The Huron University College Journal of Learning and

Motivation
Volume 50 | Issue 1 Article 6

2012

A Special Place in Our Minds: Examining the Serial

Position Effect
Alec Lowe

Recommended Citation
Lowe, Alec (2012) "A Special Place in Our Minds: Examining the Serial Position Effect," The Huron University College Journal of
Learning and Motivation: Vol. 50 : Iss. 1 , Article 6.
A special place in our minds: examining the serial position effect

A special place in our minds: examining the serial position effect

Alec Lowe

Huron University College

Previous research in free-recall tasks have demonstrated a

"serial position effect" in people's recall; that is to say words
around the beginning and end of a list are more readily
recalled than those in the middle. In the present study
participants were read out one of two 18-word lists, one
where the words either shared a common conceptual
category and where the words did n o t It was expected that
words a serial position effect would demonstrate, and that
mean word recall for the common concept list would be
higher than the mean recall of the non-common concept list.
The results of this study confirmed the hypothesis regarding
the serial position effect, but mean recall was not
significantly higher in the common concept list; however
ANOVA found a significant interaction effect between list
type and a words position on the list

The subject of memory has, for some time, been a subject of interest in the world of

psychological research, more specifically our short-term memory. Perhaps the first

researcher in the field was Hermann Ebbinghaus (1964). His Magnum Opus

"Memory: a Contribution to Experimental Psychology" (1964), includes a series of

free-recall experiments he conducted to discover whether the position of a word

on a list affected how easily i t could be recalled. In these experiments he pre-

recorded lists of "nonsense syllables" (e.g. DAX) then played them to himself and

tried to recall as many words as he could (Ebbinghaus, 1964). Ebbinghaus found

that words around the beginning and end of the list were more readily recalled

than those around the middle; he characterized this as the "serial position effect"

(manifests as a "serial position curve" on a graph) (Ebbinghaus, 1964). He

A special place in our minds: examining the serial position effect

explained this tendency to remember words around the beginning and end of the

list was the product of human long-term and short-term memory processes. In

essence, he postulated that words at the beginning of the list were more easily

recalled because they had more time to be committed to long-term memory

(Known as the "Primacy Effect"); while words at the end of the list were more

readily recalled because there was a greater chance of them still being in short-

term memory (known as the "Recency Effect") (Ebbinghaus, 1964).

Glanzer and Cunitz (1966) further examined the serial-position effect and

both long-term and short-term memory; more specifically, Ebbinghaus'

postulations that short-term and long-term memory are the cognitive processes

underlying the primacy and recency effects (what cause the serial position curve).

To test the hypothesis that long-term memory was responsible for people's

tendency to remember words around the beginning of the list during free-recall

memory tasks (where participants are read a list of words then asked to recall),

Glanzer and Cunitz varied the interval between the words as they were read off the

list and number of times the list of words were presented to participants before

they were asked to recall the list (Glanzer and Cunitz, 1966). They argued that

these processes have been demonstrated to be reliable and valid methods of

improving long-term memory and would affect the first portion of the serial

position curve (Glanzer and Cunitz, 1966), as i t is this section of the list that would

most likely be committed to long-term memory. They found that both increasing

the time between when each word is read out and the number of times a list was

read to a participant, yielded a significant improvement in recall, with regards to

A special place in our minds: examining the serial position effect

the first section of the serial position curve, but did not affect the average word

recall in the later two thirds of the serial position curve (Glanzer and Cunitz, 1966).

These results provide support for long-term memory being the process behind the

first part of the serial position curve, and therefore part of the serial position effect.

Glanzer and Cunitz (1966) also wished to test Ebbinghaus' hypothesis that

short-term memory was the process that underlies the recency effect To do this

they manipulated the time between when the experimenter read out the last word

on the list and when the participant was asked to recall the list (Glanzer and Cunitz,

1966). Although past studies have shown that the mere passage of time does not

necessarily mean that the words will exit the participant's short-term memory, as

they can rehearse the words so they stay fresh in their minds (Glanzer and Cunitz,

1966). To account for this issue they had participants perform simple math

problems during the delay (if they were in one of the experimental conditions)

(Glanzer and Cunitz, 1966). They found that a longer delay tended to decrease the

average amount of words (around the end of the list) that participants could recall.

These results lend support to Ebbinghaus' contentions that short-term memory is

the process behind the recency effect, and therefore latter part of the serial

position curve.

Previous studies in the field of memory have also focused on whether words

that belong to a common conceptual category are easier to recall. Tulving and

Pearlstone (1966) presented participants with lists of words (12, 24, and 48

words) where the amount of words that shared a common conceptual category

varied. Participants were then immediately tested on recall with or without the
A special place in our minds: examining the serial position effect

category name present; they found that the recall in the presence of the category

name was better than when the category name was not displayed (Tulving and

Pearlstone, 1966]. These results seem to indicate that having a list of words belong

to a common conceptual category can facilitate recall.

The current study seeks to see if a serial position effect will manifest when

students at Huron University College are read out an 18-word list This study will

also examine if a list of words that shares a common conceptual category is more

easily recalled than a list that does not share a common category, and whether

there is an interaction between a common conceptual category and a words

position on a list that effects recall. It is expected that the serial position effect will

be demonstrated, that is to say people will more readily recall words at the

beginning of the list and end of the list (first 6 words and last 6 words). Given the

previous results i t is expected that students in the common conceptual category

will perform better. This will likely manifest as an increase in the average recall of

words in the middle of the list, resulting in a less pronounced serial position

effect/serial position curve.

Method

Participants

The present study consisted of 20 students at the University of Western

Ontario; more specifically with 10 of these 20 are officially students at Huron

University College, one of the University of Western Ontario's affiliate colleges. 14

of the participants were female while the remaining 6 were male. 16 of the

participants were psychology majors while the remaining 4 were either majoring
A special place in our minds: examining the serial position effect

in philosophy or political science. In this study the participant's ages range from

18-22. Participants were approached by the researcher and asked to participate in

the study, and accepted based on their willingness to participate. Consequently

people were not assigned to groups randomly, but rather conveniently; that is to

say students were asked if they wished to participate in a psychology study and

upon agreeing were placed in whichever group had less participants at the time.

Materials

Two materials in this study were an information sheet/consent form and

debriefing letter. This study also had two 18-word lists that were read out to the

participants. One of the lists contained words belonging to a common conceptual

category, animals, while the other lists words did not share a common conceptual

category. Both lists' words were chosen from a list of 925 nouns located in the

appendix of Pavio, Yuille, and Allan's (1968) study "Concreteness, Imagery, and

Meaningfulness values for 925 Nouns". This was done to control concreteness

(how easily something can be referenced to sensory experience e.g. a table would

have high concreteness as most people have seen a table), imagery (how readily a

word can arouse non-verbal images, for example table would also score high in

imagery because when people hear the word they can easily picture a table in their

minds eye), and meaningfulness (refers to the average number of written

associations a person can think up within 30 seconds after hearing a word) (Pavio,

Yuille, and Allan, 1968). This was done to restrict confounds on the between

subjects variable of common conceptual category. Subjects were also provided

with a piece of paper numbered 1-18 for participants to write their answers on.
A special place in our minds: examining the serial position effect

Procedure

Researchers approached students in a classroom setting and asked them if

they were willing to participate in a psychology study. If they agreed, they were

given a letter of information/ consent form to read and sign, respectively. After

reading the letter of information participants were asked if they were ready to

proceed; after indicating they were ready, the researcher read the list of words out

to the participant with a one second gap in between the presentation of each word.

After the list was read out, subjects were given a sheet (numbered 1-18) to write

answers on and were asked to recall as many of the words on list that they could.

Subjects were told that they did not have to answer in any order so long as the

word they wrote was part of the list they were read. After all of the data was

collected, participants were scored on each third of the list. That is to say if they

recalled 3 words out of the first 6, 2 words out of the second 6, and 4 words out of

the third 6 then that person's score for the first third would be 3, 2 for the second

third, and 4 for the third of the list The average scores for each third of the last

were calculated for both the common conceptual category group and the non-

common-conceptual category group and graphed.

A 2x3 mixed factor repeated measures design ANOVA was conducted via

SPSS to determine whether the main and interactions effects of the variables were

statistically significant Planned contrasts in the form of T-tests were also

conducted through SPSS to confirm which means were significantly different.

A special place in our minds: examining the serial position effect

Results

The results of this study are displayed in figure 1. In both the common

conceptual category and uncommon conceptual category groups, average recall for

the first and last 6 words of the list is higher than the average recall for the middle

six. In the common conceptual category there is a less pronounced difference

between the average recall for the last 6 words and middle six words (2"'' section

mean recall was 2.4, while mean recall for the third section recall was 2.6), than in

the uncommon conceptual category (2"'^ section mean recall was 1.3, while 3'"''

section mean recall was 3.6). Also i t should be noted that average recall for each

section, albeit the final 6 words, for the common conceptual category list was

higher than average recall for the uncommon list (again, with the exception of the

third section which was higher for the uncommon list). Mean recall for all three

sections on each list was 2.9 (Common) and 2.5 (Uncommon).

A 2x3 mixed factor repeated measures ANOVA was conducted, via SPSS, to

assess statistical significance of the results. Maulchy's test of sphereicity was

insignificant, consequently sphereicity was assumed. The ANOVA found there to be

a significant within subjects main effect of word position, with an obtained f value

of (f(2,36)=6.86, p<0.05 two tailed). The ANOVA also determined there to be no

between-subjects main effect of common conceptual category, (f(l,18)=1.46,

p>0.05 two-tailed). However there was a significant interaction effect between the

variables of position on list and common conceptual category, (f(2,36)=5.03,

p<0.05 two-tailed).
 4 1

3.5 -

T3 Q
<U .3

^ 2.5

2 •

•Common
^ 1.5

Uncommon
1 •

0.5 -

0
1 2 3
Section of List
Figure 1: Average number of words correctly recalled out of the 1st, 2 n d , and 3rd sections (six words in each section) of
the common and uncommon conceptual category lists
A special place in our minds: examining the serial position effect

Follow up T-tests were conducted on the combined mean word score for

each third of the common and uncommon conceptual category list to determine

which third's means were significantly different. The means for number of words

recalled for the different thirds of both lists combined are: 3.05 for the first 6

words, 1.85 for the second 6 words, and 3.1 for the final 6 words. A dependent t-

test found there to be a significant difference between the mean of words recalled

in the first six and second 6 words, (t(19)=3.6, p<0.05 two-tailed). The means of

the middle and last 6 words differed significantly, (t(19)=2.6, p<0.05). However

there was no significant difference between the mean recalled of the first and last 6

words of both lists, (t(19)=0.114, p>0.05 two-tailed).

Discussion

Previous research regarding free-recall tasks have managed to demonstrate

a serial position effect, or in other words that people will more readily recall words

at the beginning and end of a list rather than those located in the middle.

Furthermore it was expected that mean recall on the common concept list would

be higher than the mean recall of the non-concept list. The results of this study

seem to confirm this hypothesis as the graph demonstrates that, in both categories,

people recall more of the first and last 6 words than the middle 6. It does not

confirm he hypothesis that mean recall for the common concept list would be

higher than the mean recall for non-concept list

The t-tests results showed that the differences between the means of the

first third/last third, and the means of the second third were significant; however

the difference between the means of the first third and last third of the list were
A special place in our minds: examining the serial position effect

found to not be significant. This demonstrates that a serial position effect

manifested and not because of sampling error; a difference between the means of

the first and last thirds of the lists would preclude the researcher from making such

an argument.

The results do not support the hypothesis that average recall was higher in

the list of words that shared a common conceptual category than recall on the list

that did not. However as a significant interaction effect between the variables

position and variable (manifested in figure one, where the uncommon list's serial

position curve intersects the common list's serial position curve) is perhaps an

explanation for why no significance was found for the main effect of common

conceptual category.

The results of the current study seem to, for the most part, exhibit scientific

adequacy. The serial position curve has been demonstrated in Glanzer and Cunitz

(1966) study and subsequently replications; additionally the ANOVA demonstrates

that the results have statistical significance. Furthermore the word lists were

controlled for "concreteness", "imagery", and "meaningfulness" (using Pavio, Yuille,

and Madigan's (1968) study) to mitigate confounds. A notable, but perhaps minor,

shortcoming in this study was the lack of random assignment of participants.

Future studies in this area should randomly assign participants to a condition with

a random number generator to make i t a random sample; as the current study was

contingent on willingness and availability of participants, this was not an option.

Future studies could further improve the strength of the results by increasing

sampling size and perhaps controlling gender of participants to an equal

A special place in our minds: examining the serial position effect

proportion (as the current study had a disproportionate number of female

participants).

Future studies could test whether increasing list size varies the curve of the

serial position curve. It is likely that the longer a word list the steeper the serial

position curve, that is to say there should be a more significant difference between

the average number of words recalled in the first and later parts of the list in

relation to words located around the middle of the list Future research could also

focus on the interaction effect between common conceptual category and a words

position on the list. More specifically why the mean number of words recalled

around the beginning of the common list was higher in the common conceptual

category (expected), but mean recall for the last 6 words was higher in the

common list (unexpected). As the ANOVA found the interaction effect to be

significant this certainly requires further investigation.

Research in this study is consistent with previous findings, with the

exception of the interaction effect between the variables of common concept and

position on list Future research needs to be conducted to examine the nature of

the interaction between the variables. As memory is a quintessential part of our

existence, should we not try to understand as much of it as we possibly can?

A special place in our minds: examining the serial position effect

References

Ebbinghaus, H. (1964). Memory: A contribution to experimental psychology. New

York: Dover Publications.

Glanzer, M., & Cunitz, A. (1966). Two storage mechanisms in free recall Journal of

Verbal Learning and Verbal Behavior, 5(4), 351-360.

Paivio, A., Yuille, J. C, & Madigan, S. A. (1968). Concreteness, imagery, and

meaningfulness values for 925 nouns. Journal of Experimental Psychology,

76(1, Pt.2), 1-25.

Tulving, E., & Pearlstone, Z. (1966). Availability versus accessibility of information

in memory for words. Journal of Verbal Learning and Verbal Behavior, 5(4),

381-391.
Appendix: Raw Data Table and Word Lists
Table 1. Raw data for uncommon and uncommon concept lists [10 participants for each list)

List Mean recall for First 6 words Mean Recall for Second 6 words Mean Recall for Third 6 words

Common 3.6 2.4 2.6

Uncommon 2.5 1.3 3.6

Alligator

Beast

Caterpillar

Fox

Elephant

Cheetah

Horse

Insect

Whale

Frog

Lobster

Mammal

Warbler

Beaver

Leopard

Bird

Python

Butterfly
Flag

Examination

Ink

Elbow

Jury

Landscape

Piano

Newspaper

Lemon

Painter

Vessel

Swamp

Tobacco

Yacht

Valley

Window

Rod

Plank