacy and recency effects independently. You need to develop a rationale to
support the dual memory hypothesis and take this into account when making your predictions.
You will conduct one analysis to look at differences between the control, delay and N-1 for the
primacy portion of the curve, and a second analysis to look at differences across the 3 conditions
in the recency portion of the curve. In other words you will complete two analyses, one for the
primacy (comparing control, delay and N-1) and one for the recency part (comparing control, delay and N-1).
Primacy – In order to demonstrate that repetition of the previously presented item during
presentation (N-1) only affects the primacy effect, and that the delay condition does not affect the primacy effect – we predict that (remember, you are looking at the differences in relation to the immediate recall control condition):
The number of words recalled in serial positions…………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
! 9
Percentage Recall
0
25
50
75
100
Serial Position
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Control N-1 Delay
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Recency – In order to demonstrate that delayed recall only affects the recency effect and that
the N-1 conditions does not affect the recency effect we predict that (as above – in relation
to the immediate recall control condition):
The number of words recalled in serial positions…………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
…………………………………………………………………………………………………………………
Each statement above will incorporate 2 predictions – thus there are four predictions, each
made in comparison with the control (free recall) condition.
PS1010 Lab 3: Investigating the Serial Position Effect
Writing your report
Abstract
Look at the abstracts in published journal articles for examples of correct style and content. An
abstract needs to be s short, succinct and clear overview of your entire research study.
Introduction
There are two important points that need to be made:
1. In our experiment we have chosen lists of 15 words and arbitrarily (but sensibly) defined the recency effect as the score for serial positions 11 to 15 and the primacy as the score for serial positions 1 to 5. However, we could have used 18 words in the lists and redefined the recency and
primacy effects as 12 to 18 and 1 to 6 respectively. The choice of 15 word lists is made in the
‘Method’. So in the introduction we need to avoid specific references to these figures and argue
and formulate our hypotheses in general terms which would be appropriate for any list length, not
just a length of 15 items.
2. We do not want a general essay on memory or even a review of research on the serial position
effect. To do the latter would be to pre-empt our own study. What is important is the development of your own rationale. As with any piece of work, you are encouraged to find your own
references, DO NOT SIMPLY COPY THIS HANDOUT! See module handbook, assessment handbook, and Blackboard for guidance on referencing and on plagiarism.
As a general guide, the structure of your intro could do the following:-
! 10
PS1010 Lab 3 Handout V1 THREADGOLD 2018
• Place the research experiment in context – briefly mention the concept of multiple memory
systems and set the general question of the relative involvement of these systems in short
term recall tasks.
• Introduce the serial position effect and argue that it does reflect the involvement of
separate memory systems and how this comes about.
• Develop the experimental rationale for independently manipulating the two components
of the effect.
• Very briefly, what did you do and then make predictions.
Method
You have answered questions above that relate to the methodology. You now need to pull this
together and present information in clear prose format in the four relevant sections (participants,
design, materials, procedure). Remember to pay attention to what information belongs in each
section, and be sure to include all relevant information, but as clearly and concisely as possible
without repetition. Remember, your methodology should allow replication of your experiment.
Guide your reader through what was done from when you, as the experimenter, began the experiment until you had finished your role as experimenter. You need to write as if you were the experimenter for the collection of all of the data. The instructions for the experiment should be
included in exactly the wording that they were given to participants.
Results
Please go to Blackboard to locate the data file and graph for your particular lab group. It is important that you run the analyses on your particular groups data. In relation to the graph, you must
say what the raw data consisted of and where it can be found. You must say how the graph was
plotted and include it in the body of the results section (NOT in the appendix). Make sure the
graph is properly labelled “Figure 1…” You should briefly describe the trends observable in the
graph (do not include interpretation of the results here) to guide you into the next section.
In relation to primacy and recency, you must say what you did with your data and then introduce,
title and present a table of means and standard deviations for the six conditions (primacy – control
/ primary – N-1 and so on). Remember to include standard deviations, and ensure you report all
figures to two decimal places, and p values to three decimal places.
After the table you need to move onto your statistics. What is the most appropriate analysis for
this data? You need to look back at your SPSS workshop 3 handout to guide you through the
analysis and write-up. You must report the results for each statistical test. If the ANOVA is significant, then you must also report the results of any post-hoc tests you undertake and do not forget
to report the direction of difference.
Discussion
Summarise the results of your statistical analyses, each test in turn – in words (do not repeat figures or statistics here). Did the control condition produce the classic serial position effect? What
effects did the N-1 and delay conditions have? Are these in line with predictions? Are there any
discrepancies?
! 11
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Link back to the research discussed in the introduction and discuss your results in relation to the
existing research. Do the results support the general belief that the serial position effect does reflect the contribution of more than one system? Did we doubly dissociate the primacy and recency effects?
Discuss any short comings of the experiment. Say why and contemplate alternatives – not just ..’it
would be interesting to see what happens if’…. .Are there any real world implications? You need
to pull it all together in a concluding paragraph.
Referencing
Remember to accurately reference both in text and at the end in your reference section. In the
‘References’ section, all references referred to in text need to be reported in alphabetical order.
Where you have two references by the same author, place them in chronological order from oldest to most recent. This is an example of APA format for a a peer reviewed journal article
reference.
Mednick, S. (1962). The associative basis of the creative process. Psychological Review, 69,
220-232.
Reading
In addition to the references listed below, please feel free to find your own references for the
write-up. General psych textbooks will cover the serial position effect, as will more specialist cognitive textbooks such as the ones on your reading list (Eysenck & Keane ‘Cognitive Psychology’
and Parkin ‘Essential Cognitive Psychology’). Alan Baddeley’s book ‘Human Memory’ is very good,
and there are lots of books of use in the library. You could also make use of google scholar to find
research articles. Remember to reference thoroughly everything you have used in your report. It is
normally a very bad idea to use websites. You should only use websites that are from a reputable
academic source and then must reference them properly (author, title, full URL, date, date accessed). However, the best advice is for you to steer clear of websites and use only published
academic sources and peer reviewed articles and books.
Warning regarding plagiarism
Do not copy from books/handouts/journals/the web/your friends in the group. You will work in
pairs to complete the data collection and results, but please ensure your report is entirely your
own work. If you are found to have copied from books/handouts/journals/the web/your friends in
the group, it will result in being brought before a disciplinary committee for plagiarism. Penalties
are severe. If you are in any doubts as to what constitutes plagiarism, consult the relevant section
of the handbook and/or arrange an appointment to talk about it with your personal tutor.
References
Baddeley, A. (Latest Edition) Human Memory. Hove: Psychology Press.
Baddeley, A. (Latest Edition). Essentials of Human Memory. Hove: Psychology Press.
! 12
PS1010 Lab 3 Handout V1 THREADGOLD 2018
Eysenck, M. (2001) Principles of Cognitive Psychology. Hove: Psychology Press.
Eysenck, M. & Keane, M. (latest edition). Cognitive Psychology: a students’ handbook. Hove:
Psychology Press.
Parkin, A.J. (2000). Essential Cognitive Psychology. Hove: Psychology Press.
Shallice, T. & Warrington, E.K. (1970). Independent functioning of verbal memory stores: A
neuropsychological study. Quarterly Journal of Experimental Psychology, 22, 261-273.
Wickens A. (Latest Edition). Introduction to Biopsychology. London: Prentice Hall
! 13

The post PS1010 Lab 3: Investigating the Serial Position Effect appeared first on My Assignment Online.

Expert Help With This Assignment — On Your Terms