A“ 7 .
‘Y'JVH-

 

téfiflfitija‘huufi. .

 

3 . .
4‘“th

'z.‘ :‘a {V -:

'«~.‘%i~*’--.i;~t3§%s:-. A41“ :3

1.9.. «aw --v--.m -—- M1
' "r: '3. . .v‘z H
g..,j»§.r.§,ufig‘ v .E

1%..“ 42;}:“tu4‘

- .’ L

1‘ .. "7:43. :i . u."- wm”? ,.
'71:: 'Efé‘féfw‘ ‘ -. (zit-gt“ "‘
“ ”$3.5; £ " L. ' " ”31°

.w Y! 7 ~ . . 5.;

2:332;
Riff ‘t
‘1 1}

. X25
. . r . .
.. q in
953.. 5;”;

. "$.32:
‘ ~37“
3:. u

2‘1". (2’9
-. -\...L
. 3. "33337513 >
.2: $qu t-‘

,1 . ‘
1;: Ila? ’17
r .7", 41$'I‘.’~€:3:’ v
.13,“ . .,‘ "L '
' -, .54”:
w

Ci
1 ,‘
<52

m .
Huv'

. .

«a
«Jaw‘fi'js
, ; } .aw .i .1: {xi
“‘3"? ' ~. ' 2;. 2
L. ”‘5‘

134...,”
L: ‘. r '3';

:56“

This is to certify that the

dissertation entitled

The Learning of English Words by Japanese Seventh-Grade
Students by Means of the Keyword Method

presented by

Mamoru Kinjo

 

has been accepted towards fulfillment
of the requirements for

PhoDo degreein EngliSh

IV .7 A l

Major professor

 

 

 

Date May 11, 1994

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY
Mlchlgan State
Unlverslty

 

 

 

PLACE ll RETURN BOXtomnovow-chockoutflomywncord.
TO AVOID FINES rotumonorbdoroddoduo.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU tummmmuowuylm
mi

THE LEARNING OF ENGLISH WORDS BY JAPANESE SEVENTH-GRADE

STUDENTS BY MEANS OF THE KEYWORD METHOD

By

Mamoru Kinjo

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of English

1994

Dissertation Advisor: Dr. Paul E. Munsell

ABSTRACT
THE LEARNING OF ENGLISH WORDS BY JAPANESE SEVENTH-GRADE
STUDENTS BY MEANS OF THE KEYWORD METHOD
By

Mamoru Kinjo

The keyword method is advocated as a way of acquiring vocabulary items of
a foreign language (Atkinson, 1975). Research on the effects of the keyword
method has shown mixed results. Those in the field of cognitive psychology have
overwhelmingly reported success from their laboratories, whereas those in the field
of TESOL have mostly reported failure, when the keyword method is used with
actual students studying a foreign language as a subject (Fuentes, 1976; Liu, 1978;
Willerman & Melvin, 1979). The mixed results seem to want further empirical data.
Therefore, the following study was designed to provide such data.

The subjects of the study were 390 seventh-grade Japanese students (12
classes) studying English in Okinawa, Japan. All of the students studied two sets
of 15 English words—one set with the keyword method and the other with the rote
method. All 390 students took a posttest immediately after their study. Then 187
students (six classes) took the same test after a month, and the remaining 203

students (six classes) took the test after two months.

Mamoru Kinjo

The students’ scores were subjected first to the Wilcoxon matched—pairs
Signed-ranks test and then to the t-test. The results revealed that at the time of the
immediate posttest. the mean score of the words learned by the keyword method
was superior to that of the words learned by the traditional rote method.

The follow-up tests Showed that all 12 classes had higher means on the
words learned with keywords.

Among three groups of students with perfect scores on either set of 15 words
on the posttest (100% by rote, the keyword method, and both methods), mean
retention was superior for the keyword set on the follow-up tests, without exception
in each group.

On five words designated as ”hard,” the rote method showed a slight
advantage at posttest, which disappeared by the first follow-up test. A greater
difference was observed at the time of the second follow-up test in favor of the

keyword method.

Copyright by
MAMORU KINJO

1994

DEDICATION

This dissertation is dedicated to the following people, who have played an
important part in preparing me for my studies at Michigan State University. First and
foremost my mother,

Sumako Kinjo,
and my two older brothers,

lwao and Kaoru Kinjo.

And the following educators:
Ritsko Shinzato (at Miyamori Elementary School)

Shunichi Tamamori (at lshikawa Junior High School)

Dr. Mary E. Green
Dr. Robert E. Shafer
Dr. Daniel Quirk

(all at Arizona State University).

ACKNOWLEDGMENTS

I wish to express my deep appreciation to Dr. Paul Munsell, my major
professor, for his guidance, encouragement, and friendship throughout the course
of this work spanning almost a full ten years. Without his decade-long. patient
assistance, this work could never have been done.

I am equally grateful to my committee members, Dr. Susan Gass, Dr. Teresa
Tavorrnina. and Dr. Marilyn Wilson, for their invaluable contributions. I would also
like to thank the outside reader, Dr. John Eulenberg, for his helpful suggestions.

I also would like to express my appreciation to thejunior high school teachers
and students in Okinawa who participated in the study.

Lorraine Hart, graduate secretary in the English Department, who always
helped me in the best possible way.

The staff at the English Language Center, John and all the others, were very
kind and helpful whenever I asked for help.

Faouzi Azzouz, graduate student at Michigan State University, offered
invaluable time and statistical assistance beyond the call of duty.

Two typists, Cathy Rynbrandt and Susan Cooley Miller, assisted me in ways

I cannot recall without gratitude.

vi

Dr. Joseph Cousins at the Office of International Education Exchange, helped
me in many ways. His family, Keiko—san, Micah, and Riku, have always given me
a warm welcome.

I also thank the Owen Graduate Center and its staff for the kind cooperation
and consideration shown to me, which provided me with the best accommodations
for the study.

Hiromitsu lshikawa, who was at the Language Center at Naha, Okinawa,
gave me much advice and encouragement in my pursuit of the doctoral degree.

Some international students/friends such as Rajesh Lal and Klengolo Traore,
and many American friends, were willing to help me and made my study at Michigan
State University pleasant.

I would like to acknowledge with the greatest gratitude the contributions of the
three organizations that made my studies in the United States possible.

The United States Civil Administration in the Ryukyus for the four-year full
scholarship that sponsored the study at Arizona State University (1968-1972). (The
Ryukyus were returned to Japan in 1972.)

The Okinawa Prefectural Government Human Resources Development
Foundation for the two-year full scholarship that made possible my graduate study
at Michigan State University (1984-1986), and the fellowship given to continue

research at MSU (March to June 1991 ), which led to the acquisition of this degree.

vii

TABLE OF CONTENTS

Page
LIST OF TABLES ................................................ xi
LIST OF FIGURES .............................................. xii
Chapter
I. INTRODUCTION ...................................... 1
Purpose of the Study ................................... 1
A Brief History of Vocabulary-Acquisition Research
Since the Beginning of the Twentieth Century .......... 1
The Vocabulary Control Movement .................. 2
Fries, Lado, and Chomsky ......................... 3
The 19703 ..................................... 4
The 19803 ..................................... 5
Three Categories Given by Carter and McCarthy
(1988) ....................................... 6
The Keyword Method and the Problems of Its
Applications to Classroom Settings ................. 7
ll. REVIEW OF LITERATURE ON THE KEYWORD METHOD ..... 9
Review of the Literature Before Atkinson’s Work ............. 9
The Role of Pictures ............................. 10
Recent History ....................................... 11
Atkinson (1975) and the Introduction of the Keyword
Method ....................................... 13
Review of the Literature of the Post-Atkinson Period ......... 16
Application of the Keyword Method to Actual
Classroom Situations .......................... 17
Developmental Constraints on the Usage of the
Keyword Method .............................. 19
The Contextual Approaches Compared With the
Keyword Method .............................. 21

viii

Effects of Elaboration on Memory .................. 23

Metamemorial Study ............................ 26
Application of the Keyword Method to Teaching
Content Materials in Various Subjects ............. 27
Mnemonics and the Keyword Method in TESOL ............ 29
Areas Needing Research ......................... 35
Hypotheses ......................................... 37
METHODOLOGY .................................... 38
The Subjects ........................................ 38
Reason for the Choice of Subjects .................. 38
The Classrooms ................................ 39
The Role of the Teacher in the Experiment ........... 39
The Experiment ...................................... 40
An Outline of the Experiment ........................... 41
Selection of English Words to Be Used in the
Experiment .................................. 41
Order of Presentation of the Two Groups of
Words ...................................... 45
Selection of Meaning of English Words .............. 45
Pronunciation .................................. 46
Keyword Production ............................. 47
Illustrators ..................................... 48
Explanation of the Experiment ..................... 49
RESULTS .......................................... 53
Purpose of the Study .................................. 53
The Hypotheses ..................................... 53
Description of the Data ................................ 54
The Four Kinds of Data in the Experiment ............ 54
Explanation of the Pretest ........................ 55
General Outline of the Experimental Findings ............... 56
Terms ........................................ 56
Data From the Entire Population ................... 57
Results of Hypothesis Testing ........................... 66
Hypothesis 1 ................................... 66
The Wilcoxon Test .............................. 68
The t-Test ..................................... 71
Comparison of the Difference ...................... 72
Hypothesis 2 ................................... 76
Hypothesis 3 ................................... 79
Chapter Conclusions .................................. 80

ix

V. DISCUSSION ....................................... 83
Introduction ......................................... 83
Possible Causes ..................................... 84

Why Does the Keyword Method Facilitate the
Immediate Acquisition and Retention of
Foreign Words Better Than the Rote Method? ....... 84
Why Does the Keyword Method Facilitate the Long-
Term Retention of the Acquired Vocabulary
Items? ...................................... 89
Studies of Imagery .............................. 92
Summary of the Discussion ....................... 95
Classroom Applications .......................... 96
Limitations of the Keyword Method and the Experiment
and Suggestions for Future Researchers ............. 99
Limitations Inherent in the Keyword Method ......... 100
Language Limitations ........................... 101
Single Exposure ............................... 101
Length of the Experiment ........................ 101
Keyword Effect on Hard Words ................... 102
Comparisons of the Verbal Keyword and the
Imagery Keyword ............................ 102
Comparisons of the Keyword Method With Other
Well-Known Methods in TESOL ................. 103
APPENDICES
A. Handouts .......................................... 104
B. The Results of the t-Test .............................. 123
C. The Data on the 12 Classes ........................... 126
D. The Data on Students With Perfect Scores on Either
Half of the Immediate Posttest or Both Tests .............. 128
E. The Data on Hard Words ............................. 131
F. Letter of Approval From UCRIHS and Principal Consent
Form ............................................. 134
BIBLIOGRAPHY ............................................... 136

LIST OF TABLES

Table Page
1. Frequency Distributions of Students’ Scores on the

Posttest and Follow-Up Test ................................. 58
2. Differences in Scores Between the Posttest and the

Follow-Up Test With the Keyword Method ....................... 62
3. Differences in Scores Between the Posttest and the

Follow-Up Test with the Rote Method .......................... 63
4. The Number of Students With Perfect Scores on Either Half

of the Immediate Posttest or Both Tests ........................ 65
5. Comparison of Differences Between the Posttest and the

Follow-Up Tests: Hard Words ................................ 67
6. Means for the 12 Classes on the Posttest and the Follow-Up

Test for the Keyword and Rote Methods ........................ 75
7. Results of t-Tests for Difficult Words ........................... 81

xi

LIST OF FIGURES

Figure Page
1. Illustration of How the Keyword Method Works ................... 15
2. Frequency Distribution of the Immediate Posttest ................. 59
3. Frequency Distribution of the Follow-Up Tests ................... 60
4. Frequency Distribution of Difference ........................... 64

xii

CHAPTER I

INTRODUCTION

This chapter states the purpose of this study. It also reviews briefly the
vocabulary-acquisition research since the turn of the century to provide a historical
context for the present research regarding the keyword method. The chapter also
discusses the problems of applying the results of keyword-method experiments to

students actually learning a foreign language.

Eumeseefihefludy

The purpose ofthis study was to conduct an empirical investigation of certain
aspects of the keyword method and, in particular, to see whether the method can
enhance the remembering of English vocabulary items. The population used was
390 Japanese seventh graders learning English in Japan. The results were
interpreted statistically, and a theoretical account is presented as a possible
explanation of the findings.

. . _ . 'i
AfiLLeiflmQ[mmm;Kim-ma[y—Am1AJ5Min—Baiemms'IIB'i[III I'IICI
One of the historical characteristics of vocabulary acquisition in foreign

language teaching during the twentieth century is that it has received widely varying

1

2
amounts of attention. In the middle part of the century there was a period of neglect,
possibly through the influence of American linguistics represented by as diverse
scholars as Fries and Chomsky, whose primary attention centered on syntax.

Reflecting on the neglect of research on vocabulary of the past decades,
Carter and McCarthy (1988) observed that “the period 1945-1970 was a limbo for
vocabulary as an aspect of language teaching in its own right" (p. 41). They
concluded their review of the period by mentioning that "[a]n article such as Higa
(1965) is a notable exception . . . in anotherwise unredeemed neglect" (p. 41). Allen
(1983) gave one reason forthe neglect of vocabulary between 1940 and 1970 as too
much emphasis on words in language classrooms in the years before that time. To
illustrate these factors, two quotations, one from the 1920s and the other from the
1940s, are provided below.

First, Sapir (1921) warned linguistic students that they “should never make
the mistake of identifying a language with its dictionary“ (p. 234). Second, Bond,
Babcock, Norman, and Andrus (1943) cited 24 categories and 1,145 entries found
in Dales’s tabulation of studies on vocabulary selection through 1938 as a reason

to doubt that in vocabulary research "any stone has been left unturned" (p. 150).

IbeMecabuIaDLQQntmlMevement

One significant area of great inquiry during this period of vocabulary neglect
is now known as "the vocabulary control movement." Two important concepts have

emerged from this movement and deserve a short explanation.

3

1. A minimum basic vocabulary as suggested in the work BaSiLEnglish
by C. K. Ogden (1930) and l. A. Richards (1943). The underlying principle for the
W list was that complex ideas can be expressed by means of simple
words. The authors identified 850 words as those elemental, yet sufficient,
vocabulary items that can fully meet the basic communication needs of foreign
learners of English. Although Dammit) is no longer widely used, a concern for
the construction of ”core or nuclear Englishes for language learning purposes”
(Carter & McCarthy, 1988, p. 6) still is an issue of vital importance in the field of ESL
or TESOL today.

2. A limited defining vocabulary recommended by Michael West and
Harold Palmer. West (1953) published WM (GSL), which
contained 2,000 words selected from a corpus of two to five million words. This list
was a product of almost three decades of major work on vocabulary selection for
teaching purposes. Carter and McCarthy (1988) maintained, with regard to the GSL,
that despite some of the criticisms, it "remains one of the most innovative examples

of foreign language pedagogy and Iexicometric research in this century" (p. 9).

Wham

When, during the 19403, researchers engaged intensely in the vocabulary—
control movement and Bond et al. claimed that vocabulary study had hardly left any
stone unturned, a new movement was in embryo. Fries (1945) published his

W in which he stressed that

difficulties in learning a foreign language should not be considered to be limited to

4

the lexicon. He stressed the importance of mastering the sound system and
grammatical structure in learning a new language. However, he did not totally
neglect the importance of the role of vocabulary in learning a foreign language. This
fact is reflected in his student’s work; Lado (1955), a decade after Fries’s publication,
dealt with the difficulty of vocabulary from the standpoint of contrastive analysis of
languages.

The field received enormous support for the emphasis on structure with the
birth of transformational grammar 12 years later when Chomsky (1957) published
his Wm. With subsequent developments in the two fields, linguistics

and TESOL both came to concur on the primacy of syntax in language study.

3121.91.05

It was not until the 19703 that voices against the long neglect of vocabulary
research began to be heard. One of these voices was that of Wilkins (1972), who
argued, "Without grammar very little can be conveyed, without vocabulary nothing
can be conveyed" (p. 111). The massive investigation by Thorndike (1973) using
100,000 students in three different age groups from 15 different countries proved
that the knowledge of vocabulary is the surest index to reading comprehension.

It was also in the 19703 that people started "viewing vocabulary learning as
a language skill. shifting the responsibility to the learner,” rather than uniformly
providing a list of words to learn based on frequency (Carter 8 McCarthy, 1988, p.

42). Even so, Meara (1980), reviewing the literature concerning vocabulary

5
acquisition, could still title his article "Vocabulary Acquisition: A Neglected Aspect

of Language Learning."

Ihe_1_9_805

In the 1980s, topics in vocabulary acquisition research developed in
increasingly divergent directions. Two important publications appeared in 1980,
according to Carter and McCarthy (1988). These are Quideflnasimfimahulam
leaching by the Regional English Language Center (RELC) in Singapore and
Meara’s survey article on vocabulary acquisition, which was already mentioned as
part of the literature of the 19703. As for the RELC publication, a list in one of the
articles contains eight "C’s" and a “G" of vocabulary teaching. They are quoted here
to Show the varieties of topics discussed:

(1) collocation, (2) clines (e.g., scales such as 221mm expressed

diagrammatically), (3) clusters (something akin to the Hallidayan notion of

Sets—see Chapter 2), (4) cloze procedures, which reinforce cluster and

collocation, (5) context (using features within the text such as definition, word

analysis, inference), (6) consultation (using dictionaries and thesauri), (7)

cards: students should keep a card index of new vocabulary, (8) creativity:

students should be given free rein to describe pictures, etc. and (9) guessing:
students should learn this skill.

Two annotated bibliographies were published by Meara. The first volume
came out in 1983, covering the years from 1960 to 1980; the second volume was
published in 1987, covering the period from 1980 to 1985.

Words were studied increasingly in their collocational possibilities, as seen

in the discussion of "prefabricated speech" (Nattinger, 1988). of "composites"

(Cowie, 1988), of "institutionalized clauses” (Peters, 1983), and of "sets" (Carter &

6
McCarthy, 1988). The period illustrated also attempts to integrate important findings
from other fields, such as psycholinguistics, as evidenced by the studies of Meara
(1984), the Birkbeck Vocabulary Project headed by Meara (1982), Channel (1988),
and from the field of linguistics (Carter 8 McCarthy, 1988).

Renewed interest in traditional word-frequency counts was seen in the 1980s
in the compilation of the “University Word List” of about 750 words (Xue 8 Nation,
1984). Nation (1990) claimed that mastery of the words in their list, together with the
2,000 headwords in the WIS. should "provide around 95 percent
coverage of academic texts” (p. 189).

Global analyses of words in their multiword contexts led some researchers
to suggest that a clear distinction between vocabulary and syntax could sometimes
collapse (Ard 8 Gass, 1987; Peters, 1983). Publications by Carter (1987), Carter
and McCarthy (1988), and Nation (1990) gave added strength to vocabulary
teaching and research in TESOL. By the end of the 1980s, Gass (1989) could say
that ”it is no longer possible to say that the lexicon is the ’neglected component’ of
second language acquisition research," but, observing the divergent directions the
research had taken, she summarized the state of the art in the 19805 in a single
phrase: "lack of focus."

II III 'G'IEI
W

Although there seems to be a lack of focus in the field of vocabulary

acquisition, Carter and McCarthy (1988) in Wing

7
mentioned three topics in which lexical-acquisition researchers have taken an active
interest. These are (a) memory, (b) word difficulty, and (c) interlanguage. In a
passage about word difficulty, Carter and McCarthy mentioned that the difficulty of
a word may result from “whether it lends itself to key-word teaching techniques."
They further stated:
One important element in learning new words is the degree of effective formal
linking learners can establish between a word in the target language and a
cognate word in the mother tongue (and because issues of memorization and
word difficulty are closely connected). According to research by Craik and
Lockhart (1972), oral repetition is not necessarily an effective way of

assimilating new words; recalling the form of a word is found to be more
productive.

WW
II! I' I' III! all

The ”key-word teaching techniques" that Carter and McCarthy referred to,
also called the keyword method, is the subject of this research. The problems of the
keyword method are discussed in the following paragraphs.

The keyword method was formally introduced by Atkinson (1975) as a
mnemonic technique to enhance the memory of a foreign word. Since the
introduction of the keyword method by Atkinson, there has been much research
concerning the keyword method among cognitive psychologists within the confines
of laboratory settings or scientifically controlled settings. These investigations
generally have suggested the effectiveness of the keyword method. In addition to
these reports of the experiments, however, there have been some experimental

reports with actual students learning a foreign language as subjects. These results

8

have not turned out to be replications of the success most often reported by keyword
advocates. For example, one of the first studies reporting the results of an
experiment conducted outside a laboratory setting—that is, in an actual classroom
setting with students learning a foreign language—was that of Fuentes (1976). His
conclusion was that the keyword method had no effect on increasing the learning of
foreign words, as compared with the traditional rote method. Therefore, Fuentes
argued, the results of the laboratory experiments do not necessarily transfer directly
to classrooms of actual foreign language learners.

Several other experiments have been carried out with actual students of a
foreign language as subjects, but the reports have shown mixed success and failure.
Consequently, the effects of the keyword method with students learning a foreign
language in actual classroom settings have been farfrom fully explored. Therefore,
the present Study was designed to provide empirical data about the effect of the
keyword method with Japanese seventh-grade students learning English as the
study population. Further discussion of the background of the keyword method and

specific issues to be studied are presented in the next chapter.

CHAPTER II

REVIEW OF LITERATURE ON THE KEYWORD METHOD

The review ofthe literature on mnemonics and the keyword method is divided
into four sections: the literature before Atkinson’s work, Atkinson (1975) and the
introduction of the keyword method, the literature of the post-Atkinson period, and

that of the applications of mnemonics and the keyword method in TESOL.

E . [ll IT I BE 9|l'i::’ill!::li

Although the present-day keyword method used for acquiring vocabulary in
foreign languages was originally promoted by Atkinson (1975), the basic principles
of mnemonics and the keyword method date back to the heyday of the ancient
Greek orators. In brief, the keyword method uses images as an aid for memory of
a vocabulary item; this technique of creating mental images as an aid for memory
was widely known to the ancient Greeks. It is said that Simonides of Ceos invented
the classic method of Loci, and it was handed down to the Romans. One of the
extant discourses on the subject of Loci from the ancient days of Rome is that by
Cicero in his De Oratore, which dealt with the art of memory as one of the five parts
of rhetoric. There have been many treatments of the method by other writers,

among which Quintilian’s is said to be the clearest, and Ad Herennium "the only

9

10

complete source, for the classical art of memory both in the Greek and in the Latin
world" (Yates, 1966, pp. 1-5).

Furthermore, although the keyword method makes even abstract words in a
foreign language concrete and visualizable to assist memory, this distinction
between abstract and concrete types associated with Paivio (1969) had been
discussed by Cicero, Augustine, and Aquinas. Even the concept of levels of
processing (Craik 8 Lockheart, 1972), which is often given as one explanation of the
efficacy of the keyword method, was discussed by Francis Bacon (Gorfein 8

Hoffman, 1987, p. 53).

W

On the other hand, the actual use of pictures and images in teaching
languages is not as ancient as mnemonics, according to the author of 25 Centuries
of Language Teaching. "Teaching through pictures was done by the ancient
Chinese. In the West it is not found consistently before the Renaissance, though
there is some evidence of it during the Middle Ages” (Kelly, 1969, p. 15).

The Orb/s sensual/um pictus by Comenius (1654) is said to be "the first fully
thought-out scheme of teaching vocabulary with pictures" (Kelly, 1969, p. 17).
Comenius recommended five steps in using the book:

1. The pupils were to familiarize themselves with the book itself.

2. They were to make sure they knew the vernacular names of everything
depicted in the book.

3. If possible, the teacher was to show them the real thing.

11

4. The pupils were to copy the illustrations.
5. Finally, they were to color in their own copies and even the etchings
in the book.

These steps illustrate Comenius’s idea that language learning is an active
process employing as many senses as possible. In other words, pictures held the
possibility of expanding the context in the course of language learning and teaching.
However, ”two centuries of almost total neglect of pictures followed Comenius”
(Kelly, 1969, p. 18). Then there was a renewed interest in his works at the end of
the nineteenth century, which produced a reissue of his writings. However, the
effect was negligible, for his work was regarded more as an inspiration for the
teacher than as materials for students.

Kelly (1969) also pointed out that, during World War I, the United States Army
received many conscripts from minority groups, who did not speak English and were
ignorant of many things concerning European countries. Language teachers used
photographs cut primarily from the National Geographic Magazine to teach them
about languages and real things. This was a notable development as the teaching
put into practice the basic approach of Comenius to language teaching, taking

advantage of pictures as an aid in language teaching.

Becemhistery
As for the historical origin of the keyword method itself, Desrochers and Begg
(1987) placed it, based on their study, "between the epoch of the first celebrated

modem mnemonist, Raimond Lulle (1236-1315), and the end of the 19th century”

12

(p. 56). They mentioned two related figures from the last century, Fenaigle and the
Reverend J. H. Bacon. "Fenaigle’s (1813) New/4’1 of Memory, probably the most
complete account of previous mnemonic systems," devoted a whole chapter
(Chapter V) to the learning of foreign languages, emphasizing the utility of cognates
for memorizing new foreign words. Here, one can see that the value of associating
an unfamiliar word form with a familiar one was recognized. As for the Reverend J.
H. Bacon, then at the St. Bees College of Cumberland, England, he is known to
have advocated what he called the "phonetic link method" for the study of French
vocabulary.

Paivio and Desrochers (1981) further mentioned Loisette (1895) and Sayer
(1877) as those working along a similar line. They gave an example Sayer used to
teach the Latin word rana("frog"), using what he called "the method of assimilation,"
as seen in the following sentence: "A frog caught in the rain" (p. 780).

Desrochers and Begg (1987) summarized their study of the history of the
keyword method up to the end of the nineteenth century by writing that "what
remains uncertain is when and by whom the second ingredient of the keyword
method, the semantic link between the two concepts, was introduced."

In studies of memory in the early twentieth century, occasional references
were found to the utility of the keyword method. Desrochers and Begg (1987)
mentioned Estabrooks (1927) as one, and Paivio and Desrochers (1982) referred

to Julliot (1919).

13

In the late 1960s, a series of experiments was conducted to help retarded
children. According to Martin (1978), who gave a summary of the results, mnemonic
verbal mediators were found to help the children’s recall.

Paivio and Desrochers (1981) cited, among the advocates of something
similar to the keyword method in the early 1970s, Furst (1972), who labeled his
method the ”linking word” method, and Lorayne (1974), who named his the
”substitute word” method. About the latter author Paivio and Desrocher wrote:
”Lorayne proposed specifically that foreign words be learned by associating them
with picturable English words that have phonetic resemblance to the foreign word."

It was not until the 19705, however, that serious empirical investigations on
mnemonics were conducted by psychologists. Otto, Butler, Blake, and Ball’s (1973)
report on the effectiveness of the keyword method was the first of such experimental
assessments (Desrochers 8 Begg, 1987, p. 58).

9|)“ “925: III II I l' f
Wind

”The publication of Atkinson’s studies,” according to Pressley, Levin, and
Delaney (1982), "set the stage for all subsequent keyword investigations" (p. 64).
Accordingly, one ofAtkinson’s (1975) articles, "Mnemotechnics in Second Language
Learning,” is reviewed in detail here.

In the introduction to his article, Atkinson explained that he was involved in
pedagogical application of psychological theory to optimize computer-assisted

instruction. One of the experimental projects he was engaged in happened to focus

14

on “the restricted but well-defined problem of optimizing the teaching of foreign
language vocabulary" (p. 821). A series of vocabulary learning experiments he
conducted with students at Stanford University displayed large between-subject
differences, even among this apparently select sample of American university
students. Atkinson’s search for strategies to help students led to a series of studies
on mnemonic techniques. In concluding the introduction, Atkinson mentioned the
contribution of his colleague, Michael Raugh, who had been intrigued by the power
of mnemonics and convinced him that “this line of research was worth pursuing” (p.
821)

Atkinson then showed how the keyword method works, first explaining the
term "keyword" as follows: ”By a keyword we mean an English word that sounds like
some part of the foreign word.” Then two links follow, integrating acoustic and
imagery components as quoted here: "The spoken foreign word is linked to the
keyword by a similarity in sound (what I call the acoustic link) and in turn the
keyword is linked to the English translation by a mental image (what I call the
imagery/ink)" Atkinson’s actual examples of the keyword method then follow. One
of them is the following: "The Spanish word for duck is para (pronounced something
like "pot-o"). Using the English word potas the keyword, one could image a duck
hiding under an overturned flower pot with its webbed feet and tufted tail sticking out

below.” (See Figure 1.)

15

 

PATO — pot — DUCK

Figure 1: Illustration of how the keyword method works.

Atkinson then reported the results of an experiment teaching 120 Russian
words to 52 students at Stanford University. The students learned 40 words a day
for three consecutive days, and took a test a day after the last day, and again after
six weeks. The results showed that the keyword group recalled 75% of the items,
whereas the control group recalled 46% of the items on the first test. On the second
test after six weeks, the keyword-method group recalled 43% of the items, and the
control group recalled 28% of the items.

Atkinson then discussed the following five questions on the basis of the data
available from pilot experiments and other research in progress: (1) Should the
experimenter supply the keyword, or can the subject generate his own more
effectively? (2) Does supplying the imagery link for the subject facilitate learning?
(3) When a foreign word is presented, does the time to retrieve its English translation
depend on the method of learning? (4) Are the imagery instructions critical in the

keyword method, or can the subject do equally well when told to associate the

16

keyword and English translation by generating a meaningful sentence connecting
the two words? (5) How useful is the keyword method if the subject is asked to
retrieve the foreign word when given its English translation?

In the next section, Atkinson deals with the effectiveness of a keyword.
Atkinson recommended a committee approach in generation of the keyword, as
'[elxperience indicates that individual experimenters can come up with some pretty
bizarre keywords that work for them but for no one else.” He then points out that the
words students learn through keywords will, when mastered thoroughly, form a direct
link between the foreign words and the native translations. Therefore, the keyword
method should be as effective as any other ways of learning foreign vocabulary.

In the concluding section, Atkinson stated, ”If our instructional applications
prove as successful as the experimental work," then the keyword method deserves
a role in Ianguage-leaming curricula.

In the following pages is the review of literature on the keyword method so as
to Show the general trend of the experiments conducted from 1975 to the present.

They are arranged primarily chronologically to highlight the developmental patterns.

B . [II IT I [II E l-Ell' E . I

Since the introduction of the keyword method by Atkinson (1975), there has
been active research in this area of teaching, or remembering vocabulary items
through the method. Some of the interesting topics of investigation are reviewed in

the following paragraphs under these headings:

17

1. Application of the keyword method to actual classroom situations.

2. Developmental constraints on the usage of the keyword method.

3. The contextual approaches compared with the keyword method.

4. Effects of elaboration on memory.

5. Metamemorial study.

6. Application of the keyword method to teaching content materials in
other fields.

These divisions and the order of the topics were created partially on the basis
of chronological order of research developments, but studies reviewed in them do
not always conform to the chronological restriction because of the topical divisions:
Studies done many years apart may appear in a group because of their thematic
proximity.

. .
AW 9 I | [1| ET I'

The experiments by Atkinson (1975) were followed by those of Fuentes
(1976); Liu (1978); Levin, Pressley, McCormick, Miller, and Shriberg (1979); Merry
(1980); and Pressley, Levin, and Delaney (1982), who investigated the applicability
of the keyword method to a real classroom situation with mixed results.

Fuentes (1976) compared the keyword method with a rote-rehearsal group.
One hundred thirty-five students taking second-year high school Spanish were
asked to learn five new words given ten minutes on the first four days of the school

week. He found no difference in performance between the two groups.

18

Liu (1978) found that the keyword method of teaching vocabulary to first-year
high school students in Taiwan proved to be ineffective. However, Pressley, Levin,
and Delaney (1982) criticized this report because it provided insufficient details
concerning procedures and the results.

Willerman and Melvin (1979) taught 40 French words in a beginner French
class. But their results did not indicate the significant differences keyword-method
advocates would expect.

Levin et al. (1979) conducted a series of experiments to investigate the
classroom application of the keyword method. They confirmed the claim of Fuentes
(1976) with high school students, but succeeded in producing striking keyword
effects with elementary-school-age children.

Merry (1980) reported five different ways of teaching French words in a real
classroom setting. The findings showed that the keyword method showed
superiority over all the other methods for both tests, one immediately after the
learning session and the other a week later. However, there was no difference
between keywords used with pictures and keywords used with verbal images.

Pressley, Levin, and Delaney (1982) provided a comprehensive review of the
keyword-method literature, covering more than 200 studies with a view to potential
applications ofthe keyword method to teaching a variety of curricular contents. One
of the sections included a review of group-administered keyword studies as follows.

The results of the experiments reviewed showed a mixture of success and

failure. Fuentes (1976) and Willerman and Melvin (1979) were criticized; Merry

19
(1980) was cited as a report in favor of the keyword method. The authors reported
”additional failures of the keyword method among high school students according
to a self-paced, group-administered format" in their laboratories.

Two general observations were made from the analysis of these mixed
resuns:

1. Based on their findings of the failure with high school students, the
authors suggested that less sophisticated learners may benefit more from the
keyword method, such as students in elementary schools or junior high schools.

2. The overall results of the application of the keyword method seem to
show that laboratory experiments cannot necessarily be directly applied to real
classroom settings with actual students as subjects of investigation.

Qeyelepmentalfimstraintscmhe.
Wham

Various stages in the cognitive development of children interact with the
results of the keyword method. Some of the results of studies on this topic are
reviewed below. One of the earliest studies of the keyword method with children
was conducted by Pressley (1977). His subjects were second— and fifth-grade
students. Students in both grades benefited from using the keyword method.

Pressley and Levin (1977) examined the differences between fifth graders
and ninth graders on the basis of Rohwer’s hypothesis concerning adolescents’
change in strategy in learning noun pairs. A majority of the fifth graders reported

that they simply rehearsed, whereas few ninth graders relied exclusively on

20
rehearsal. Pressley and Levin (1978) further reported thattheirsixth-grade subjects,
when provided with the keyword, could generate their own imagery links, but their
second-grade subjects could not create images of their own.

A successful way to help second-grade students was reported by Pressley,
Levin, and McCormick (1980): they used a sentence variation of the keyword
method. In their experiments, second-grade subjects were asked to make up a
sentence, integrating in their mind the keyword and a meaning in order to remember
the meaning of the Spanish words. When they were asked to remember the
sentences constructed, and allowed to utter their created sentences, they showed
learning gains of 180%. The authors noted that these second graders benefited
from self-generated sentences. In the pilot research the subjects did not benefit
from experimenter-provided sentences in the pilot research.

One of the largest learning gains ever reported was a case of facilitation
ranging from 200% to 1270% over no-keyword subjects as children tried to acquire
foreign verbs (Miller, Levin, 8 Pressley, 1980).

Spontaneous transfer of the keyword method to other tasks was observed
among 18-year—old subjects but not among 12-year-old students (Pressley 8 Dennis-
Rounds, 1980). However, if 12 year olds were told to apply the mnemonic aid they
used in a study task to another study task, they could do so.

The youngest subjects studied were children three to six years of age who
were required to learn ten new words. The three year olds with the keyword method

scored a mean of 5.5, and the control group 0.6. In this experiment, "the lowest

21
scoring keyword subject outperformed the highest scoring control subject" (Pressley.
Samuel, Harshy, Bishop, 8 Dickenson, 1981).
On the basis of past experiments, Pressley, Levin, and McCormick (1980)

suggested three ways that were successful in helping young children learning

vocabulary:
1. The experimenter provides interactive pictures.
2. The experimenter provides separate pictures of the keyword and the

foreign words and lets the student link (interact) them.
3. The experimenter allows the students to use a sentence variation

(student generated) of the keyword.

WWW
When

Many experiments have compared the keyword method and the contextual
approaches for their effectiveness as vocabulary-remembering strategies. However,
these experimental results have been reported mainly by keyword-method
advocates, and the results have tended to report advantages for the keyword
method.

Levin etal. (1982) reported 50% more retention of definitions by fourth-grade
subjects with the keyword method as compared to subjects with the context
approach. In one experiment (Levin et al., 1984), the contextual support for the word

”angler" (a person fond of fishing) was the following passage: "The angler carried

22

a lot of things down to the stream. He carried a net and a tackle box, as well as his
fishing pole. He hoped to catch a lot of fish that afternoon."

In this instance, the keyword method had an illustration in which two
characters had a dialogue relating the keyword to the definition of the vocabulary
word. The major figure of the two was an angel who was fishing, in order to provide
an acoustic cue for the word “angler."

With college students, tasks of sentence judgment concerning correct usage
of the acquired vocabulary items, and sentence creations using them, are tested.
The results of both tasks showed the superiority of the keyword method (Pressley,
Levin, 8 Delaney, 1982).

Multiple contextual exposures such as providing synonyms, sentence
elaborations, illustrations, and reading of sentences were provided in an experiment
to see their effects on vocabulary remembering. This treatment proved inferior to the
keyword method: a 62% retention by the keyword subjects and 28% by the context
subjects (Pressley, Levin, Kuiper, Bryant, 8 Michener, 1982).

McDaniel and Pressley (1984) reported that even sentence constructions of
both groups of subjects at a later date reflected the initial superiority of the
vocabulary acquisition by the keyword method. They reasoned that contexts
facilitate meaning discovery, but the discovery of the meaning of a word by itself
does not ensure the subsequent retention of the vocabulary item. In other words,
inferring the meaning of a word is different from remembering the sound of a word.

The following experiment seems to Show some limitations of the keyword method.

23

McDaniel and Tillman (1985) reported that the keyword subjects’ superiority
disappeared if students were simply asked to construct synonyms from elaborated
definitions. The authors concluded that the keyword method’s superiority, relative
to the context method, lies in subjects’ acquisition of meaning links associating
vocabulary items and their definitions.

Reviewing the issues of contextual approaches to vocabulary learning versus
the keyword method, therefore, Pressley, Levin, and McDaniel (1987) emphasized
the importance of the distinction between getting the meaning from context, on the
one hand, and remembering the word form and meaning, on the other. The keyword
method enhances the ability to go from meaning to form, or from form to meaning.
They thus asserted that ”effective vocabulary-inferring processes and effective
vocabulary remembering strategies are complementary vocabulary acquisition

components, with one’s strength being the other’s weakness, and vice versa.”

WNW

Some empirical findings have Shown that an item of information is stored
better in memory as cognitive activities relating to the item increase, and that
ambient data encoded at the time of acquisition of an item of information in memory
are strong cues for retrieval. The former is known as the levels (or depth) of
processing theory (Craik 8 Lockhart, 1972), and the latter, the encoding specificity
principle (Tulving 8 Thompson, 1973). These findings led researchers to investigate
possibilities of introducing some extra materials into learning processes in the form

of elaboration for the sake of better acquisition and retrieval of information.

24

Consequently, some keyword experts have taken an interest in the effect of
elaboration. Some of these are reviewed in the following paragraphs.

Shriberg, Levin, McCormick, and Pressley (1982) reported that
transformational imagery instruction is much superior to representational imagery
instruction. Representational imagery simply represents facts as they are, whereas
transformational imagery is so constructed as to enhance the memory of the given
facts. For example, the Great Lakes may be represented with the five lakes of their
exact shapes with reduction in sizes with their respective names by representational
imagery. On the other hand, transformational imagery may have a lake with some
villas around it, or on it with the legend: "My homes on the lake.” The word
”HOMES” functions as the first-letter mnemonics for H-Huron, O-Ontario, M—
Michigan, and so on. Or the name of Lincoln’s assassin will be remembered by a
picture of Lincoln shot by a man standing in a booth, instead of a mere theatrical
background as a representational picture might have it.

The success of elaboration techniques in enhancing memory has led some
scholars to designate the focus of early research on memory strategies as rehearsal
and organization, and the new trend as associative elaborations or elaborative
associations (Pressley, Borkowski, 8 O’Sullivan, 1984).

Facilitation of learning by means of extratextual information is studied in the
forms of precise elaboration versus imprecise elaboration, and the elaborative

interrogation to acquire confusing facts (Pressley, McDaniel, Turnure, Wood, 8

25

Ahmad, 1987; Pressley, Symons, McDaniel, Snyder, 8 Turnure, 1988). For
example, consider the following three sentences.

The hungry man got into the car.

The strong man helped the woman.

The brave man ran into the house.
These sentences can be confusing as they are, if the task is to remember the
relations of the three agents and their actions.

These sentences can be expanded with precise elaboration to make the men
and the actions more memorable as follows:

The hungry man got into the car to go to the restaurant.

The strong man helped the woman carry the heavy packages.

The brave man ran into the house to save the baby from the fire.
The relations that emerge from the hungry man going to the restaurant, the strong
man carrying heavy packages, and the brave man saving the baby from the fire
make it easy to find or remember to whom the car, the woman, and the house
belong.

The sentences can be elaborated imprecisely as follows:

The hungry man got into the car to go for a ride.

The strong man helped the woman find the newspaper stand.

The brave man ran into the house to find some ketchup.
Examples of precise elaboration in the first group are more effective for the purpose

of remembering these confusing relationships, as they can answer elaborative

26
interrogation, which can be called a ”why" question. The latter group does not
facilitate the recall of needed information as well because the sentences are

elaborated rather imprecisely.

Metamemmialfitudx

The natural ability of the learner may often influence the efficiency of study,
or study ability in a natural situation. One way to increase the efficiency of study is
recognizing the limitations of one’s memory and successfully coping with them. This
knowledge is an understanding of how one’s memory works--metamemory. Some
metamemorial studies by keyword-method theorists are reviewed in the following
paragraphs.

Pressley, Borkowski, and O’Sullivan (1984) pointed out that "good and poor
learners differ in the extent to which they are aware of their own memory limitations."
Therefore, they asserted that the keyword method can be a way to make students
aware of 'metamemory acquisition procedures."

Studies by Pressley, Ross, Levin, and Ghatala (1984) of 10 to 13 year olds
investigated the possibility of metamemorial effects on the subjects’ cognition of
Ieaming strategies. The findings confirmed an increase of the keyword method over
the contextual method by means of both (a) giving performance feedback and (b)
asking them to think back to decide which strategy made a task easier.

An investigation of memory-strategy monitoring in adults and children by
Pressley, Levin, and Ghatala (1984) revealed some differences between the two

groups of subjects in their ways of monitoring strategies. A mere description of the

27
keyword method as well as the repetition method as a possible way of remembering
vocabulary did not predispose adults to increase their use of either method,
however, when given a description of the keyword method and the repetition
method. Children (11 to 13 year olds) tended to try the keyword method. What
changed the strategies of adults was (a) an experimenter’s recommendation and/or
(b) results from practice and test performance. As for children, they accepted
inappropriate recommendations even after practice. The authors cited Bjorklund and
Zeman ( 1982) to explain the finding: ”Children are often insensitive to their own use
of strategies, and the effects of those strategies.”
. .

WW I . S I . |

Shriberg, Levin, McCormick, and Pressley (1982) conducted one of the
earliest studies on the application of the keyword method for prose learning. Eighth-
grade students were asked to learn the purported accomplishments of some
individuals from several short passages. The results showed that gains over the
control subjects were 100% by the self-generated keyword-image group and 200%
by the imposed-imagery group.

The possibility of teaching content materials other than language was
explored by Konopak and Williams (1988). They introduced their own examples of
a lesson on minerals fora fifth-grade science class, in which they provided a general
guideline to implement the keyword method in the classroom. The authors

considered the strength ofthe keyword method as association between known and

28
new information. However, at the time of their research, there were no other studies
in which the keyword method or mnemonics in general had been used as the basis
for teaching a whole content area, such as the history of World War I. Therefore,
Scruggs and Mastropieri (1989) developed a 31-page booklet on World War II to
teach 30 mildly handicapped adolescents.

In this lesson, for example, to show that the Central Powers were Turkey,
Austria, Hungary, and Germany, a picture was Shown of people playing 'tag” (the
first-letter mnemonics for Turkey, Austria-Hungary, and Germany) in Central Park
(the keyword for the Central Powers). The result showed that the reconstructive
elaboration group outperformed the control students by a factor of nearly 2 to 1 on
immediate recall.

Investigations of the possibility of teaching content materials can be seen in
many of the following cases: an experiment in which Desrochers, Gelinas, and
Wiegand (1989) showed that the grammatical genders of German nouns were
taught with the keyword method; teaching complex botany concepts to college
students (Rosenheck, Levin, 8 Levin, 1989); teaching 45 Old English words to
college students to compare the keyword method with the contextual approach to
vocabulary acquisition (McDaniel 8 Pressley, 1989); teaching tenth-grade students
some biographical information about the United States presidents (Dretzke 8 Levin,
1990).

Rosenheck et al. (1989) taught botany concepts to college students and

concluded from their experiment that "mnemonic superiority found after a two-month

29

delay verified the staying power of the mnemonic instructions, and mnemonic
superiority appears to increase with time."

These experiments seem to Show that the keyword method can be applied
to many kinds of learning tasks that demand memory as an important part of
learning. It can be surmised that the keyword method and many of the variant forms
of this method will be explored in teaching many more content materials in the

future.

ll . I” I; III” I' IESCI

Among scholars in the field of second-language teaching, mnemonics is
largely a forgotten topic. To illustrate, Blair (1982) could name only Andrew Cohen,
Michael Tatton, and Delbert H. Groberg as experimentalists serious about
mnemonics. Munsell et al. (1988), reviewing the field as a whole from the
perspective of the study of the brain and language learning, pointed out that ”the
field of language teaching seems relatively silent on the issue of memory, with the
exception of Stevick (1976, 1986) and a few others" (p. 270).

Stevick, to whom Munsell et al. referred, was undoubtedly one of the greatest
pioneer researchers of memory in the field of second-language teaching, as
evidenced by his two publications, Memory, Meaning and Method (1976) and
Images and Options in the Language Classroom (1986). These two books do have
much information to inspire those interested in memory in the field of second

language. However, even though the basis of the keyword method is using images

30
as an aid for memory of words, Stevick, in neither of the aforementioned books,
discussed the keyword method.

Apparently the first discussion of the keyword method by Stevick appeared
in his Success With Foreign Languages (1989), as one of his interviewees, Fred,
reported that he had finally managed to remember one of the Japanese words that
gave him “the hardest time remembering.“ However, his concluding remarks
concerning the keyword method only provided his own very guarded support:

I do think, though, that it is important to remember that at best, stockpiling

through mnemonics builds resources that have a very short shelf-life; as

Dexter told us, if the words are not used in context very soon, they will

evaporate and the time and energy that went into the mnemonics will have

been wasted.

Aside from the study of the keyword method by Fuentes (1976), one of the
earliest studies that referred to the keyword method as a vocabulary learning aid and
reported the data of empirical research concerning the role of mnemonics for the
second-language learner came from Andrew Cohen and Edna Aphek (1980). In
their experiment, 26 native-English-speaking students learning Hebrew at a
theological seminary in Jerusalem were given a short training in the usage of
mnemonic associations and were given a list of associational types they could use
to learn 20 new words the subjects selected from given passages. Then they were
asked to generate their own associations, writing down each time they used an
association for a word across classes over a period of a month.

Cohen and Aphek’s report showed five types of associations students

generated and included the authors’ analyses of those types. Original words

31
students picked to learn new words and the types of associations used for them
were also reported. Another finding revealed that the beginning students used
sound as a dominant associational basis, and the advanced students also used
structural associations as an aid for memory.
Students reported what kinds of associations they used for each word. There
were four groups of recall strategies, according to the kinds of associations they

used at the time of the recall, as shown below:

mm mm
1. Used original association 46% 86%
2. Used a new association 13% 78%
3. No longer used an association 31% 61%
4. Used no association at all 10% 72%

Some observations made from the study are as follows:

1. ”It might be that almost any learner, no matter what his capabilities,
could improve his vocabulary learning through the aid of associations."

2. Generating associations to learn a word “may enhance the learner’s
attitudes toward the language itself.“ The processes may stimulate the student’s
creativity and tap ”an innate capacity for enjoying the new language.”

3. Some of the subjects expressed a desire to receive more thorough

mnemonic instructions.

32

Blair (1982), in his 'Easification," mentioned that ”in the field of language
teaching surprisingly little experimentation has been done on mnemonic training."
He mentioned Michael Tatton and Delbert H. Groberg as the few experimentalists
he knew of, and referred to Andrew Cohen as 'coiner of the term ’easification’ in
language learning,” and as "the contact person" among those interested in
mnemonics in the field. His article explained in detail how textual memorization of
Russian expressions could be facilitated through mnemonic aids. At the end of the
article, he answered a question about the problem of retrieving words acquired
through mnemonics. As the theory answers an important question with regard to the
keyword method, the whole passage is quoted below:

What then of the objection typically raised that association mnemonics force

the language learner to take extra, unnatural steps and in the end only hobble

his learning? My answer is that none of the learning researchers I know or
have read believe this. M. David Merrill, a noted author in the field of
instructional science, explains that the brain seems to use what is called the

”push-down” principle where once something is learned, regardless of how

it was learned, memory access is ”pushed down“ to the automatic level.

Ultimately there is no difference in access time between one thing learned

through association mnemonics and a similar thing learned through other

strategies.

Cohen (1987) expressed some of the critical issues with regard to the use of
mnemonic associations in vocabulary learning, giving particular attention to the
keyword method and offering possible solutions. In one section, he observed that
mnemonic experiments “have not tended to discuss the nature of associations that

the learners actually generate." As one exception, Cohen introduced his new

experiment conducted over a 100-day period. The students’ reports gave rise to

33

nine types of associations, and were given in tabular form. These nine associations
are more fully developed and explained in Cohen (1990, pp. 26-27).

Cohen referred to differences between laboratory studies and classroom
learning by citing three reasons given by Peterson and Swing (1983):

1. Many questions raised in this paper can only be resolved in the natural
environment of a classroom.

2. Answers to questions that hold true in the laboratory are not
necessarily applicable to the classroom, where there is loss of control of significant
variables.

3. Classroom learning is qualitatively different from laboratory learning.

Cohen’s conclusion that "there is not as yet enough information about the
results of such strategies to recommend that all learners use them" sounds more
conservative than Carter’s (1987) that "research in this domain points quite
conclusively to the value of what has been termed the Ire y-word technique” (p. 153).
However, Carter did mention that the technique should be considered "as only one
among several possible techniques" (p. 154). His view of the keyword method was
that an effective usage of context to learn words is not a good possibility among
learners in the initial stages of ESL learning, as words may "be assimilated as single
(or paired) items." Therefore, the mnemonic aid has some utility. However, the
method could be time consuming in cases where finding the right keyword and

picture interaction is not easy. There may be some interference with pronunciation

34

and spelling, but mnemonics, according to Carter, can sometimes "assist with
irregular spellings such as ’island = is land’ etc."

Nattinger (1988), speaking of memory of words, stated that "form may be
more important than meaning in remembering a vocabulary item. . . . We need to
teach strategies that take form as the principal path of meaning." As for the keyword
method, he believed that "the technique is valuable for students at both advanced
and beginning levels of ability."

Nation (1990) thought that "mnemonic techniques make learners independent
of the teacher." He also mentioned that one difficulty with the keyword method is
that it is often difficult to think of keywords that sound like the foreign word.
However, even with this difficulty and time involved, students may benefit by
knowing mnemonic strategies, as Cohen (1990) observed:

In conclusion, then, remember that the mind may close off fruitful options to

learning vocabulary based on justifications such as, "l have so many words

to learn that I don’t have time for gimmicks." The reality is that some of these
activities—like learning words through association—can save an enormous
amount of time.

Furthermore, Channell (1988) included this suggestion in her concluding
section of "Psycholinguistic Considerations in the Study of L2 Vocabulary
Acquisition," referring to the works of Cohen and Aphek (1980) and Cohen and
Hosenfield (1981): "The prediction that associations which are both semantic and
phonological in nature are most helpful In aiding recall of vocabulary invites empirical

testing." Apparently, the best known and researched approaches used in learning

a vocabulary in second language is the keyword method.

35
WED

Peterson and Swing (cited in Cohen, 1987) pointed out some of the
differences that should be noted between laboratory studies and classroom learning.
These differences may be illustrated by some of the comments made by the
keyword-method researchers in the past. For example, Levin et al. (1982) cited their
personal communication with Borton with regard to the encouraging results of a
national survey of teacher reactions to keyword-method application to classroom
learning:

The results of a recent national survey of teacher reactions to the present

keyword method adaptation were most encouraging (Borton, Note 4). About

three-quarters of the upper elementary grade teachers who responded were
enthusiastic about the method to the point of requesting an opportunity to
explore it further in their classrooms.

Pressley et al. (1990) stated that "no systematic research has been produced
on how to prepare teachers to teach strategies." One of the reasons for the above
situation was given by Pressley et al. themselves: "Perhaps most regrettably, many
experimental psychologists and researchers testing the potency of strategies in
laboratory settings have little interest in educational implementation of strategy
instruction."

Another marked difference between classroom and laboratory situations is
the presumed precision associated with the lab. To illustrate, researchers argue
about differences one second can make in an experiment:

In his reply to Hall and Fuson (1986), Pressley (1987) reported effective use

of the keyword method with a presentation rate of 3 s per item, whereas Hall,
Owens, and Wilson (1986) reported that their college subjects experienced

36

considerable difficulty in the use of the keyword method with a presentation
rate of 4 s per item. (Hall 8 Fuson, 1988, p. 251)

Such scientific pursuit of precision in the laboratory is lopsided from the
standpoint of classroom teachers. This point becomes even clearer when one
considers a comment by Cohen (1990) on the length of laboratory experiments in
the past. He stated that "the longitudinal studies conducted by cognitive
psychologists may include a time span of only a week or so" (p. 142). Therefore, he
mentioned that "there is room for research conducted for longer periods of time" (p.
142). Therefore, one of the main themes of the experiment was to measure
longitudinal effects of the keyword method by testing at intervals of one and two
months actual students learning English.

It is also interesting to note with regard to the keyword method here that
Raugh and Atkinson (1975) said that "it would be interesting to evaluate the keyword
method on a language, such as Russian or Japanese, that has few cognates." They
did conduct experiments with a vocabulary in Russian, and the results are available.
However, to this researcher’s knowledge, no report on an experiment using the
keyword method in Japanese or with Japanese junior high school students has yet
been conducted. It would be interesting to see whether the positive effects of the
keyword method would obtain among Japanese learners of English through an
empirical investigation. With the aspects of the keyword method discussed above,

then, this study undertook to investigate the following three hypotheses:

37
Dilemma

The following hypotheses concerning the effects of the keyword method were
tested in this study:

Hypmnesislz The keyword method used with Japanese students studying
English in the seventh grade in Japan will lead to higher scores than the
traditional method on the immediate posttest and the follow-up tests designed
for this study.

flypmngsm: The keyword method will lead to higher scores than the
traditional method on the same tests used as a follow-up test one or two
months later. The subjects of the study are those students with perfect
scores on either half of the immediate posttest.

Hypothesis}: The keyword method will lead to higher scores than the
traditional method on the subset of words on the test designated as difficult
words.

Chapter III presents the methods used to investigate these hypotheses.

CHAPTER III

METHODOLOGY

To investigate the effect ofthe keyword method in vombulary acquisition, and
to test the hypotheses developed for this study and listed at the end of Chapter II,

an experiment was designed with the following conditions and procedures.

Ihefiubjegm
The subjects were seventh graders in 12 classes, each with about 30 to 40
students, in two kinds of junior high schools in Okinawa Prefecture, Japan: four

classes at a national junior high school and eight classes at public schools.

8 E II El . E S I . I

The choice of seventh-grade students as subjects was based on an
observation made about the effectiveness of the keyword method: "It may well be
that supplying the keyword is most helpful to the beginner and becomes less useful
as the subject gains familiarity with the language and the method" (Atkinson, 1975,
p. 824). Pressley, Levin, and Delaney (1982, p. 79) reiterated the point and further
identified such less sophisticated learners as elementary school children and junior

high school students.

38

39

Therefore, the seventh-grade students fit the description of the ideal
population who may benefit most from the keyword method: seventh graders in

Japan have just begun formal education in English.

Ibe_QlaSSLong

The experiment was conducted in ordinary Japanese junior high school
classrooms in the form of the daily English lesson. A class in a junior high school
in Japan usually consists of 30 to 40 students and a teacher. In this case, the
teacher who supervised the whole class hour of the experiment was always an

English teacher.

Wet
'IIE 'l

The 12 teachers who participated in the experiment were individually given
instructions concerning the experiment; they were expected to take part in only a few
activities except for supervision of the class, to distribute the material to the students,
and to time the length of each section of the experiment according to the
experimental schedule. After the explanation of the keyword method given in the
handout discussed below, the teachers took the role more or less of proctors.

The experiment was designed in this way to reduce the teachers’ influence
on the students’ performance. In other words, if a teacher were involved in a vital
way in the explanation of the keyword method, the ability of the individual teacher

would become an uncontrollable variable. The handout given to each student was

40

intended as a uniform explanation and guide for the use of the time and materials

in the experiment.

W

The experiment was conducted after the second week in May, the second
month of the 1991 academic year in Japan. There were four kinds of tests that
provided the data for the study. They were as follows:

1. A pretest of the 432 possible subjects to test the ability to supply a
Japanese translation for the 30 words in the study. Of these students, 32 were
considered to know too many words and were not included in the tabulations. Ten
others were excluded because of irregular participation.

2. A posttest of the 390 students to determine the rate of immediate
recall of the 30 words immediately after the learning session. In the posttest and
follow-up tests, students selected, rather than supplied a Japanese translation, from
a list of 30 Japanese words (translations).

3. A follow-up test 1 of 187 students to determine their retention of the
acquired words after one month.

4. A follow-up test 2 of 203 students to determine the retention of the
acquired words after two months.

Two kinds of follow-up tests were given. Follow-up test 1 was given after one
month, and test 2 was given after two months. One-half (6) of the total (12) classes

of students took test 1, and the remainder took test 2. In other words, six of the 12

41
classes (187 students) took follow-up test 1, and the remaining six classes (203

students) took follow-up test 2.

I D II' [II E . I

One class hour (50 minutes) was used to introduce 15 English words by the
keyword method, as well as 15 words by the rote-memory method, a traditional
method in Japanese schools. The pretest began the hour, and the posttest ended
the hour. One and two months later, the test was repeated.

. .
WW! I . II E . I

The Ministry of Education in Japan designates 490 English words as essential
vocabulary items to be mastered over the course of three years in junior high school.
A list of these words can be found in the course guidelines issued by the
aforementioned agency in Japan (The Ministry of Education, 1977, pp. 102-107).
Most of the words that were used as vocabulary items in the present experiment
were derived from this source. However, not all of these 490 English words could
be used. Two steps of elimination of a certain group of words preceded actual
selection of the words that were used for the experiment. One set to be eliminated
consists of English loan words, such as "table" or "radio," which happen to appear
on the list of essential English words for junior high school students. Because the
students’ knowledge of these words might have influenced the experimental data,
it was thought best to delete them. The other set of words that was eliminated

included such words as articles, pronouns, prepositions, and numbers. As the

42

experiment was expected to last until almost the end of the first semester, the
students might have been exposed to most of these words by the end of the
experiment, in addition to the exposure during the experimental treatment.

After these two steps of elimination, the remaining words consisted primarily
of nouns, verbs, adjectives, and adverbs. From these words, 30 words that could
effectively be matched with a keyword needed to be chosen for the experiment. The
number of words to be used in the experiment was decided in consultation with
English teachers atjunior high schools in Okinawa, as well as with consideration for
the number of words reported in previous keyword experiments: 40 words a day for
college students (Atkinson, 1975, p. 823) and 18 Spanish words for sixth-grade
subjects (Pressley 8 Levin, 1978, p. 359).

In the actual selection of the 30 words, two considerations were made to
prevent a biased collection of words. One criterion was that about half the list should
be verbs. The other was that only one word should be chosen for each letter of the
alphabet. This manner of selection would produce 26 words, one for each letter.
However, this was not done for two reasons. First, there were no words on the list
beginning with the letters "X” or "Z." Second, the assistants helping to design
keyword prompts could not come up with a good keyword for words beginning with
the letter "I." Therefore, only 23 words were selected from the list of 490 words in
the guidelines. The remaining seven words came from two sources. One was

based on an observation Atkinson (1975) made about the usefulness of the

43

keyword, and the other was a new list of essential words issued by the Ministry of
Education.

The first selection of the remaining words was based on an experiment
Atkinson (1975) called a "free-choice procedure" (p. 824). In this study, subjects
received a keyword only if they requested one. It was found through the experiment
that the likelihood of requesting a keyword would increase if a subject missed the
item on the previous trial. This finding seemed to indicate that subjects were likely
to use a keyword if they had some difficulty with a vocabulary item. Therefore, to
investigate the role of the keyword in helping students with difficult vocabulary items,
five closely related and presumed difficult words were added to the list: hen-rooster
and bull-cow—ox. These words may be confusing and difficult for Japanese students
because the Japanese words that denote these two animals are unmarked with
regard to the gender of the animals; thus, hens are called female chickens, and
roosters are called male chickens. So it is also with cattle in Japanese.

The remaining two words came from a new list of essential words that the
Ministry of Education in Japan designated for the preparation of new textbooks. The
new list included 30 new words; 13 words from the original list were deleted from this
new list. This mandate took effect in April 1993. However, in the transitional period,
words from both lists could be taught to students using some textbooks based on the
new guidelines. For this reason, the new list was considered a reasonable source

for the two words.

44

For the word beginning with the letter "I," the new list had only three words:
"ideas," "important," and "introduce." The word "idea" is used as a borrowed word
in Japan, and the word ”introduce” may be well known to students as American
English teachers introduce themselves in class at the beginning of the academic
year. Therefore. the two aforementioned words were deleted, leaving "important"
as the selected word. As for the last word to be selected, a verb was sought
because verbs represented less than half the list. Among several verbs found in the
new list, enjoyseemed to represent interesting cultural and syntactic differences in
English and Japanese.

The following are the 30 words chosen for this experiment, in alphabetical
order, divided into two groups of 15 words each:

ask, both, cry, daughter, early, enjoy, forget, grow, holiday, important, July,

know, lake, hen", rooster"

March, nose, other, put, quickly, ready, send, try, use, visit, want, yard, bull",

cow", ox"

* = "hard words"; not in alphabetical order.

All 30 words were presented to all of the subjects. However, each subject
studied half of the words in the keyword method and the other half in the rote
(traditional) method. Therefore, subjects who studied the first 15 words by the
keyword method also studied the last 15 items by the traditional method; subjects

who studied the last 15 words by the keyword method studied the first 15 words by

45

the traditional method. This means that, instead of using two groups of subjects, the
control and the keyword groups, the 30 vocabulary items could serve as either the
control or the keyword words. This division was chosen so as to avoid forcing all of
the subjects at one school to memorize 30 words by rote, depriving them of an
educational innovation, or a possible benefit from the study. Similarly, the design
avoided having all of the subjects at another school use an unfamiliar experimental
approach.

QrdeLQLEresentatlenmneTwc
(impediments

In presenting these two groups of 15 words to students, the order was
crossed so that if one class in a school was presented the 15 words by the keyword
method first, the other class in the school received the same 15 words by the
keyword method second. The mixed order of presentation was chosen so as to
balance any effects of novelty or fatigue derived from a uniform order of

presentation.

SI I. [ll . [E I'lllll

The guidelines issued by the Ministry of Education in Japan provide a list of
essential English words for junior high school students. However, this list is
accompanied by no information as to which meanings of an English word Should be
taught to students. Therefore, textbook authors use their discretion in the choice of
meanings of English words that are introduced. Consequently, meanings of an

English word introduced to students could vary from textbook to textbook. (There

46

are six different textbooks approved by the Ministry of Education. At the time of this
study, three textbooks were being used by different educational districts in Okinawa.
Public schools in Japan are not allowed to use any other kinds of textbooks in class
except as supplements.)

In choosing the meanings of the 30 words selected for this study, the
researcher looked through all three textbooks currently in use in Okinawa and chose
the meanings that were most often mentioned when the word first appeared. When
all of the textbooks used different Japanese words for an English word (sometimes
not necessarily a different meaning), the researcher extended the search of the
Japanese word orthe meaning to select by looking through the other three textbooks
approved by the Ministry of Education. In the few cases when it so happened that
equal numbers of the same Japanese words or meanings turned up from the
investigation of the six textbooks, the researcher looked for the meaning of the
words by referring to dictionaries (about 20) designed for junior high school students.
What most of the dictionaries listed as the first meaning (or the Japanese word for

it) was selected.

E . I.

Junior high school students (seventh graders) in Japan have just begun their
formal education in English. Therefore, they are not yet quite sure about
pronunciations of new English words. Thus, it was necessary for the 30 selected
words to be accompanied by pronunciations written in Japanese so that they could

at least have some idea as to how they were pronounced. However, no textbooks

47

give pronunciations of English words in Japanese. The students are to learn them
from the teachers’ pronunciations. For this reason, the researcher had to resort to
dictionaries designed for junior high school students, for they are the only reliable
sources that provide phonetic pronunciations of English words using Japanese
writing. However, these dictionaries do not express pronunciations (British and
American) or a word identically. In some cases, the word "enjoy" was written in five
different ways, all sounding close enough to the English pronunciations. Even
dictionaries published by a single agent differ in the Japanese descriptions of
pronunciation, probably because of the idiosyncrasies of lexicographers.

Most of the dictionaries use only katalrana (the most common "alphabetic"
Japanese system used to write the pronunciations of loan words) letters or a mixture
of katakana and hiragana letters. The printed letters could be all of equal size or
varied, bigger and thick, to show stressed sounds. What is most often seen is
katakana with the print sizes varied to show the stressed syllable. Therefore, this

form of describing pronunciation was adopted for the experiment.

Wilden

In compliance with Atkinson’s (1975) recommendation that a committee
approach should be taken to ensure good keyword production, the researcher
formed a three-member committee for the production of good keywords. The
committee members consisted of two illustrators ofthe keywords (described below)

and the investigator.

4s

The committee members were first given an introduction to the keyword
approach to vocabulary acquisition and were then given the list of 30 English words
and asked to produce as many keywords as they could. Then, less than a week
later, the committee brought all the keywords they had produced together for
discussion. A final list of keywords for the 30 words was selected. The criterion of
the choice was what sounded good and was considered memorable.

Occasionally, one or another member came up with a new keyword for a word
for which the committee had already finished the illustration. When the three
members agreed that the new keyword was an improvement, it was added, deleting
the old one from the selections. Several instances of such a refinement process

arose until the completion of the printing of the experimental materials.

Illusttalms

The two illustrators of the keyword associations on the committee were Yuriko
Toyama and Mayumi Machida. They were chosen for their excellent talent and
availability. They are students at Okinawa Kokusai University, so they were
available for discussions and meetings more often than would have been the case
with professional illustrators.

Before the illustrators began their drawings, the group met to discuss what
kinds of images or ideas should be included in the pictures. The illustrators then
made the illustrations with the results of the discussions in mind. On some

occasions, the illustrations were revised when the three committee members agreed

49

that improvements could be expected from revision. The keywords and illustrations

are all provided in Appendix A.

E I I' [II E . I
The experiment took the form of a 50-minute session that included the

following components:

Minutes)
1. Pretest 5
2. Explanation of the keyword method 5-7
3. Studying 15 words by the keyword method 10
4. Studying 15 words by rote memory 10
5. Posttest 10_
Total 40-42

The vocabulary learning materials developed for this one-hour study
consisted of six pages of handouts. The content of these handouts is explained in
detail below and is provided in Appendix A.

Eage_QDe_Qf_tbe_bandQutS. The first page of the handouts contained the

following information in simple words that junior high school students could

understand:
1. Title: Learning English Words: Using Mnemonics.
2. Explanation of the value of acoustic association in learning an English

word, and its integration into a pictorial form to render visual support to vocabulary

50
memory. Two examples of the keyword technique using specific English words
follow, using two English words, "spider" and "hill."

3. Recommendation to students that they will avail themselves of acoustic
associations and images to learn English words in the future, and that they are about
to learn 15 English words by this method in the class.

4. Explanation of the schedule for the rest of the class hour: ten minutes
for Ieaming 15 words by the keyword method, and another ten minutes for learning
15 words by rote. Then, at the end of the class, ten minutes for a test to determine
their retention.

5. An example of the matching test format, so as to alert students that

mere recognition (and not spelling) will be required.

6. Instructions that students follow the teacher, as he or she tells them
what to do.
7. The researcher’s name, Mamoru Kinjo, and the name of his university,

Okinawa Kokusai University, at the bottom.

EagemQ. The second page is a pretest, listing all 30 English words without
Japanese translations. Students are asked to supply Japanese translations of the
English words if they know any of them.

WM. These two pages contain 15 English words to be
learned by the keyword method. The items to be learned by the keyword method
are called "Group A." The instructions given in this section concern the amount of

time allotted to learn them (ten minutes) and how pronunciation is described.

 

Eng
des

Ital

wt

ar

51

English pronunciation is given by Irata/rana. The thick, slightly larger letters in the
description of pronunciation show the stressed syllable in English. English words
transcribed in Japanese kata/rana are underlined.

On the third page, there are eight frames (8 cm x 6 cm) (see Appendix A) in
which English words are presented by the keyword method; that is, inside this frame
is a pictorial representation of an acoustic association. At the bottom of the frame
are English words to be learned, katakana pronunciations of the words in brackets,
and the Japanese equivalents/translations. Under this is another line of a phrase or
a sentence in English that integrates the association. These eight frames are
numbered top to bottom, one to eight. The next page (the fourth page) follows a
similar format except that it contains seven forms of the words to make up the total
number of words (15) introduced by the keyword method.

Eageflye. On the fifth page is a list of 15 English words and their Japanese
equivalents to be studied in ten minutes in any way students choosenprobably by
rote. This group of words is called "Group B." It should be noted, however, that
equal numbers of groups used Group B words first with the keyword method and
Group A words second with the rote method.

Eaggij. This last page is a test of all 30 words the students had just
studied. The 30 English words appear in alphabetical order (with the exception of
"hen" and "rooster" at the end of the first group, and "bull," "cow," and "ox" appearing
at the end of the second group). The first 15 words occupy the top half of the page.

A column containing the 15 Japanese translations of the first 15 words, in random

52

order, appears to the right of the English column at the top. The last 15 words are
on the bottom half of the page on the left, with Japanese translations in random
order in a column to the right. This arrangement was used to save some processing
time and effort.

The answer sheets provided the data forthe analyses. They were sent to the
researcher unscored by the teachers who agreed to conduct the experiment with

their classes.

CHAPTER IV

RESULTS

Chapter IV contains the results of the analyses of the data. The first section
reintroduces the purpose of the experiment and the three hypotheses under
investigation to summarize the purpose and the conditions under which the
experiment was conducted. The chapter then presents the data, without statistical
analysis, to provide a general outline of the study. Finally, the three hypotheses are

examined statistically.

EumosecflbeStudu

The purpose of this study was to provide an empirical investigation
concerning the keyword method in a classroom setting. The subjects were 390
Japanese seventh graders studying English in Japan. The task was to remember
the pronunciations and meanings of the words presented in this experiment. To
measure important differences between the keyword method and the rote

(traditional) method, three hypotheses were investigated.

Initlxmthsses

The following were the hypotheses of this study:

53

54
flypmnesisjz The keyword method used with Japanese students studying
English in the seventh grade in Japan will lead to higher scores than the

traditional method on the immediate posttest and the follow-up tests designed
for this study.

W: The keyword method will lead to higher scores than the
traditional method on the same tests among those students with perfect
scores on either half of the immediate posttest.

Hypothesisjz The keyword method will lead to higher scores than the

traditional method on the subset of words on the test designated as difficult
words.

IbeEouLKindsnLQata
IaniEanleem

The data in the experiment were obtained from four tests. They are:

1. The pretest

2. The immediate posttest

3. The one-month follow-up test 1

4. The two-month follow-up test 2
A brief explanation of these four tests is given in the following paragraphs.

Ibeprejest. The pretest (five minutes) was taken by a total of 432 students
at the beginning of the experiment. It required students to supply Japanese
equivalents for the two sets of 15 English words.

Iniimmedlatemngsl The immediate posttest (ten minutes) was taken by
all of the subjects in the study at the end of the class hour. It required students to
find the correct match for each of the 15 English words from a list of 15 Japanese

equivalents for both sets.

55
3. W. The one-month follow-up test (15

minutes) was given to the students in 6 of the 12 classes (187 students). It was
identical to the posttest except for the time limit.

4. W. The two-month follow-up test (15
minutes) was given to the students in the remaining six classes (203 students). It

was identical to the one-month follow-up test 1.

Explanaflmcflbeflretest

The pretest was administered to ensure that the students had little prior
knowledge of the vocabulary items to be learned: 15 items to be learned by the
keyword method and another 15 by the traditional (rote) method. The study
established a score of 3 as the cutoff point on either of the two groups of 15 words.
Thirty-two students exceeded this maximum and were excluded from the data to be
reported in this study. (A few had very high scores.) An additional ten students
were excluded because of missing data. Among the 390 students in the study were
some students who answered one or two words correctly on the pretest.

The following gives an account of these numbers and how they are treated.
First of all, the total items that all the students would deal with by means of the
keyword method, or the rote method, was 5,850 items (390 students x 15 items).
From the group of words (5,850) to be learned by the keyword method, the students
already knew a total of 148 items: this is about 2.5% of the total. For the group of
words to be learned by the rote method, 161 items of the 5,850 were correctly

answered by students, about 2.8% of the total. Both groups. therefore, appeared to

56

know roughly equal numbers of the English words to start with, and no distinction

was made of words as "known" or "unknown."

G lDll' [II E . llE' I.

This section provides data of a general nature. The following tables and
graphs Show general trends of the data with a brief description. The tables contain

some abbreviations, which are explained below.

Terms
In the following tables and graphs, the following notation is used for different
types of methods and test scores:
K = keyword
R = rote
Kpost (or KP) = the posttest grades of students using the keyword method.
Kfoll (or KF) = the follow-up test (both 1 and 2) grades of students using the
keyword method.
Rfoll (or RF) = the posttest grades of students using the rote method.
Rpost (or RP) = the posttest grades of students using the rote method
Kdiff (or KD) = the difference between the posttest and the follow-up test of
students using the keyword method.
Rdiff (or RD) = the difference between the posttest and the follow-up test of

students using the rote method.

57
QatafimmJbeEmireEQDulafion

Table 1 shows the frequency distribution of the students’ scores from 0 to 15.
The numbers of students are given across the four columns representing four
variables: (a) the posttest of students using the keyword method, (b) the posttest of
students using the rote method, (c) the follow-up test of students using the keyword
method, and (d) the follow-up test of students using the rote method.

The first row of scores Shows that, at the time of the immediate posttest, 9
students scored 0 with the keyword method, and 11 students scored 0 with the rote
method. The rest of the same row indicates the numbers of those who scored 0 at
the time of the follow-up tests: 29 for the keyword method and 33 for the rote
method. The next 15 rows provide information about the numbers of students who
received the scores indicated.

W. The frequency distributions Shown in Table 1 are presented
in Figures 2 and 3 (bar graphs). Figure 2 represents the frequency distribution of the
immediate posttest, and Figure 3 that of the follow-up tests.

In Figure 2, the horizontal line shows the possible scores from 0 to 15, and
the vertical line, the frequency counts of each score. The keyword score is shown
with the black column (the black column/bar hereafter) and the rote method score,

with a white column (the white column/bar hereafter). According to the chart, in the
areas showing the scores of 0 to 6, the white columns are taller (more rote students

scored in this 0—6-point range). But from 7 to 15 points, the black columns

58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 1: Frequency distributions of students’ scores on the posttest and follow-
upmfi.
Posttest Follow-up Test

Score Number Percent Number Percent
KN RN K% R% KN RN K% R%
0 9 11 2.3 2.8 29 33 7.4 8.5
1 6 7 1.5 1.8 27 37 6.9 9.5
2 9 9 2.3 2.3 39 60 10.0 15.4
3 6 13 1.5 3.3 27 47 6.9 12.1
4 5 3 1.3 0.8 37 41 9.5 10.5
5 12 19 3.1 4.9 32 44 8.2 11.3
6 12 21 3.1 5.4 42 31 10.8 7.9
7 19 13 4.9 3.3 32 29 8.2 7.4
8 21 20 5.4 5.1 29 18 7.4 4.6
9 18 24 4.6 6.2 26 25 6.7 6.4
10 16 28 4.1 7.2 17 8 4.4 2.1
11 30 27 7.7 6.9 22 6 5.6 1.5
12 22 29 5.6 7.4 10 2 2.6 0.5
13 49 34 12.6 8.7 14 5 3.6 1.3
14 55 55 14.1 14.1 1 0.8 0.3
=15 101 77 25.9 19.7 4 3 1.0 0.8

 

Note: K = Keyword, R = rote, N = number of students

 

59

.5258 oEooEé 9: Co 8:29me Sconce”. ”N 9sz

meom

8.9. 006? co. 3 00d oofioo :ooo.m v09m 8...
86¢ 00. Nw 8. or oo. o

8
FZDOOQE D -.. - - - fie
PZDOOQV. I
.ON

 

 

 

 

.00..

 

 

our

 

enjeA

60

5:093 D
528%. i

  

oo.m_.

859 3-26:8 9.: Lo 5:39.86 55:09“. ”m 2:9“.

CON oo.m 09m
006 CON

comp oo. 2. cod
00.0w cod 00¢

8.3. OONP

meOm

oo. _.
oo.

 

en's/t

61

are taller except 9, 10, and 12 points, with 14 points being equal (55 and 55). The
keyword method seems superior at the upper ranges.

Figure 3 represents the frequency distributions of the follow-up test. The
graph shows a different pattern from the results of the immediate posttest. In this
graph, the dividing point seems to be around points 5 and 6: in the 0-5-point area,
all the white columns are taller; in the 6-15-point area, all the black columns are
higher. In other words, the rote method scores tend to be in the lower part and the
keyword-method scores in the upper part.

ONO-.I 0| 0 l‘ 'i‘ ‘l ‘ .0‘h “I I‘ 00 l‘ «.10. l‘ 0 .A".. ‘
Tables 2 and 3 and Figure 4 Show the differences between the posttest and the
follow-up test. In other words, the points shown are the amount of loss that took
place between the posttest and follow-up tests at one month or at two months. The
chart shows that, generally speaking, the greater loss took place with the rote
method and the smaller loss took place with the keyword method. Figure 3 shows
that a few students actually had higher scores (1 to 5 points) on the follow-up tests.
Three students lest 15 points, meaning they went from 15 on the posttest to 0 on the
follow-up test. The tables Show that the average loss was about 5 points (5.3 for the
keyword method and 5.9 for the rote method).

Oll0-.. 0| 0 I‘ 1‘: AI. -. o ' "o u. ‘0, .- ‘0. According
to the data given in Table 4, 101 students had perfect posttest scores with the

keyword method, and 77 students with the rote method. A total of 33 students

ill.
.I
Huh
II-

62

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 2: Differences in scores between the posttest and the follow-up test with
the keyword method.

Value Frequency Percent Valid Percent Cum. Percent
-5.00 2 .3 .3 .3
-4.00 1 .3 .3 .5
-3.00 3 .8 .8 1.3
-2.00 4 1.0 1.0 2.3
-1.00 8 2.1 2.1 4.4

.00 25 6.4 6.4 10.8
1.00 22 5.6 5.6 16.4
2.00 32 8.2 8.2 24.6
3.00 30 7.7 7.7 32.3
4.00 40 10.3 10.3 42.6
5.00 27 6.9 6.9 49.5
6.00 44 11.3 11.3 60.8
7.00 41 10.5 10.5 71.3
8.00 29 7.4 7.4 78.7
9.00 34 8.7 8.7 87.4

10.00 21 5.4 5.4 92.8
11.00 9 2.3 2.3 95.1
12.00 7 1.8 1.8 96.9
13.00 8 2.1 2.1 99.0
14.00 2 .5 .5 99.5
15.00 2 .5 .5 100.0

Total 390 100.0 100.0

 

 

 

 

 

 

 

Mean = 5.336 Mode = 6.000

63

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3: Differences in scores between the posttest and the follow-up test with
the rote method.

Value Frequency Percent Valid Percent Cum. Percent
-2.00 4 1.0 1.0 1.0
-1.00 7 1.8 1.8 2.8

.00 21 5.4 5.4 8.2
1.00 25 6.4 6.4 14.6
2.00 24 6.2 6.2 20.8
3.00 24 6.2 6.2 26.9
4.00 33 8.5 8.5 35.4
5.00 43 11.0 11.0 46.4
6.00 30 7.7 7.7 54.1
7.00 43 11.0 11.0 65.1
8.00 33 8.5 . 8.5 73.6
9.00 25 6.4 6.4 80.0

10.00 29 7.4 7.4 87.4

11.00 23 5.9 5.9 93.3

12.000 12 3.1 3.1 96.4

13.00 9 2.3 2.3 98.7

14.00 4 1.0 1.0 99.7

15.00 1 .3 .3 100.0

Total 390 100.0 100.0

 

 

 

 

 

 

 

Mean = 5.954 Mode = 5.000

60:28.6 Lo 5:39:66 Damaged ”v 9:9...

DEE

8.9 8.2 8.: 8.8 8R 8.m 8.,” 8.. 8..- 8.9 8H.
HznocuomD _ _

 

ODIBA

Encode. a

or

ON

64

on

0?

cm

65

 

 

 

 

 

 

 

 

 

 

 

000. 000. oooo.m—. Pmoam
000. w NM CO. GOO. COO. 0000. mm
000. 000. 8006—. Chou!
a
Hag
GOO. COO. 0000...: Fmoam
000. ms mvfi: QMN. mNON mm Pm. F- KN
0mm. mNON momvdw PWOQ!
gum“
rNN. mNNN NVmew Fwomm
CDC. 9: wad wNN. mNNN mmva For
GOO. GOO. 800mg Pmoav.
3mg
=<I=E
.nea 02a) Em .3m 500 .2m cams. Em $00 330
5 . . cams. . 20th>
=34. .. 8:285 Em Bm Lo 02

 

 

 

 

 

 

 

 

 

.33. £02 .5 «mozmoa EmfioEg 55 Co to; 556 cc mmeoom Seton 53> 8:256 ho LmnEac on... ”v 2an

 

66

received perfect scores with both methods. Further discussion is given in the
analysis section.

W. The data concerning the
results of the effects of the hard words are presented in Table 5. Without looking at
the detailed parts of the figures, readers can see from the table that, at the time of
the posttests, the scores deriving from the rote method are higher. On the other
hand, the follow-up test scores Show that the scores deriving from the keyword
method are all superior. Consequently, there is an apparent reversal in the effects
of the two methods.

This concludes a discussion of the general outline of the experimental

findings. Further analyses of the data are provided in the following section.

8 II [II II 'I l'
tlxuotbeslsJ
The keyword method used with Japanese students studying English in the
seventh grade in Japan will lead to higher scores than the traditional method
on the immediate posttest and the follow-up tests designed for this study.

To test the first hypothesis of the experiment, the data from the entire

population were analyzed by means of the following procedures:

1. The Wilcoxon matched-pairs signed-ranks test.
2. The t-test.
3. Comparison ofthe difference (lost points) between the posttest and the

follow-up test of the keyword method and the rote method.

4. Comparison of 12 classes using the means of each class.

67

 

 

 

 

 

 

 

to. an. ER. joumzf
o8. 8m 88 NS. 84. 88. O8. «8. 8m
80. 8m. 88. 3885:
8%
04—80% m
as. men. 8:. 50%;:
8o. 8m Ba- 80. Se. :8. 08. 8m. 08
:8. Be. 88. 508;):
Wollfidm
I81... 26
:82 mean.
:00
an E 2...? co mm am .2 cams. ow cam—2 .Qw Eco 8.0 ofimcm>
__e.~ -3 Lo mm :98 .o .
30:99.20 ocean. z

 

 

 

 

 

 

 

 

 

 

 

 

.86; Ba... 862 9.32.8 9: new 5288 9: comics mooccuoEc Co $83.50 no 2an

68

The following discussion deals with the results of the analyses through these

procedures, in the order listed above.

IDQJMIQQXQDLLQSI

The data gathered from the entire study population were first analyzed using
the Wilcoxon matched-pairs signed-ranks test (hereafter called the Wilcoxon test)
in consultation with advisers at the Statistical Consulting Service at Michigan State
University. Although this is not considered a powerful tool, it has the advantage of
being useful where the assumption of normal distribution cannot be made with
certainty. Because no other studies were available to establish the normal
distribution of this population, the Wilcoxon test seemed appropriate as one of the
tests.

The Wilcoxon test separates the scores of the students into three groups and
gives the numbers of subjects who fall into each category. The three categories are
as follows: (a) students’ scores with the keyword method that are lower than scores
involving the rote method (K LT R); (b) students’ scores with the keyword method
that are greater than those with the rote method (K GT R); and (c) students’ scores
with the keyword method that are equal to those with the rote method (K EQ R).
The analyses of the posttest scores, follow-up test 1, and follow-up test 2 are

provided in the following:

69

The Results of the Posttest:

% Cases
32% 125 - Ranks (K LT R)
48% 188 + Ranks (K GT R)
20% .12 Ties (K EQ R)
390 (2-tailed probability = .0000)

The analysis by the Wilcoxon test showed that, at the time of the posttest, the
keyword method showed a superiority: 48% of the subjects scored higher learning
the new vocabulary items with the keyword method, as compared to 32% of the
students who did better with the rote method. The net difference between the two
treatments was 16%.

Because the 2-tailed probability in the Wilcoxon test was .0000, the observed
significance was sufficient to reject the hypothesis of no difference. Therefore,
Hypothesis 1, that the keyword method leads to higher scores on the posttest,
cannot be rejected in the case of the immediate posttest. I

The Results of Follow—Up Test 1:

% Cases
28% 51 - Ranks (K LT R)
58% 110 + Ranks (K GT R)
15% .26 Ties (K EQ R)
187 (2-tailed probability = .0000)

The data from follow-up test 1 showed that, after a month, 28% of the
students who took test 1 scored higher with the rote method, and 58% of the
students did better with the keyword method. The 2-tailed probability was .0000 (p

< .05).

70

The percentage difference between the subjects who did better with the
keyword method orthe rote method was 30%, in favor ofthe keyword method. This
seems to be an indication that after a month the words learned by the keyword
method were retained better.

The data so far have shown that the scores resulting from the keyword
method were higher both at the time of the posttest, which tested initial acquisition,
and at the time of test 1, which tested the retention of the acquired vocabulary one
month afterward, with a 2-tailed probability of .0000 for both times. Furthermore, the
percentage difference between the keyword method and the rote method rose from
16% to 30%, in favor of the keyword method, on follow-up test 1. There is yet one
more part of the data to consider: follow-up test 2, which would show the results of
the scores after two months.

The Results of Follow-Up Test 2:

% Cases
25% 51 - Ranks (K LT R)
60% 122 + Ranks (K GT R)
15% _3_Q Ties (K EQ R)
203 (2-tailed probability = .0000)

According to the data for follow-up test 2, there was a slight rise in the
differences of the percentages between the two populations: It was 30% a month
earlier, but rose to 35% a month afterwards. The slight rise was again in favor of the
keyword method. Therefore, Hypothesis 1 cannot be rejected in the case of the

follow-up tests.

71

In summary, the posttest suggests that many students did better at the time
of the acquisition, as seen in the data: 48% of the students did better with the
keyword method, as compared to 32% of the students who did better with the rote
method, and the percentage of students who did better with the keyword method
kept rising throughout the tests: the posttest and follow-up tests 1 and 2—48% to
58%, and then to 60%. Conversely, there was a declining trend of the percentages
of students who scored better with the rote method: 32% to 28%, and then to 25%.

Conclusion. All three test scores of the two populations of the experiment
subjected to the Wilcoxon test yielded a 2-tailed probability at the .0000 level for all
three tests; the keyword-method groups were superior each time. Therefore, the
data suggest that the keyword method is superior as compared with the rote method,

and Hypothesis 1 is tentatively accepted.

IleLt-lest

The next statistical procedure used to test the hypothesis was the t—test. To
use a t-test, one needs to assume that the targeted population is normally
distributed. This fact is examined through the normal P-P plot of all variables in
consideration. Despite the caution expressed above regarding normal distribution
in the discussion of the Wilcoxon test, the t-test seems justified with this population,
as suggested by Figure 4. The data from the t-test are included in Appendix B.

The keyword posttest average score was 11.1103 and the rote posttest
average score was 10.400. The difference between them is .7103. The t-test

suggests that there was a significant difference between the keyword post and the

72

rote post (2-tailed probability = .000). The keyword post was significantly higher than
the rote post.

Results of the analyses of follow-up test 1. As for the 187 students with
the one—month follow-up test, the t-test indicated that the average of keyword post
was different from the average of rote post. The mean of the keyword population
was 11.2460, and that of the rote group was 10.1176 (mean difference = 1.128).
The students’ follow-up test showed that the mean of the keyword group was 5.786,
and that of the rote group was 4.524 (mean difference = 1.262). There were
differences in both the posttest and follow-up scores (for both, 2-tailed probability =
.000), and the keyword-method group showed superior scores each time.

Results of the analyses of follow-up test 2. As for the 203 students with
two-month follow-up tests, the t-test suggested that there was a nonsignificant
difference in the averages of the posttests for both methods. The mean of the
keyword-method group was 10.985, whereas 10.660 was the mean of the rote-
method group (2-tailed probability = .170; mean difference = .3251). However, the
average of the keyword-method follow-up (5.76) was significantly higher than that
of the rote-method follow-up (4.37) (2-tailed probability = .000; mean difference =

1.389).

C . [II D'[[
Between the posttest and the follow-up tests for all 390 subjects, the keyword
group lost, on average, 5.3359 and the rote group 5.9538, for a difference of .618,

with a 95% confidence interval between 1.042 and .194. The t-test showed that

73

there was a significant, if modest, difference between the amount lost by the
keyword-method population and the rote-method population (2-tailed significance =
.004).

Greater insight was gained into the effect of the keyword method when the
overall population was analyzed according to the subpopulations of the one-month
follow-up test and the two—month follow-up test.

Results of the analyses of the one-month follow-up test 1. As for the
group of students with the one-month follow-up test, the t-test showed that there was
no difference in the average number of points lost between the posttest and the one-
month follow-up test. The keyword group lost 5.4599 and the rote group lost 5.5936
(difference of 1.064) (a 95% confidence interval between .763 and .495; 2-tailed
significance = .675).

Results of the analyses of the two-month follow-up test 2. However, for
the population of students with the two-month follow-up test, the t-test indicated that
there was a significant difference in the drop between the posttest and follow-up
tests. The keyword population lost 5.2217 and the rote population lost 6.2857
(difference of 1.064). The keyword-method population was superior to the rote-
method population, with a 95% confidence interval between 1.635 and .493, and 2-
tailed significance = .000. This suggests that students with the keyword method
were much more likely to score higher when tested after two months than those

students with the rote method.

74

Conclusion. The comparison of posttest means of thetotal population of 390
students by two different methods, keyword and rote, suggests that the keyword
method has an advantage in the initial acquisition of vocabulary items. Furthermore,
the analysis of the difference in the amount of loss from the initial acquisition,
measured by means of the loss between the posttest and the follow-up tests,
suggests that the smaller amount of loss in the means of the follow-up tests of the
keyword group is indicative of the superior retention with the keyword method.

Beauflsoflheanalxseseflhmmmflsensefiflasses The whole population
was also analyzed as a unit for a comparison of means. All the scores of the total
population of 390 students were computed and are shown as 12 units of the data
(Appendix C) with their means and standard deviations based on class membership.

The data for the 12 classes are provided in the Appendix. However, what is
of interest here is a comparison of the means of the keyword method and the rote
method. Therefore, the means of all 12 classes were extracted from the data and
listed below for comparison purposes. (See Table 6.)

The first two rows of the table Show the scores of the posttest of the keyword
method and the rote method. A glance at the scores of the 12 classes Shows that
out of the 12 classes compared, there are only 3 classes (classes 2, 5, and 8) in
which the scores of the rote method on the posttest are superior to those of the
keyword method. On the follow-up tests, out of 24 sets of scores, there is no single

case in which the rote method showed superiority.

75

 

 

 

 

 

 

mmvm wmm.m mama mvmw omwd name :3.” oboe mos...e mhmé nwmé mmwN “E
3nd wwwfi 9.0K oovN omoé mvmé comm ooo.m mmm.m Fmoé mmvd ems.”~ mv.
mmo. : mmv. E. N 3.0? Km. 3 nmwfi ooo.o_. mmmdr wmo. Z. wmmd vmhd oovdr wad am
NmoNF thNr 3a.: mNo.N_. with vmndp mnwér mama: mmméw mmnd oowd EN: ax

NP 5. or m w n m n v m N _. mme

 

 

 

 

 

 

 

 

 

 

 

 

 

.8058... So. new 20239. 9: .2 70.9 3-30:2 on. new Bounce 55 :0 3890 NF 9: .2 mane—2 ”m 2an

 

76
In conclusion, using a variety of evaluation procedures, Hypothesis 1 seems

to be upheld.

Hypothesis}
This section provides the analyses of the second hypothesis of the
experiment:
The keyword method will lead to higher scores than the traditional method on
the same tests used as a follow-up test among those students with perfect
scores on either half of the immediate posttest.

This section of the study compares the scores of the following three groups

of subjects to compare the levels of their retention:

1. The 101 students who received perfect scores related to the keyword
method.

2. The 77 students who received perfect scores related to the rote
method. I

3. The 33 students who received perfect scores with both methods.

All 145 of these students received perfect scores through one or both
methods. Their initial acquisitions of words are considered to be equated. The data
from these students give the following results.

Ale. ‘OI‘qu'l‘. .Q‘l AI"“"“ 0‘ ‘q‘O
te_the_kemrd_m_ejhed. Out of the 101 students who scored perfectly with the
keyword method only, 47 students took follow-up test 1 and 54 students took follow-
up test 2. Their data and the results of the t-tests of these 101 students are included

in Appendix D.

77

The means of the three tests show that those of the keyword method are
always higher, and the 2-tailed probabilities are always smallerthan .05 (.0000, .001,
and .0000, respectively). Therefore, the hypothesis of no significant difference is
rejected: The keyword method seems to be superior.

When the immediate posttest was given, the advantage of the keyword group
averaged 2.7443; after a month, the average advantage was 1.4468, and two
months afterwards the average advantage was 2.3519. These figures seem to
suggest that the keyword method sustains its advantage over time.
an. ‘ o rt... l‘ 0‘1 AI. O‘OO‘I‘OI ill-0‘
method. Of the 77 students who on the posttest scored 100% on words learned by
means of the rote method, 30 students took follow-up test 1 and 47 took follow-up
test 2. It is interesting to note that among the 77 students with perfect score on the
rote method set at the time of the posttest, the keyword method showed an
advantage when the students were tested one month and two months afterwards.

Analyses of the data of the 77 students with the immediate posttest.
The 2-tailed probabilities of the analyses show that, at the time of the posttest, the
rote method was definitely better (.000), but test 1 showed that the keyword method
showed an advantage (.000). Again at the time of test 2, the keyword method was
better (.003).

Analyses of the data of 30 students with follow-up test 1. As for the

posttest showing the 1.9667 average advantage in favor of the rote method, the

78
difference shifted to 2.1033 points in favor of the keyword method at the time of test
1 (after one month).

Analyses of the data of 47 students with follow-up test 2. As for the 47
students who took test 2 after two months, they had an average advantage of 1.234
points over the keyword method at the time of the posttest. However, by the time
of follow—up test 2, the keyword method had an advantage of 1.5531 points above
the rote method.

An. ' o I‘M-.0 I‘ 1': AI. 0'0. 0‘I"! AI..|
methods. Ofthe 33 students who scored perfectly with both methods, 11 students
took test 1, which was the follow-up test after a month, and 22 students took test 2,
which was the follow—up test after two months. The results are presented in the
following paragraphs. (See also Appendix D.)

Analyses of the data of 11 students with follow-up test 1. These
students started with equal scores (15), but after a month, the scores of the 11
students with test 1 averaged 8.7273 forthe keyword method and 6.3636 forthe rote
method. The mean difference was 2.3637, and the 2-tailed probability was .012. As
p < .05, the hypothesis of no difference was rejected; therefore, the data suggest
that the keyword group was superior to the rote group.

Analyses of the data of 22 students with follow-up test 2. For the 22
students who took test 2 after two months, the difference of the means was 2.1819,
also in favor of the keyword method, and the 2-tailed probability was .01. The

difference was again in favor of the keyword method.

79

Conclusien. It can be said from the above observations that the data
supported the hypothesis, which predicted that even with students whose initial
acquisition was very high, the keyword method seems superior for helping in

remembering the meaning of English words. Hypothesis 2 thus seems to be upheld.

Hypothesis}
This section provides the analyses of the data regarding the third hypothesis
of the experiment:

The keyword method will lead to higher scores than the traditional method on
the subset of words on the test designated as difficult words.

Raugh and Atkinson (1975) reported an increasing trend for subjects in their
experiment to ask for the keywords in learning words that were hard to remember.
Therefore, to investigate the keyword effect on hard words in this experiment, five
hard words were included in the list. These were two groups of words: (a) hen and
rooster and (b) bull, cow, and ox. All 390 students learned one of the groups of
words with the keyword and the other with the rote method, or vice versa. The
percentages of successful mastery of these words as values and their frequencies
were calculated. As there were two groups of words, one of which consisted of two
items and the other of three items, the following adjustments in calculations were
made in computing the students’ scores. All of the students received their
percentage of correct answers as their scores. To illustrate, if all the answers were
correct (100%), 1 point was given. In the case of one correct out of two, 0.5 point

was given as the score. Then, scoring was done as follows: one out of three, 0.33

80

point; two out of three, 0.67 point; and three out of three, 1 point. (See Table 7;
further data on the subpopulation are available in Appendix E.) The data are given
below.

.11. ‘OI‘OOI‘ 0‘-I°I‘°°t-0 O'OII-0.00.One
of the important findings is that the rote method Showed a superiority in acquisition
of hard words on the posttest (rote 74% and keyword method 68%). However, the
scores of the follow-up tests Show that the rote method lost its initial superiority (rote
27% and keyword method 36% correct).

Cemeaflsenseflbediflemmebebaeenhebuememeds. The data show that
there is a 15% difference ([74-27] - [68-36] = 15) in the amount of loss between the
two methods in favor of the keyword method. In other words, hard words learned
by the rote method lost 15% of their advantage over those learned by the keyword
method when their retention was measured after a month or two months.

To summarize, the results of the analyses of the data of our samples seem
to support the hypothesis that an increased effect of the keyword method is seen in

Ieaming words that are hard to remember.

ChapteLCanluSlenS
As for the first hypothesis, the conclusions from the Wilcoxon matched-pairs
signed-ranks test, the t-test, and other observations yielded strong support to the
hypothesis that the keyword method does lead to greater acquisition and retention

of new vocabulary.

81

.mmfi. .Nvo.v HO wmm

 

 

 

 

 

 

:33. mmm cm.m mmc. mmv. ammo.
a_m meu-m HO c2~m>tu :mcz HO mm mm :mcz
mdc:cpwuu«a Ocufimc
hfic. 5mm. vfinm. JJOEEZ:
ooo. mom. can
cmc. vmm. vwmm. JJOme:
:mcz “O mm mm :mcz mam DDOO m~ficm wfinmfium>
meulm HO Ewsssz
.mfio.- .wofl.-. HO wmm
mac. amm >e.m- Mme. Hmv. fiaco.-
e.m ..mc-~ cc e:.a>-c :coz Oo mm om :moz
mdc:cuwuufia Ucufimc
mac. men. oven. BmOamZ:
oco. own. 0mm .
amo. Rev. ammo. EmOaxZI
:md: we mm mm :ccz uwm Leo; muwma cdnmfiam>
Hflculm LO acnszz

.mEoz, gsofifi Lo.— mEoE Lo 9581 K 532.

82

AS for the second hypothesis, the follow-up test scores of the three groups of
students with perfect scores on either set of 15 words showed that the keyword
method scores were always superior. Even in the case of the 77 students with
perfect scores on the rote portion of the posttest, their follow-up scores favored the
keyword method. This is a strong indication that the keyword method led to higher
retention even among the students who had a perfect score on the rote portion of the
posttest.

As for the third hypothesis, the data showed that in the initial acquisition of
hard words measured by the immediate posttest, the rote method was superior, but

the effects were reversed by the time of the follow-up tests.

CHAPTER V

DISCUSSION

IntLQdumien

This section of the study considers the data in terms of their possible causes
and classroom applications. The chapter also considers the limitations inherent in
the keyword method itself and limitations concerning the present experiment, and
consequent suggestions for future researchers. The first part of the discussion
provides possible theoretical causes of the findings by integrating facts and theories
found in the experiment, in the review of literature, or in readings not covered in the
review of literature.

The first set of interpretations is based on findings from such fields as
cognitive psychology, applied linguistics, or TESOL. The specific theories or
hypotheses are (a) the levels of processing view (Craik 8 Lockhart, 1972) and
"apperceived input as an activator of selective attention" (Gass, 1988); (b) the tip of
the tongue phenomenon (Brown 8 McNeill, 1966) and the cohort model of word
recognition (Meara, 1987); and (c) dual code theory (Paivio 8 Desrochers, 1981 ) and

the organization-redintegration hypothesis (Desrochers 8 Begg, 1987).

83

The results of the experiment strongly suggest that the keyword method does
help vocabulary acquisition as far as remembering the pronunciation and the
meaning of a word is concerned. The keyword method is different from other
methods ofteaching vocabulary because of two kinds of links, which Atkinson (1975)
referred to as the acoustic link and the imagery link. These two devices will be
discussed in the following pages in connection with a theory called the levels of
processing view, and a notion of apperceived input as an activator of selective
attention.

Leyeis_ef_pm_ceseing. Craik and Lockhart (1972) proposed a theory now
known as the levels of processing theory. According to this view, any incoming
stimulus generates some amount of mental processing on the part of the receiver.
The mental processing can be on many different levels, from shallow to deep. The
level of processing of the target information is hypothesized to influence both the
initial acquisition and the strength of retention of the information in memory.

Craik and Lockhart gave as an example of shallow information processing
what is known as maintenance rehearsal, and as an example of deep processing,
elaborative rehearsal. A familiar form of maintenance rehearsal in the case of
vocabulary learning is a simple, mechanical repetition of the words. On the other

hand, elaborative rehearsal involves thinking about the word in much more complex

85

ways during learning. The learners need to use other aspects of their prior
knowledge. For example, train could be rhymed with rain, making a phrase like "the
train in the rain." Or a train is classified as a vehicle, and the vehicle is known to be
faster than a horse.

If it is true that an item of information is acquired better with the deeper
processing of it, we have a possible explanation for the success of the keyword
method. In the keyword method, students must process an item of target vocabulary
more deeply than in mere repetition, or when merely jotting down the word as many
times as necessary until it is acquired. Specifically, the two components of the
keyword method, the acoustic link and the imagery link, appear to involve
elaborative rehearsal.

Amuslichnls. As for the acoustic link of the keyword method, students
learning an English word have to find some Japanese words that sound similar.
Therefore, it is necessary for them to pay close attention to the individual phonemes
and their order in the English word so that they can find Japanese words that
properly represent the order of the sounds in the English word. Then, students have
to examine which of the Japanese words enable them to replicate the order of the
phonemes of the English word most closely, once it is chosen as the keyword. The
closer attention the students pay to the word, the more mental investment can be
expected. Therefore, these processes of acoustic analysis of the phonemic features
of an English word could add to a deeper processing ofthe word and hence to more

effective acquisition and stronger retention.

86

Furthermore, the final selection of the keyword does not always depend on
an exact copy of the phonemes of the English word. In some cases, because of the
connotations or individual associations with the word, the second-best word or
phrase may be chosen. The choice may be the result of a negative association of
the first keyword. or a positive quality of the second best keyword that can override
the advantage of the exact ordering of the English phonemes in the Japanese word.

For example, the exact ordering of the phonemes of the English word invite
can be done with a phrase in Japanese meaning "with a prostitute." If students
recite aloud in class the English word followed by its meaning, it would sound in
Japanese like "invite (you) with a prostitute." The keyword "with a prostitute" is not,
therefore, a good one for teachers to provide in the seventh grade.

Such a process of selection of the keyword entails mental negotiation of the
acoustic advantage of one word over the meanings and associations of other
keyword candidates. This will again add to the levels of processing of information
in the acquisition of vocabulary items.

The selection of the keyword from many acoustically similar Japanese words
requires students to draw on their knowledge, and such usage of prior knowledge
is considered to be an important process of changing input into intake. Gass (1988)
referred to this process as "activators of selective attention," which leads to
"apperceived input." Apperception entails application of prior knowledge to the

intake of new information.

87

The importance of the usage of prior knowledge in language learning may be
seen in a report of a study by Pressley, Levin, and McCormick (1980). In their
experiments using the "sentence variation” technique within the keyword method,
their pilot research revealed that experimenter-provided sentences failed to benefit
a group of second-grade subjects, whereas self-generated sentences did benefit
them. When the students were allowed to utter their constructed sentences aloud,
they showed learning gains of more than 180%. The results reported here may
illustrate the involvement of different levels of prior knowledge as a cause of success
or failure in their learning.

As discussed above, the usage of prior knowledge that is required for input
to become apperceived input constitutes one of the crucial factors in successful
learning. Thus, it could be said that the keyword method and the process of finding
the acoustic link serve as ”activators of selective attention" and lead to more
apperceived input. This aspect of the usage of prior knowledge with the keyword
method should result in some learning gains. What is discussed here about the
acoustic link and activation of selective attention applies to an imagery link, as well.

imageiyjihls. When the Japanese keyword is chosen, the student now has
two words to combine in a single image—one coming from the Japanese word and
the other from the meaning of the English word. This process of creating an
effective image allows for more individual freedom and variation than mere repetition
and rote memory may offer. To illustrate, Atkinson (1975) used the English word

"pot" as the keyword to memorize the Spanish word meaning "duck." His example

88

of the keyword image was a duck having an overturned pot on its head, or a duck
carrying an overturned pot on its head. However, the image does not have to be
that way: it can be of almost any type that can allow for students' creativity in
visualizing the combination. Some examples of these possible images could be a
pot in which a duck is sitting on its eggs or its ducklings with a legend: a pot-duck
party or a pot with a cut-up duck inside: a pot-duck dead. There are as many
images as could be created by students; the process of creating various images and
selecting one that is the best for each of the students will increase the depth of the
processing of the vocabulary item.

Furthermore, the keyword method opens up for students a greater possibility
of expressing theirfeelings while learning vocabulary. For example, a happy student
may like to use or create images of an optimistic note, and an angry student may
want to vent his or her anger on images of a darker mood, or of something
destructive, which may be particularly appealing on certain days, or hours, or
because of their personal adversities. Such free accommodations of individual
psychological make-up or mood may make it more effective than methods with
rather uniform presentations. Students can possibly process deeper what appeals
to them than what they are less able to relate to.

As discussed above, the usage of the acoustic link and the imagery link
seems to result in cognitive processing that will enhance the acquisition of
information. In working on these links, students will also activate their apperception,

thus (probably) increasing the efficiency of vocabulary-learning activities. Some of

89

the positive results of this experiment may reflect the effects ofthese processes: the

levels of processing view and apperception as a motivator.

III -I fill [III 'I
Mammal

The discussion of this question is divided into two sections. The first section
deals with implications stemming from the tip of the tongue phenomenon and the
cohort model of word recognition. The second section deals with implications
derived from studies of imagery.

I‘ ' o I‘ on .‘ OI‘IOII‘IOI «.l0 I‘ can “00‘ o A. o - urn
One notable experimental finding is that all the scores of the follow-up tests showed
a superiority for the keyword method over the rote method. At the time of the
posttests, which measured immediate recall, this was not necessarily so. The rote
method produced higher results in some cases (in three of the classes, and with
hard words). That is, even when the posttest scores favored the rote method, the
superiority had vanished, without exception, in about a month. This result is
interpreted to indicate that the keyword method increased long-term retention of the
acquired vocabulary items or that words learned by rote were more easily lost.

The following discussion is an attempt to provide some theoretical
interpretations. Studies from the tip of the tongue phenomenon and a theory called
the cohort model of word recognition (Meara, 1987; see Gass, 1989 for the review

of Meara) will provide the basis for the discussion.

90

Brown and McNeill (1966) studied their subjects when a word was on the tip
of their tongueS--that is, when they were not quite able to give the word they were
sure they knew. These subjects, although unable to provide the word, could supply
such information as the first sound, the number of syllables, or the suffix of the word.
This means that a word is a composite unit integrating a plethora of information.
Horowitz, White, and Atwood (1968) discovered that not all of the components of the
word are effective as cues for the retrieval of the word when it is on the tip of the
tongue. They identified, among the syllables of the word, the first syllable as the
best cue for retrieval of the word, the last as the second, and the middle syllable as
the least effective.

If there is some identifiable facilitation in some of the acoustic properties of
a word, such as the first syllable of the word, the keyword method that uses the
sounds of the foreign word may be taking advantage of this acoustic facilitation of
word retrieval. It is most often the case that the keyword copies the first few sounds
(phonemes) of a foreign word (in this experiment, all of the keywords [Japanese]
included sounds similar to the initial sounds of the English words). The results can
possibly therefore be interpreted as this facilitation of acoustic quality at work.

In terms of facilitation of retrieval through the keyword method, a theory called
the "cohort model of word recognition" (Meara, 1980; see Gass, 1989, for the
review) may provide some insight into a possible advantage of the keyword method

over the rote method.

91

According to the cohort model of word recognition, word recognition proceeds
in a linear left-to-right fashion. Identification of the initial phoneme of a word
eliminates all the other words that do not begin with that phoneme in the mental
lexicon, thus reducing a cohort of possible words to those starting only with that
initial sound. Then the identification of the second phoneme will further reduce the
cohort to a smaller cohort sharing the two segments in the order given. The process
of phoneme identifications, and subsequent cohort reductions, continues until a
single candidate emerges.

The effect of cohort reduction by a serial identification of the phonemes is
such that, for instance, if the phoneme la! is given as a cue for the name of a
country, some readers may already have some countries in mind. With the second
phoneme Im/ given, most people can terminate the lexical search forthe country and
say "America." To give another example, given "fruit" as a cue and lael and lpl as
the first two phonemes, identifying the fruit becomes easy. There are not many fruits
that begin with this particular string of phonemes. Either lpl or M as the next
phoneme terminates the lexical search: the word sought is apple or apricot.

As discussed above, no more than the first few phonemes are necessary to
activate enough cohort reductions to zero in on the word(s) sought. Therefore, the
following discussion deals with what the cohort model of word recognition might
suggest when Japanese students are learning English words through the keyword

method.

92

The keyword method requires a student to use the pronunciation of the
English word as a retrieval cue for the meaning. Therefore, the student must first
find Japanese words sounding like the target English word. One possible process
of the recognition of English words is to process a few initial phonemes of the
English word. Using these phonemes to search for the Japanese words, only a few
Japanese words that copy the exact serial cluster of phonemes in the word remain
in the cohort. With a fourth or fifth phoneme given, fewer Japanese words will
remain. Once the Japanese word is chosen and matched to one used to learn the
English word, the image combining it with the meaning of the English word should
come to mind without much difficulty. The memory of the image can subsequently
result in a retrieval of the meaning of the English word.

Even ifthe meaning of the English word is not retrieved instantly, the student
can continue trying to retrieve the meaning of the word. Since the Japanese
keyword is available, it can be kept in working memory easily and continue to serve
as a cue. With a method such as the rote method, which does not provide learners

with heuristic devices, retrieval is often an all-or-nothing process.

Studieseflmageui

One of the important components of the keyword method is an image that
integrates the native word and the meaning of a foreign word. If images enhance
memory, as seems likely, the keyword method could have some memory-enhancing
effect because of its image or picture. The following discussion reviews a few

studies concerning the superiority of pictures or images over strictly verbal stimuli.

93
The discussion relates the dual code theory (Paivio, 1971) and the organization-
redintegration hypothesis (Desrochers 8 Begg, 1987).

In terms of memory, or memorability, pictures and words are known to be very
different. To illustrate, in one experiment, subjects inspected 2,560 pictures of
complex visual scenes for 10 seconds each and could identify 93% of them in the
recognition test. This means that the subjects correctly remembered a total of 2,380
pictures out of 2,560 (Standing at al., 1970). Verbal materials may not show such
accurate identification. A comparison of the recognition of pictures and verbal
materials is available from a study by Shepard (1967). In this experiment, subjects
were shown pictures first one by one; then they were asked to identify the one they
had seen from a pair of pictures. Their recognition error rate (misidentifications) was
1.5%. With the verbal materials using sentences to identify in a similar fashion, the
recognition error rose to 11.8%.

These experiments seem to Show that pictures (images) have different levels
of retrievability from verbal materials once they are stored in memory. Interestingly,
the different strengths of retrievability from memory between images and verbal
materials appear to be reflected in the mental representations of words. To
illustrate, Paivio (1971 ) discovered that concrete words are better remembered than
abstract words. Concrete words such as pen, desk, cat, dog, and apple involve
material images of the referents, but abstract words such as idea, freedom,

identification, and promise seem to have few of these concrete, material images.

94

Paivio (1971) accounted for this difference by postulating the dual code
theory. This theory holds that there are two kinds of information coding in memory;
one is verbal, and the other is imaginal. Concrete words are stored in these two
codes, but abstract words are stored only in a single code, verbal codes. If a word
is stored in two codes, the possibility of retrieving the word becomes higher than
when it is stored only in a single code. Therefore, concrete words that are stored in
two codes are easier to remember and are retrieved with more ease than abstract
words.

There is from the field of education some evidence showing that the concrete
quality of memorable pictures can be a potent means of helping mentally retarded
children develop basic skills of literacy. For example, the letter "f" can be taught
using an image of a flower with leaves at the cross bar and a flower at the top
(Laubach et al., 1971). lsgar (1975) used a similar technique to teach reading-
disabled children and obtained 100% success at letter identification after only 5 to
10 minutes of practice per letter.

The results of these experiments Show that pictures and images add
something to verbal memory and strengthen retention of certain information. This
longer retention becomes even more important when it is considered in connection
with 8999’s organization-redintegration hypothesis. The hypothesis is discussed
below.

According to this hypothesis, whenever an item of information is processed

in the brain, it constitutes a unique cognitive and physical event. Each of these

95

events constitutes a "trace." Traces may be organized into higher-order units such
as chunks, schemata, scripts, and so on. The hypothesis further holds that
"[bjecause each trace is assumed to be unitary, access is redintegrative, meaning
that access with an y part ofa trace makes the entire information content of the trace
available for use."

The keyword method appears to require students to remember a word in a
foreign language using both imagery links and acoustic links, at least in the initial
stages of learning the word. If the picture or image part of the keyword method is
well integrated into a trace-like unit of the word memory, it may redintegrate the
entire information as soon as the keyword image or the picture is retrieved. While
it might not be desirable to keep this image cue as a major part of the long-term
memory trace, having access to it during the period in which the word is being
consolidated in memory clearly seems to provide an advantage in lexical acquisition.
With this assumption, it is conceivable to interpret the results of the experiment as
reflecting the enhancement of the memory because of images/pictures and
redintegrative quality of the memory trace. Rote memory has fewer means of

preserving the memory trace.

5 [II D' .

The discussion in this section dealt with two questions regarding the keyword
method: (a) Why does the keyword method facilitate the immediate acquisition and
retention of foreign words better than the rote method? and (b) Why does the

keyword method facilitate the long-term retention of the acquired vocabulary items?

96

The first question was addressed through the frameworks of the levels of
processing view and apperceived input as an activator of selective attention. Both
views hold that elaboration of information, or relating new information to prior
knowledge, generally enhances intake of information. The keyword method normally
involves more elaboration than the rote method.

The answer to the second question involved such theories as the tip of the
tongue phenomenon and the cohort model of word recognition. Using the initial
phonemes of a foreign word as an internal cue for retrieving a Japanese keyword,
which then leads to the meaning of an English word, is a possible process. The dual
code theory of Paivio (1971) and the organization-redintegration hypothesis
(Desrochers 8 Begg, 1987) both point to the possibility that the retrieval of
information could be facilitated either through images, or some tightly chunked
information unit such as the unit of information that integrates an image,

pronunciation, and the meaning of the target word.

CI 9 I. I'
In terms of actual classroom applications, the keyword method lends itself to
varieties of vocabulary-learning activities, as it allows students to have some choices
in finding the keywords and creating some images combining the keyword and the
meaning of a target foreign word.
Students may be asked to find good keywords for a foreign word as part of
their homework. Then the teacher may use these keywords in some other class

period. Or several English words may be given to students, to let the students find

97

the keywords and images. The answer can be put on uniform-Sized sheets of paper
like an index card, and placed on the wall of the classroom. Students can look at
them during class breaks to study and amuse themselves and learn in the process.

The students may also be instructed to choose the keyword production that
strikes them as the best. The creator may even be nominated for some titles like the
keyword king or queen of the week, the month, or the year.

Atkinson (1975) reported that an individual can have idiosyncrasy in
preference of some peculiar keywords. Therefore, students may be encouraged to
ask their family members to provide their favorite keywords for the English words
they are studying. They can collect keywords from younger or older brothers and
sisters, or from parents. Therefore, the keyword-collection activities and the
discussions involved in them will enhance family communication and parents'
understanding of what their children are learning in their language classes. This
seems to be healthy application of a teaching method called the keyword method.

However, a word of caution may be in order here. Finding the keyword or the
combination of the keyword and the foreign word takes some imagination. As a
result, it is conceivable that some less imaginative students may never find any of
the keywords or combinations, no matter how much time they spend on the task.
Therefore, the homework Should not be forced, nor should slower students be
embarrassed regarding their efforts. Therefore, it may be done as group homework,

if group work is desirable.

98

Another area in which the keyword method seems to be of special help is
remembering some difficult words. For example, students learning English may
soon encounter groups of words that are confusing because of their homonymy,
such as (a) son and sun; (b) vane, vain, and vein; and (c) pair, pear, and pare.
Students find learning the correct spellings of these words very difficult. Then trial-
and-error processes may make students think that their efforts are wasted and the
study of these items very painful; there is no guarantee that they will not be forgotten
again. The keyword method can help to overcome or ease the difficulties involved
in these groups of words. Some examples this researcher has been using may help
make the point easier to understand. The three groups of the homonymous words
mentioned above will be used as a sample of the mnemonics:

Son and sun: The question is what goes in between the S and the N?
1. For sen-Every father thinks his son is Qk (sen-9k).

2. For sum-The sun is the life of our universe (the sun-universe).

Vain, vane, vein:
3. For vain-I am vain.
4. For lane—Jane’s house has a pretty vane.

5. For yeihul found an octopus with eight veins.

Pair, pear, pare:
6. For pair—God breathed air into His first pail. Adam and Eve.

7. For pear—Beans eat pears.

99

8. For pare—Paring knives are raLe.

As the examples Show, the sentence-variation technique within the keyword
method has the possibility of making tractable some difficult words. These particular
homonyms may seem trivial, and helping students learn some similar pairs may not
be our biggest challenge. However, the possibility of helping students feel confident
that they will not easily forget the confusing items is not trivial. This variation of the
keyword method can give students some confidence that if they try hard enough to
find some clues and elaborate their learning methods, there is a way better than
mere rote.

In conclusion, the keyword method seems to provide students with a way to
retain vocabulary items. This possibility can be explored by students, or teacher, or
researchers. However, there are some limitations with regard to the keyword

method. They are discussed in the following section.

I. 'II' [II I; lllll I III E . I
andfiuggesfienedeEutureBeseaLchets

Future researchers may wish to explore theoretically or empirically some
areas concerning the validity of the keyword method as a way of teaching foreign
language vocabulary. So far, the discussion has dealt with some advantages of the
keyword method. In the next section of this paper, therefore, some limitations or
disadvantages inherent in the keyword method and in this study will be briefly
discussed. These limitations and disadvantages raise empirical or theoretical

questions. Among these limitations are the limitations inherent in the keyword

100
method itself, the language used in this experiment (Japanese), a single exposure,

and the keyword effect on hard words.

I. 'II' II I. II I; llllll

First, the keyword method is a mnemonic device for remembering the
meaning of a foreign word. Because of this restricted purpose, the expression
"vocabulary acquisition" used by most keyword-method advocates only means the
successful remembering of the meaning of the vocabulary item or its sound. In other
variations, the keyword method could help students remember some other aspect
of information. It is not, however, designed to teach many aspects of words in their
natural, valid contexts. When words are learned in context, students can also learn
syntax, connotations, collocational possibilities, and many other discourse features
that go with the word. However, such a rich possibility is missing from the normal
keyword method, clearly one of the great limitations inherent in the method itself.
Its great advantage is helping students (perhaps especially beginning students who
do not yet have enough knowledge of English to avail themselves of the contexts of
the English material) to remember English words, and to remember difficult or
confusing words with some assurance they will not forget as easily as they might
with the rote method.

A second drawback is that it takes a great deal of time to find a proper
keyword in Japanese for some English words. Then there are some English words
whose pronunciations almost defy the finding of keywords. This means that there

are some words that can be taught with ease with keywords, and some others for

101

which it is almost impossible to find keywords. Therefore, the advantage of the

keyword method cannot extend to every word to be learned.

I I. 'II'

There are a few limitations in applying the findings of the present research to
classrooms. One obvious limitation stems from the fact that the study was
conducted with Japanese seventh graders as subjects. Therefore, the results may
not be replicated with other subjects using other languages. This limitation gives rise
to the suggestion to conduct further research of a similar kind with other English

students of different languages.

Singlefixmsum

The present study was an experiment with a single exposure of the keyword
method. Even though results and the analyses strongly suggested that the keyword
method does have a positive effect on students’ learning, these results cannot be
used to predict the effects of multiple exposures with the keyword method.
Therefore, some experiments using multiple exposures of a word using the keyword
method, or sustaining the use of the method over a long period of time, can add to

the body of knowledge concerning the keyword method.

Lengtheflhejxueriment
The present study had two follow-up tests to measure students’ retention of
the acquired vocabulary. The results showed that the relative advantages of the

keyword method over the rote method tended to get greater as time went on, even

102
though the number of words remembered continued to go down. It would be
interesting to conduct experiments to measure effects of the keyword method
covering more than three months to see whether there is further widening of the
differences. The findings could be of great interest in regard to the claims of the
keyword method with regard to retention. Even ifthe findings are negative, the study
could make a contribution by finding the time spans involved in the widening of the

differences.

WW

The results of the experiment with the hard words in the study showed that,
at the time of the first posttest, the scores of the rote method were slightly higher, but
the two follow-up tests showed that the keyword method showed an advantage in
the retention of hard words. The percentage of difference in scores between the
keyword method and the rote method was greater with hard words than with the
other words. However, the hard words used for the experiment consisted of only two
(hen and rooster) or three (bull, cow, and ox) words. An experiment could be
designed with a greater number of hard words, and with the definition of "hard" more
carefully specified, to provide better data with which to ascertain the effects of the
two methods. A valuable reference on hard words is Higa (1965).

CemearisensetthelerbaLKemord
WWW

Merry (1980) reported that the keyword with verbal images and the keyword

with pictures show no significant difference. If mere verbal images are as good as

103

actual illustrations or pictures, it could save teachers the time and expense of
preparing actual productions of the keyword illustrations, which is the hardest part
for many teachers in using the keyword method. Consequently, empirical data on
this aspect of the keyword method seem desirable.
C . [II I; III II Illl'll
W

The present study only compared the keyword method with the rote
(traditional) method, but there are many other techniques of teaching foreign
languages with which the keyword method could be compared. Some possible
candidates for the comparison may be total physical response, language experience,
communicative language teaching, or methods involving audio-visual materials. At
more advanced stages, where students are able to use the available contexts
effectively, the keyword method could be compared with the contextual approaches
to teaching vocabulary. Advantages and disadvantages or limitations ofthe keyword
method found in these comparisons would contribute to the body of empirical

knowledge available in the field of vocabulary acquisition.

APPENDICES

APPENDIX A

HANDOUTS

HM

1. Explanation of the keyword method and the class activities

fleeces-mnemour

EekliififiéxiatSREIEL'ChsE/Lh‘. teasing) ainfill?!
156:1(Iiéflatfibnthéctfifilo'clni‘ll'h‘.l't‘l’afibtfj
euanaaefi (3 3-i:)oa.8573%5(8 7-n+)0
stratauxxfi-c-r. ca) roaches“ 1:4 72 +£1.18?me
at.aeLmFommnflStceMmenrutt.rm «a: cor:
abbfiltfi§9b9t§$F0$fltflflEKohT awasnoea
cfllfi-htLS‘I’. .

8'IIIEII1N4 r] _ t I: MIME») E: ,
(bvzu4r . E»0h-.

 

 

  

\ '..’./O.

 

 

 

 

 

A.) i

.szDWiV-béatfilz H t. (5291) J I: it“? fsrwalllmn
aims-r. in? 83,41: r< SJ euoifitfi f< scz_,___~e4yj
that cabalteeersuaaxamLst.

strfiltmoifiu.aéfiuwé.loanrualaofiamab
f I: Iv (IIILL: 5) 05:] the marshal exist.

Ros—Van.lswfififimltétbfilcaéfiotiifihml
use.tolsoeeaiaaua.nmgesseannmraarca
abfilo£fl(&59x5)éiiTT$h.it.%03bm8150§
meanemrfiarseust.fiac.assuagmiafiex<xa
enebrzlLrbiT.

 

 

Etna) 311.9838 1116152589 5202-51.)“me am frat.»
heiracaalsasstlunclitbéwst.fifiL.abet
ibfififiBOLohrlOfifimleéfihif.

rztuTothoLfiflictbe,fiL$3§Ooo' n(z«uyr)
éfiiélfltfi') sex...

:i.
till €11.80YL-fl4] $3sz
:5 2. GIRL [77 -llr] In, IMF.

fi$filfibltéflffiiTmaod—vfiflhfflhflifik.

. Mamoru Ringo
Okmawa Kokusai Un versity

1 05
2. Pretest

rt

‘
~11:

L Li]

F7: 15 iii 31' {’3 _ , E? _ 1'1

*{1

Hi
I 4“
EN

15,.

 

JX>kL®‘%%1qthnw.%mfih &*B$mgxzofi gg
.t“
C.

0N)
/?T
5'- X L 1

ASK ['7' X7]
130111 [as ex]
CR1 [7 =7” 43
DAUGHTER [ i‘“--9‘J
EARLY [77—11]

1
2

3

4

5.

6. ENJOY [:2 721-"
7

8

9

1

1

“7'

“I
|_.l ”X
A _ l

FORGET [that ’7‘“ 'y
GROW [7:1 '7}

. HOLIDAY [2k "5'" 4:3

0. IMPORTANT [4 773‘: —7 '/ b:

1. JULY [7’ .3 ‘7 4:

12. KNOW [/ '7]

13. LAKE {L/ ,f 71

14. HEN [A y]

M—m -

7" 2 i~

PC)

'1. MARCH [7 -- 2‘}

2 NOSE [/ 77C]

3. OTHER" '7’ ‘fj

__ 4. PUT [7 ‘7 1‘3

5, QUICKLY 3:7 '7 .r v '7 U i
6. READY 1 L/ 7'” ’13

7. SEND [viz I/ F]

 

_.__ 8. TRY [i‘ ‘3' 4
~____9. L38 [J— —2‘}
______ 10. V181? {9‘ ” 1’ 7.” 4 ., b]
11. IMTL /i~}

_______12. YARD [JV—y]
1. BULL [ ‘7“Jb]
14. C011 {7: <7 3

0X [71‘ "171 E

y—a
(1|

106

3. A group of words to learn by the rote method

7‘ 21/ ~—- 7' A
' m gnudééhdvfidfibifik.%@6109fif.p<q N
IggigmwararTeg-eéseepeqaga1so.gumm«—y
£3330 $13, LirL. fiEQf‘é/‘b‘ipéaiflimQN—infZa’CliLIlfi'E/LO
hand-mtuili. fagbiifi<figénécéé§Lthifo

1. ASK[7X7]t‘1’i’-‘l6

2 BOTH [’3’ oz] @750)

3. GRY [754]W£.i\1(

I. DAUGHTER [ P—N 1e

5. EARLY ['7—9] $<

& BMW [Iya='54]¥bv
7. FORGET [rm ,7», H 51%
8. GROI [7x2 '7] DEE-9‘6

9. HOLIDAY [as U?" 4] (*8
10. llPORTANT [4 y :1-2 —-9 y l] fiat
11. JULY [V .3 =5 4175

12. 11110! [/ a] filo'tlné

1L um [Lr471m

1L HR[’~V]bAED

15. ROOSTER [JV-29] SIDED

107
4. A group of words to learn by the keyword method

«0150mmé.CéabetméfierfihéIOfiWTfiir
Tau.(caabeommoxau.re:»oajoxamny»¢
tflfifiabn.7yf—54V(T®Mfl)fimbnthif.flaw
+0Xu$u.fofififi<fifi$nactéiLru§t.)

 

 

   
      
     
      
  
 
     
  

  

1
ARR?“

\

\\

\
I.

  
  
 

     
  
  

1»—
39,"): ‘
1“"1'2w
«\JH
\ K11
w'IIARCHI'i’“+] 3n -

3n 7—+ (iii—5) if

 

 

_~>~
. .. . <
(5) QUICKLYU r‘} 4,911] “if

Tli‘P< 94-771) (fingg)

 

 

~ -———

121 11031: [/ or] n a?) READY ("V i" 4.1—35%;-‘2‘3 r;
1'11 I az' (col) u-r' 4 Immwgt

 

 

 

a) SEID [4: 71'11'86
{:7 (4‘) PIP 3E6

 

 

 

.. ‘4 4'“. ‘
119 11111117 “Lbs“ x.‘~"~
r5 (1E) 45 ”"3”” ”7)

 

 

(4) PUT l 7"? H EH
’3 WM

71

 

 

108

5. A group of words to learn by the keyword method

 

 
  
       
     

   
   
 

—‘ - ® HULL[7‘»]3¢<£5L)
‘6) ELSE—[zigflfrga 11:11:.) :5 12¢

 

 

 

 

.1 ‘ /"/’ ”'5
0 VISIT[O' 47+]lfil'JH'6 ‘ 739] INF (”51.)
If")? (an t gar-.1115 ® csowgtaoaw 22

 

 

w 2.» _
«Dumb? 911111.“
72 (lid) a? (111.)

 

    

I;

0 YARD l"\""|"l E
'l'— F (iii) 03$!

 

 

6. Posttest

109

815.8

f3 E

.__1

awifitfitfi%uuaaxgé$fiL.€®E%é§i®ficfi

AL‘C'FM‘.

 

7'!" 2 l» 1

11511 [7293

30111 [#:on

GRY [7 ’7’ 4]

DAUGHTER [ l~*—5’]
EARLY i-r-U]
EHJGY [aw-9'“ .4]

'FORGET [3"th l‘]

GRGI [i=2 .0}

.HOLIDAY [r31 '1 f' 4]

iIPORTAliT [4 yd! ~9 '2 1]
JULY [3" .1 a 4]

1110! i/ 93

LAKE [V493

Ill-:11 [AV]

ROOSTER {JD—2y]

-'r' z l~ 2

mmmer—f]

nose 1/ 9x]
OTHEM'rfl

PUT {7.7”
QUICKLY[7'7 4779]
READYEi/i'" 4]-
SEND[-lz‘/F]
nun-54]
USB[=——z‘]

VISIT ['7' 41" 47+]
lAll‘i['7 71]
YARD['\"-i~']
EULL17’Ivi

emu: 9]

OX {73' 771]

.\'.'*."..\".‘*.”.‘*."."‘.””'“°*‘*

§-(
75
$115
SIDED
7.?th

mo'Cl‘B

bbc‘f')
iLt‘
@750
“LR. i51<
fifi‘i’b

E

l<
u¢<»5L)
5

3H

tab?
£¢185L)
we
(tfi.fififi®)&¢
Ea

Tfle<
BLu

fifi

mama
fifififét

HO

1. Explanation of the keyword method and the class activities

ififiwfigrfimfimour
Eékuififiéliéténmerwiehb.FraabeJéflm
7oti(fiiénétgbntu6Ctém¢Th§#boFféébfiJ
euafisaefi (3'3-££)0B.857B§l(8-7—n¢)®
attaxfifix1311.calcaabuaJL45zltLtfiémi
5t.éénfifiwflmhfiitctmménTUET.Tu.amcor:

65be1tfiéobotfimfioigéfiufimomr.awestoea
CmfihtLiiu

 

splosllxnr r], 4 I. lllLL(t"|') E: .
and)". -,\|--’.

< Lanna r

       

 

 

 

 

JUDWiC336J15iC H O ($491) J 11. its}? FSPIDBR[XI*45’]J
éhhit.%n?.§$AuF<OJthfifififiéf<emzzs4iJ
thorcaabfiJtaéersflxaxénLit.

itFEJkhaigm.kéfifiwé.l®&krhaflfiwfiémfiL
H: w (HILL: 3) (Mil tho 16515219111 viii-r.

 

mos—Van.15®§$%01665bfijtfiéfiatfiiEOWs
vii.f®15®$¥éfiififlé.fififié$£t$Lfi%?$61:6
5bfiJ®Efi(&50x5)éfiiT?$ho$t.%@fibK615®§

mfiénfiarfiirLEMET.fi&L.£59fi¥%mififiéz<ii
enamstLreit.

EmmrcéabujTHiélSfiéaszmsuLL®10fi.R0
nmarsmialsabztlOfivfiirsemirofifim.ébéfi
i5§$$30KOUT10fifiQFXléfihi?o

izhuFowm;omfimav?hé.fiL$%®oo”O(2&977)
éfiiéfiimbbifik.
g1
W x I 1. BOY [#54] i). a»;
5 2. GIRL [if—IV] u. NE

fiéfif%b1téfltéiruaos—Véwurmuuaea.

Mamoru Kin'o
Okinawa Kokusai Unlversity

1 11
2. Pretest

2 1 [11”??? E Jig 52% F! El

figwfifiémorwnd.Emfifiéitafifiégimngé
251».

1+1
”if,
\{I

 

”11
P38
718

(I

-7- x 1 1
$7;
A511 ['7 27]
BOTH [73's 01]
GRY [7 =5 4]
DAUGHTER [ I~"-—7]
EARLY ['7—11]
EHIOY [Iv-5‘" 31:1
FORGET [an/7"? 1]
CEO! [7:2 a]
HOLIDAY [7.1: U?" 4)
IMPORTANT Myrsf—yw]
JULY [f' a. "7- 4]
111101 1/ a]
LAKE [V47]
HEH [ay]
ROOSTER [JV—Z7]

p—n H y.— p—o H
A w M H c:
o u o o o

.—
U1

'7' X 1* 2

O1
P14

MARCH {—7 ——-7-]

NOSE [/ 1771']
max-2M7 4f]

PUT [7.7 1‘]
QUICKLYL7 ‘7 4 '77 '1]
READY IV?” 4]
SBND[—Iz:/I~']

TRY [1‘5" 4]
USEEa—z‘]

VISIT [17" 47." 4., 1]
IANT['7 yr]
YARD[-\"—l~']

BULL [ 7"»)

001 [7‘3 17]
“17:777. 1

SDOJQODU‘AQNH

H .— H H r—o H
U! A m N H O
a o o o o n

112

3. A group of words to learn by the keyword method

7" 21/ —- 7" A

awn Selma-a C'éésbfitffiéfllillbfé‘bé 1 Oémfl‘cfii'c
T31», (c‘afibfimfxipmfismm F}: zvmfij wifilzfifi'flj—
vfigmi—gwm‘ 7y7—54 y (Tamas) ififilbd‘t'thi'd’. 7:97:
+®Kb$u‘éwfififi<figén5CtéfiLThiTo)

 

 

 

 

    
    
    

.. m V93 ”‘30
EARLY[7—'l‘]51(%)'

T—U(fl) 1$<

 

 

CD BOTH
1M 11'

 

 

 

 

 

 

(3) CRY [7 =5 4] llflfit‘ ii?
234 (llfih) 5145‘ is“ .

\ (9
f‘ I") '.

 

 

I

 

Q) DAUGHTERI l~“ "

-.-. if"; \\ -‘ " - . <4
— m -- ~ \ (fl) 61ml [7:2 0.11M???»
VA Y"; (AL

I ,
m '4" ata'a— (3:93) mmwz, ““131.

"f

 

 

 

 

 

113

4. A group of words to learn by the keyword method

 

 

 
  
  
 

 

  
 
 

\\\\\\\‘:'°"

1
E

  
    
    

   

: \N
-\

~._.... ~

‘4‘- \"_F

 

  

I
3:? 4'60:
,IV ‘1_ / I, .1“ . 1" ll..\ ‘11 (
(ED UOLTDAY n ‘ ‘ \ 0 IMF-1v Yaymy Y’l ’
Ragnar $0511) u—ru 3‘,‘ (Rik) my" f//

 

‘;/,T

  

’,///// ‘ 1’1? // /I,/
. / ,1 ,
//;I J '

tw/

0/

‘65;

AC}
)1

m IIPORTANT[‘1'/‘J’-‘;V|‘] fiflfi w I'mfl[-r\y] 3)ng

may: ijVFVL—"(fi —$Km) /.;7y!. t'!(ZEf.n> ED
" //
///1:>// /\'/// /‘/ 5 k
”I 7 , .31 ~

    

    

 

  
 

[7" 1'

Q N D,
Ly; .
I”

     

\
\

 

7 l .- - ,' w
(up /// /////’
/ / l /, "
7 // / f .r ‘ '
AID JULY + 417?! // // <15) ROOSTER Urn—29] ins/van
--‘4 (981$) 75 SIDED“ F—29—?

——rf

-M

 

 

V:

® [110! [/ a] filo'thé
L2 (fl) b§mOTU5

 

 

 

114

5. A group of words to learn by the rote method

7’ )1/ --— 7° B
&®15®$%mmcaébfiefifi503eko#beIOfimftfi
‘Pi'éébfibx'fiéfiéffiltxfitzfiiénébfiiL’Cb‘CTévo

1‘ MARCH [7 w-fl 3F!

‘ NOSE [/ 92'15

‘ OTHER['7' $11820)

\ PUT [7°v HEX

2
3
4
5‘ QUICKLY[7'.7 4-77'11tci¢<
6t READY Eu?” 4] igmfi'cgt

7‘ SEND {-ta y F1136

8‘ TRYH“? 43fzbbd’

9t USE[-‘3—-—x’]f§’)

10‘ VISIT ['7' 4X" 47 b] 375%???»
11‘ um [7 yr; mm

12‘ YARD["\”—HE

13L BULLi 7‘2Li&¢(85L)

14‘ (301172 0 m4: (boat)

15‘ 0X173‘v72 ] (fiffit fifiFfiUD) $4"—

6. Posttest

JAR-L: a: 13

115

it; ._

_.__—_

L E!

moigtfiufiaumaaxgeamLtE03§2§iomm£

A L173 1‘.
33:.

H H g—n p..-
N N H a
O O O O

D-fi
:fi

15.

'77" 1 1‘ 1

11511 [7273

BOTH {3! 91]

GRY [a 5 4]

DAUGHTER [ P—y]
EARLY [T-U]
ENJOY [Jr/9’“ :4]

'PORGET [dz 1’ a: v t]

GROY [7a a}

.ROLTDAY [3: U?“ 4]

TIRORTAUT [4 yrs! -7 y t]
JULY [9" .1 a 4]

nor [/ 93

LAKE [V473

REA [“7]

ROOSTER [xv—29]

’7'- 1 !~ 2

MARCH [*7 -i']

NOSE 1/ OX]
mu'rrf]

PUT [7.11M
QUICKLY[9'7 4770]
BEADYIi/‘i" 4]-
SERD[—t:‘/I~'3
mus“
USEEJ—z’]

YTSTT ['7' 42' 47+]
IAN‘H'? yt]
YARDHv—r]
DULLI7’Iv]

C0l[7'.7 a]
0X17r-7711

.\'.'*.>4.\":*.“7*>’.*.=‘-.‘*“°*~1

§<
75
En;
sAEO
tins

flofl‘5

35kt")
iLts‘
337503
“Ha iSI(
E5111
(*8
fife?

E

l<

[W (:05 L)
5

35

in)?

$41: (8’) L)
(the

(amt memo) fitt-
5&5

itiP<
tiLh

(fir)

man
fifim‘z‘érz

116

Explanation of the Keywords

In the following pages is an explanation of the keywords and the
phrases/sentences that were used to assist students’ remembering of the 30
words tested.

Abbreviations: K = Keyword

JP = Japanese phrases/sentences
M = Meanings of the phrases/sentences

The following is the format of explanation. First, for each word (1-30), the
keyword(s) is (are) given (K: ___).

Second, the Japanese phrases/sentences that incorporate associations

of the pronunciations of the English words and the keywords are given in Roman

characters, underwhich English equivalents appear (JP: ).
English equivalent

Third, the translations or meanings of the phrases/sentences are provided

(M: ).

 

1. ASK K: ASU (tomorrow) & KURU (come)

JP: ASU KURU KA TAZUNERU
tomorrow come if ASK

M: Ask (if she) comes tomorrow.

2. BOTH K: BOSU (boss)
JP: RYOHOU NO BOSU DA
BOTH of boss are

M: (They) are the bosses of both (groups).

. CRY

. DAUGHTER

. EARLY

. ENJOY

. FORGET

. GROW

JP:

JP:

117

KURAI (dark)

KURAI TO NAKU
darkness due to CRY
Cry due to dark/ness.

DOTA DOTA (noisily)

DOTA DOTA HASHIRU MUSUME
noisily run DAUGHTER

Daughter galloping.
ARI (ant)

ARI WA HAYAKU
ant EARLY

Ants (come) early.
ENJYO (allowance. aid)

ENJYO llNA TO TANOSHIMU
allowance good ENJOY

Enjoy good allowance.
BOKETTO (absent-mindedly)

BOKETTO SHITE WASURERU
absent-mindedly FORGET

Forget absent-mindedly
KIGUROU (anxieties)

KIGUROU DE SODATSU
anxieties by GROW

Grow by (overcoming) anxieties.

9. HOLIDAY

10.

11.

12.

13.

14.

IMPORTANT

JULY

KNOW

LAKE

HEN

K:

JP:

K:

JP:

JP:

118

HORI (dig) and DE" (OK)

KYUJITSU WA (ANA) HORI DEII
HOLIDAY hole dig OK

(As for) holiday, digging (a hole) is OK.
IPPON (one) & TANTOU (short sword)

JYUYOUNA IPPON TANTOU
IMPORTANT one short sword

One important short sword.
JYURAI (up to now)

TANABAT WA JYURAI . SHICHI GATSU
the Star Festival has been up to now in JULY

The Star Festival (July 7) has been in July up to now.
NOU (the brain)

NOU WA SHITTEIRU
the brains KNOW

The brains know.
LEI (spirit, ghost) KURU (come)

LEI KURU IZUMI
the spirit come LAKE

The lake that the spirits visit.
HEN (strange) & EHEN (hem)

HEN TO NAKU HENNA MENDORI
hem strange HEN

A strange hen hemming ”Hen!”

15. ROOSTER

K:

JP:

1 19
DOSHITA (what happened to)

ONDORI WA DOSUTA
rooster what happened to

What happened to the rooster?

. MARCH

. NOSE

. OTHER

. PUT

. QUICKLY

. READY

JP:

JP:

1 20
MACHIMASU (wait for)

SAN-GATSU MAC HIM ASU
MARCH (month) wait for

(We’ll) wait for March.

N0 (of) 8. ZU (picture)
HANA NO ZU
NOSE of picture

Pictures of noses.
AZANA (nickname)

HOKA NO AZANA
OTHER of nickname

The other nickname.
POTTO (hastily)

POTTO OKU
hastily PUT

Put hastily.
KUI (eat) & KURI (chestnut)

SUBAYAKU KUI KURI
QUICKLY eat chestnuts

Eat chestnuts quickly.
REDII (a woman)

LEDII WA JYUNBIGA DEKITA
lady ready

The lady is ready.

7. SEND

8. TRY

9. USE

10. VISIT

11. WANT

12. YARD

JP:

JP:

121

SEN (one thousand) & DORU (dollars)

SEN DORU OKURU
one thousand dollars SEND

Send one thousand dollars.
TORA (tiger) & IRU (is there)

TORA IRU KA TAMESU
the tiger is there if TRY

Try if the tiger is there.
Y UZU RU (give, share)

YUZUTTE TSUKAU
share USE

Share to use.

BIJIN (a beauty)

BIJIN TO HOUMONSURU
beauty with VISIT
Visit with a beauty.

WAN (bowl) & TO (and)

WAN TO HASHI GA HOSHII
bowl and chopsticks WANT

Want a bowl and chopsticks.

YADO (hotel)
YADO NO NIWA

hotel of YARD

The yard of the hotel.

13. BULL

14. COW

15. OX

JP:

JP:

122
ERABURU (bossy)

ERABURU OUSHI
bossy BULL

Bossy bull.
KAU (buy)

MIRUKU TO MEUSHI O° KAU
milk and cow BUY

Buy milk and a cow.
OOKUSU (increase)

(SHOKUYOU. NIGURUMAYOU NO)
(for meat, for transportation) of

OUSHI O OOKUSU
male bovine animal increase

Increase male bovine animals for meat and
transportation.

APPENDIX B

THE RESULTS OF THE T-TEST

123

 

 

 

 

 

 

 

 

 

Overall
Number Of Q-tail
Variable pairs Owrx Sig Mean SD SE of Mean
KPOST 11.1103 4.145 .210
390 .649 .000
RPOST 10.4000 4.306 .218
Paired Differences
Mean SD SE of Mean t-value df 2-tai1 Sig
.7103 3.546 .180 3.96 389 .000
953 CI (.357. 1.063)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
KFOLL 5.7744 3.764 .191
390 .621 .000
RFOLL 4.4462 3.203 .162
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
1.3282 3.077 .156 8.53 389 .000
951 CI (1.022. 1.635)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
KDIF ‘ 5.3359 3.697 .187
390 .342 .000
RDIF 5.9538 3.727 .189
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
-.6179 4.259 .216 -2.87 389 .004

95% CI {-1.042. -.194)

124

One-month follow-up

 

 

 

 

 

 

 

 

 

Number of 2-Lail
Variable pairs Corr Sig Mean SD SE of Mean
KPOST 11.2460 3.940 .288
187 .602 .000
RPOST 10.1176 4.314 .315
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
1.1283 3.696 .270 4.17 186 .000
95% CI (.595. 1.662)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
KFOLL 5.7861 3.827 .280
187 .680 .000
RFOLL 4.5241 3.391 .248
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
1.2620 2.913 .213 5.92 186 .000
95% CI (.842. 1.682)
Number of 2-ta11
Variable pairs Corr Sig Mean SD SE of Mean
KDIF 5.4599 3.833 .280
187 .330 .000
RDIF 5.5936 3.699 .271
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
-.1337 4.360 .319 -.42 186 .675

95% CI (-.763. .495)

125

Two-month follow-up

 

 

 

 

 

 

 

 

 

Number of 2-fail
Variable pairs Cori Sig Mean SD SE of Mean
KPOST 10.9852 4.332 .304
203 .696 .000
RPOST 10.6601 4.292 .301
Paired Differences
Mean SD SE of Mean t—value df 2-tail Sig
.3251 3.364 .236 1.38 202 .170
95% C1 (-0141: .791)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
KFOLL 5.7635 3.716 .261
203 .558 .000
RFOLL 4.3744 3.026 .212
Paired Differences
Mean SD SE of Mean t-value df 2—tail Sig
1.3892 3.226 .226 6.14 202 .000
95% CI (.943. 1.836)
Number of 2-tail
Variable ‘pairs Corr Sig Mean SD SE of Mean
KDIF 5.2217 3.573 .251
203 .363 .000
RDIF 6.2857 3.731 .262
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
-1.0640 4.124 .289 -3.68 202 .000

95% CI {-1.635. -.493)

APPENDIX C

THE DATA ON THE 12 CLASSES

126

The posttest (overall)

- - Description of Subpopulations - -

Summaries of KPOST

By levels of CLASS

Variable Value Label Mean
For Entire Population 11.1103
CLASS 1 11.7143
CLASS 2 9.8000
CLASS 3 9.7586
CLASS 4 11.9677
CLASS 5 10.8947
CLASS 6 11.2778
CLASS 7 10.7241
CLASS 8 7.1786
CLASS 9 12.0286
CLASS 10 11.9412
CLASS 11 12.2973
CLASS 12 12.6571

Total Cases = 390

- - Description of Subpopulations - —

Summaries of RPOST

By levels of CLASS

Variable Value Label Mean
For Entire Population 10.4000
CLASS 1 9.8571
CLASS 2 10.4000
CLASS 3 8.7241
CLASS 4 8.9677
CLASS 5 11.6579
CLASS 6 10.5556
CLASS 7 10.0000
CLASS 8 7.8571
CLASS 9 11.9714
CLASS 10 10.4118
CLASS 11 11.4595
CLASS 12 11.6286

Total Cases = 390

Std Dev

NUUUb-fiwbwwbb

.1453

.2588
.7879
.5421
.9957
.9472
.9829
.7876
.3890
.4682
.0445
.2047
.5775

Std Dev

4

bUbNU‘A-bh’hmfib

.3057

.0980
.4613
.1610
.2464
.8923
.2456
.5826
.0164
.8126
.4389
.2197
.0663

Cases
390

28
30

31
38
36
29
28
35

37
35

127

The follow-ups (overall)

- - Description of Subpopulations - -

Summaries of KFOLL

By levels of CLASS

Variable Value Label Mean
For Entire Population 5.7744
CLASS 1 4.0357
CLASS 2 4.4333
CLASS 3 4.6207
CLASS 4 5.9355
CLASS 5 5.0000
CLASS 6 5.3056
CLASS 7 6.3448
CLASS 8 4.0357
CLASS 9 7.4000
CLASS 10 7.6176
CLASS 11 7.3243
CLASS 12 6.3143

Total Cases = 390

- - Description of Subpopulations - -

Summaries of RFOLL

By levels of CLASS

Variable Value Label Mean
For Entire Population 4.4462
CLASS 1 2.8929
CLASS 2 4.3667
CLASS 3 4.2759
CLASS 4 4.0645
CLASS 5 4.0789
CLASS 6 3.4167
CLASS 7 4.8966
CLASS 8 3.2857
CLASS 9 5.5429
CLASS 10 5.5882
CLASS 11 5.0270
CLASS 12 5.4286

Total Cases : 390

Std Dev

UUUNAwabUWLa-D

.7643

.3387
.5592
.6195
.0409
.6981
.2586
.4772
.1854
.9921
.9236
.5360
.5378

Std Dev

U

NNMWNUNNWUUN

.2028

.9732
.9347
.7502
.0543
.9074
.4187
.5591
.8785
.3549
.6756
.9673
.9930

Cases

390

28
30
29
31
38
36
29
28
35
34
37
35

APPENDIX D

THE DATA ON STUDENTS WITH PERFECT SCORES ON EITHER

HALF OF THE IMMEDIATE POSTTEST OR BOTH TESTS

128

 

 

 

 

 

 

 

 

 

 

88. u 38808 __s-m SB. A 8:22.00.
.88. A £3868 :68 ”.8. A 88.680.

mm... 8.2” Baum. 30mm
08. mm :3. 84. 83 28.. 8

84. 8:” 88: 30$.

8m. .8.“ 88.9 50%
o8. 8 EB 8m. :3 88w 8

08. 08. 88.3 50%.

.02 N 3-30:0”.

8... 89m 858 308
8. 8 88 63.. 83 83.?

OS. Sam :88 t. 3on

84. 23 88S 508m
08. 8 8.6 mm... 9%.. SEN

88. 80. 88.9 t. 50%.

.05. F 5-30:0“.

in. mm; 88.8 508m
08. 89 as .8. 8:. 83a 5.

88. 88. 88.2 50%
.DOLQ .C 03.0) Cotw .Em SOD .3m :00: Cotw SOD COOS— mwmmo 0 @—
__m.-m -. oucBoED .Bw .Ew .0 .02 .D LN>

 

 

 

 

 

 

 

 

 

.6059: 2°59. on. o. 88.2 «98m .8th 8282 on; $538 Pop on. .8 San

 

129

So u £5868 .85 «we. A 88.880.

.80. A £3208 :28 68. u 88.2.8...

 

 

 

 

 

 

 

 

 

 

va. mmmN mvovh ...:Oum
moo. mv w—.m va. vsmd Nmmmé xv

mwv. med vnmmm cJJOmX

GOO. GOO. COCOWF hwomm
OOO wv mom..- wvw. vmwe ovmmé- xv

mvw. vow; owmn .mw HMOQ!

.02 N 3-26:0...

wow. vaN 000m.m «440mm
000, mm mad 00v. ONmN mmmmé om

mam. QNNh mmm...® 340...!

GOO. GOO. oooomp PWOQE
000 mm VNAT vow. ommN hmmmé- om

vmv. mmmd mmmodw FWOn—X

.05. F 3-30:0“.

000. 000. cocoa: HWOQK
000. ms mv.®- mmw. mNoN mmwmé- KN

wmm. mNON mowvflw Hwoav.
poi o:_m> Lotw .05 $00 .Em cam—2 Lotw 500 330

- a - . . :85. . m_nm:m>

__ma N u 00:2”go vum ”um *0 OZ .

 

 

 

 

 

 

 

 

 

.noEmE 22 on. o. 8:29 «.208 Boron 8289. on; 2538 K on. .8 San

 

130

 

 

 

 

 

 

 

 

 

 

.38. n 5588 __s-m 36. A 88.680.
.08. u 8.888 __s-~ no.2. A 88.680...
NS. 35 3.8.4 30%.
8 a a; Em. 8:. 33w 8
So. 3.8a 8m: .3on
o8. 88. 88.3 50%.
08 3 a 8. o8. o8. 88. mm
88. 08. 88.3 5on
.22 N 93-326...
So. 23 88.8 3on
m8. 3 88 wt. 83 88d :
8: Ed 2m; 38x
88. 88. 88.3 52¢
83 3 8. 8o. 80. 88. 3 3
88. 80. 88.3 508.
.22 _. 9.30:0“.
08. 88. 88.3 5on
80 3 mm 8. 08. 80. 88. 8
88. 8o. 88.3 508.
teem .2m .60 .sm :8: .
nod o:_m> Lot >o moms
- a - . w . o :82 . o mBmcm>

 

 

 

 

 

 

 

 

 

3053: 58 65 2 626.2 3388 .8th 8282 on; 3:258 mm on. .2 3mm.

 

APPENDIX E

THE DATA ON HARD WORDS

131

 

 

 

 

 

 

Overall
— — — f-fes‘s . paired samples - -
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKPOST .6829 .407 .021
390 .340 .000
HWRPOST .7440 .378 .019
Paired Differences
Mean SD SE of Mean t-value df 2—tail Sig
-.0611 .451 .023 -2.67 389 .008
95% CI (-.106. -.016)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKFOLL .3564 .394 .020
390 .302 .000
HWRFOLL .2714 .337 .017
Paired Differences
Mean SD SE of Mean t-value df 2-tail Sig
.0850 .435 .022 3.86 389 .000
95% CI (.042. .128)

132

One-month follow-up

t-tests for paired samples -

 

 

 

 

 

 

Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKPOST .7130 .379 .028
187 .331 .000
HWRPOST .7264 .400 .029
Paired Differences
Mean SD SE of Mean t—value df 2-tail Sig
-.0134 .451 .033 -.41 186 .685
95% CI (-.078. .052)
Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKFOLL .3690 .398 .029
187 .304 .000
HWRFOLL .2968 .371 .027
Paired Differences
Mean SD SE of Mean t-value df 2-tai1 Sig
.0722 .454 .033 2.17 186 .031
95% CI (.007. .138)

133

Two-month follow-up

— - - t-tests for paired samples - - —

 

 

 

 

 

 

Number of 2-tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKPOST .6552 .430 .030
203 .362 .000
HWRPOST .7603 .356 .025
Paired Differences
Mean SD SE of Mean t-value df 2—tail Sig
-.1051 .448 .031 -3.34 202 .001
95% CI (-.167. -.043)
Number of 2—tail
Variable pairs Corr Sig Mean SD SE of Mean
HWKFOLL . .3448 .391 .027
203 .299 .000
HWRFOLL .2479 .303 .021
Paired Differences
Mean SD SE of Mean t—value df 2-tail Sig
.0969 .417 .029 3.31 202 .001

95% CI (.039. .155)

APPENDIX F

LETTER OF APPROVAL FROM UCRIHS AND

PRINCIPAL CONSENT FORM

 

134

MICHIGAN STATE UNIVERSITY

 

orrICE OI VICE PRESIDENT EOR RLSEARCH [A57 LANSING . Mycmgm . “82.40“,
AND DEAN or THE GRADUATE SCHOOL

April 17, 1991

Mamoru Kinjo
E-714 Owen Hall

RE: AN INVESTIGATION INTO lTflZ KEYWORD ETFCTIVENESS AMONG JAPANESE SEVENTH
GRADERS IN LEARNING A FOREIGN LANGUAGE VOCABULARY, IRB#9I-176

Dear Mr. Kinjo:

The above project is exempt from full UCRIHS review. The proposed research
protocol has been reviewed by another committee member. The rights and welfare
of human subjects appear to be protected and you have approval to conduct the
research.

You are reminded that UCRIHS approval is valid for one calendar year. If you
plan to continue this project beyond one year, please make provisions for
obtaining apprOpriate UCRIHS approval one month prior to April 15, 1992.

Any changes in procedures involving human subjects must be reviewed by UCRIHS
prior to initiation of the change. UCRIHS must also be notifed promptly of any
problems (unexpected side effects, complaints, etc.) involving human subjects
during the course of the work.

Thank you for bringing this project to my attention. If I can be of any future
help, please do not hesitate to let me know.

Sincerely,

David E. Wright, Ph.D., Chair

University Committee on Research Involving
Human Subjects (UCRIHS)

DEN/deo

cc: Dr. Paul Munsell

MSU is an A/firmatiw Action/Equal Opportunity Institution

B5

Safety of students involved in the study

The techniques and purpose of ‘this statistical study on
vocabulary learning and retention/acquisition by Professor Mamoru
Kinjo at Okinawa Kokusai University have been explained to me. I
recognise that the activities involved in this study are safe and

are not incompatible with normal classroom procedures.

mfifluafixgkwfiiifi«®£$fimomr

WQE%X$QQWQ$%K;5TE$§®$E~E§$K0MTJ®
W%&H®75X§AKODI,Exbtwfim.fi&.$M$®fimfi5
Ditto %n%mfifi®#§.%gfifi®$2EE®Efl.Xuemfi
Efituaabotfiman.E§$fimbamfi§éMiétocuuu
CtéfifihttiTo

fiz ¢$fi me E 8
NAME SCHOOL TITLE Principal DATE

Mamoru Kinjo
vflWflfl-KWXKP

em .

WREfiXf

BIBLIOGRAPHY

BIBLIOGRAPHY

Allen, V. F. (1983). Techniques in teaching vocabulary. Oxford: Oxford
University Press.

Ard, J., & Gass, S. M. (1987). Lexical constraints on syntactic acquisition.
Studies in Second Language Acquisition, .9, 233-2 52.

Atkinson, R. C. (1975). Mnemotechnics in second-language learning. American
Psychologist, 30, 821 -828.

Blair, R. W. (1982). Easification. In R. W. Blair (Ed.), Innovative approaches to
language teaching (pp. 219-225). Rowley, MA: Newbury House
Publishers.

Bond, 0. F., Babcock, J., Norman, H., & Andrus, L. (1943). Teaching the

romance languages: French, Spanish, Italian. Review of Educational
Research 13, 142-161.

Brown, R., 8. McNeill, D. (1966). The tip of the tongue phenomenon. Journal of
Verbal Learning and Verbal Behaviour, 5, 325-377.

Carter, R. A. (1987). Vocabulary. London: Allen & Unwin.

Carter, R., 8. McCarthy, M. J. (Eds). (1988). Vocabulary and language teaching.
London: Longman.

Channell, J. (1988). Psycholinguistic considerations in the study of L2
vocabulary acquisition. In R. Carter & M. McCarthy (Eds), Vocabulary
and language teaching (pp. 83-96). London: Longman.

Chomsky, N. (1957). Syntactic structures. The Hague: Mouton.

Cohen, A. D. (1987). The use of verbal and imagery mnemonics in second-

Ianguage vocabulary learning. Studies in Second Language Acquisition,
.9, 43-62.

136

137

Cohen, A. D. (1990. Language learning: Insights for learners, teachers, and
researchers. New York: Newbury House.

Cohen, A., 8 Aphek, E. (1980). Retention of second language vocabulary over
time: Investigating the role of mnemonic associations. System, 8, 221-
235.

Cohen, A., 8Aphek, E. (1981). Easifying second language learning. Studies in
Second Language Acquisition, 3, 221 -2 36.

Cohen, A., 8 Aphek, E. (1987). The use of verbal and imagery mnemonics in
second-language vocabulary Ieaming. Studies in Second Language
Acquisition, .9. 1, Cambridge University Press.

Cohen, A. D., 8 Hosenfeld, C. (1981). Some uses of mentalistic data in second
language research. Language Learning, 31, 285-313.

Cowie, A. P. (1988). Stable and creative aspects of vocabulary use. In R.
Carter 8 M. McCarthy (Eds), Vocabulary and language teaching (pp. 126-
139). London: Longman.

Craik, F. I. M., 8 Lockhart, R. S. (1972). Levels of processing: A framework for
memory research. Journal of Verbal Learning and Verbal Behavior, 17,
671-684.

Curtis, M. E. (1987). Vocabulary testing and vocabulary instruction. In M. C.
McKeown 8 M. E. Curtis (Eds), Nature of vocabulary acquisition (pp. 37-
51). Hillsdale, NJ: Erlbaum.

Desrochers, A., 8 Begg, l. (1987). A theoretical account of encoding and retrieval
processes in the use of imagery-based mnemonic techniques: The
special case of the keyword method. In M. A. McDaniel 8 M. Pressley
(Eds), Imagery and related mnemonic processes: Theories, individual
difl'erences, and applications (pp. 56-77). New York: Springer-Venag.

Desrochers, A., Gelinas, C., 8 Wieland, L. D. (1989). An application of the
mnemonic keyword method to the acquisition of German nouns and the
grammatical gender. Journal of Educational Psychology, 81, 25-33.

Dretzke, B. J., 8 Levin, J. R. (1990). Building factual knowledge about the US
presidents via pictorial mnemonic strategies. Contemporary Educational
Psychology, 15, 152-169.

138

Estabrooks, G. H. (1927). A handy memory trick. Journal of Genetic
Psychology. 34. 615-619.

Fenaigle, G. von. (1813). The new art ofmemory. London: Sherwood, Neely, 8
Jones.

F ries, C. C. (1945). Teaching and learning English as a foreign language. Ann
Arbor: University of Michigan Press.

Fuentes, E. J. (1976). An investigation into the use of imagery and generativity
in learning a foreign language vocabulary (Doctoral dissertation, Stanford
University). Dissertation Abstracts International, 37: 2694A. (University
Microfilms No. 76-25, 997)

Furst, B. (1972). Stop forgetting. New York: Doubleday.

Gass, S. (1988). Integrating research areas: Aframework for second language
studies. Applied Linguist/cs, 9, 198-217.

Gass, S. M. (1989). Vocabulary in a second langauge. Studies in Second
Language Acquisition, 17, 121-122.

Gorfein, D. S., 8 Hoffman, R. R. (1987). Memory and learning: The Ebbinghaus
Centennial Conference. Hillsdale, NJ: Erlbaum.

Hall. J. W., 8 Fuson, K. C. (1986). Presentation rates in experiments on
mnemonics: A methodological note. Journal of Educational Psychology,
78, 233-234.

Hall, J. W., 8 Fuson, K. C. (1988). The keyword method and presentation rates:
Reply to Pressley (1987). Journal of Educational Psychology, 80, 251-
252.

Higa, M. (1965). The psycholinguistic concept of "difficulty“ and the teaching of
foreign language vocabulary. Language Learning, 15, 167-179.

Horowitz, L. M., White, M. A., 8 Atwood, D. W. (1968). Word fragments as aids
to recall: The organization of a word. Journal of Experimental Psycholo-

gy, 76, 219-226.

Isgur, J. (1975. Establishing letter-sound associations by an object-imaging-
projection method. Journal of L earning Disabilities, 8, 16—20.

Julliot, Ch.-L. (1919). Léducation de la mémoire. Paris: Flammarion.

139

Kelly, L. G. (1969). Twenty-five centuries of language teaching. Rowley, MA:
Newbury House.

Konopak, B., 8 Williams, N. L. (1988). Using the keyword method to help young
readers learn content material. Reading Teacher, 41, 682-686.

Lado, R. (1955). Patterns of difficulty in vocabulary. Language Learning, 6. 23-
41.

Laubach, F. 0., Kirk, E. M., 8 Laubach, R. S. (1971). The new streamlined
English series: Teacher's manual. Syracuse, NY: New Readers Press.

Levin, J. R., Johnson, D. D., Pittelman, S. D., Hayes, B. L., Levin, K. M., Shriberg,
L. K., 8 Toms-Bronowski, S. (1984). A comparison of semantic— and
mnemonic-based vocabulary-learning strategies. Reading Psychology, 5,
1-15.

Levin, J. R., McCormick, C. B., Miller, G. E., Berry, J. K., 8 Pressley, M. (1982).
Mnemonic versus nonmnemonic vocabulary-Ieaming strategies for
children. American Educational Research Journal, 19, 121-136.

Levin, J. R., Pressley, M., McCormick. C. B., Miller, G. E., 8 Shriberg, L. K.
(1979). Assessing the classroom potential of the keyword method.
Journal of Educational Psychology, 71, 583-594.

Liu, l.-M. (1978). Appraisal of the keyword method in second-language learning.
Proceedings of the National Science Council (Taiwan). 2. 437-442.

Loisette, A. (1895). Memory: A scientific practical method of cultivating the
faculties of attention, recollection and retention. New York: Funk 8
Wagnalls.

Lorayne, H. (1974). How to develop a super-power memoor. New York: Signet.

Martin, C. J. (1978). Mediatational processes in the retarded: Implications for
teaching reading. In N. R. Ellis (Ed), International review of research in
mental retardation (Vol. 9. Pp. 61-84). New York: Academic Press.

McDaniel, M. A., 8 Pressley, M. (1984). Putting the keyword method in context.
Journal ofEduwt/onal Psychology, 76; 598-609.

McDaniel, M. A., 8 Tillman, V. P. (1985). The keyword method versus learning
tram context: Where's the link? Manuscript submitted for publication,
Department of Psychology, University of Notre Dame, South Bend, IN.

140

McDaniel, M., 8 Pressley, M. (1989). Keyword and context instruction of new
vocabulary meanings: Effects on text comprehension and memory.
Journal of Educational Psychology, 81, 196-207.

McKeown, M. G., 8 Curtis, M. E. (Eds.). (1987). The nature of vocabulary
acquisition. Hillsdale, NJ: Erlbaum.

Meara, P. (1980). Vocabulary acquisition: A neglected aspect of language
learning. Language Teaching and Linguistics: Abstracts, 1.3, 221-246.

Meara, P. (1982). Word associations in a foreign language: A report on the
Birkbeck Vocabulary Project. Nottingham Linguistic Circular, 11(2), 29-38.

Meara, P. M. (Ed.). (1983). Vocabulary in a second language (Vol. 1). London:
Center for Information on Language Teaching and Research.

Meara, P. (1984). The study of lexis in interlanguage. In A. Davies, A. Howart,
8C. Criper(Eds.), lnterlanguage (pp. 225-235). Edinburgh: Edinburgh
University Press.

Meara, P. (Ed). (1987). Vocabulary inasecond language (Vol. 2). London:
Center for Information on Language Teaching and Research.

Merry, R. (1980). Keyword method and children’s vocabulary learning in the
classroom. British Journal of Edumtional Psychology, 50, 123-136.

Miller, G. E., Levin, J. R., 8 Pressley, M. (1980). An adaptation of the keyword
method to children’s learning of foreign verbs. Journal of Mental Imagery,
4, 57-61.

Ministry of Education, Japan. (1977). Chyugaklrou gakushyu shidou youriyou
[Guidelines for junior high schools], 102-107.

Munsell, P. E., Rauen, M. G., 8 Kinjo, M. (1988). Review essay:
Comprehensive survey of recent conclusions. Language Learning, 3&2),
261-278.

Nation, l. S. P. (1990). Teaching and learning vocabulary New York: Newbury
House.

Nattinger, J. (1988). Some current trends in vocabulary teaching. In R. Carter 8
M. McCarthy (Eds.), Vocabulary and language teaching (pp. 62-82).
London: Longman.

141

Ogden, C. K. ( 1930). Basic English: A general introduction. London: Kegan
Paul, Trench 8 Trubner.

Ott, C. E., Butler, D. C., Blake, R. S., 8 Ball, J. P. (1973). The effect of
interactive-image elaboration on the acquisition of foreign language
vocabulary. Language Learning, 2, 197-206.

Paivio, A. (1969). Mental imagery in associative Ieaming and memory.
Psychological Re view, 78, 241 -263.

Paivio, A. (1971). Imagery and verba/processes. New York: Holt, Rinehart, 8
Winston.

Paivio, A., 8 Desrochers, A. (1981). Mnemonic techniques in second-language
learning. Journal of Educational Psychology, 73, 780-795.

Peters. A. (1983). The units oflanguage acquisition. Cambridge, NY:
Cambridge University Press.

Peterson, P. L., 8 Swing, S. R. (1983). Problems in classroom implementation
of cognitive strategy instruction. In M. Pressley 8 J. R. Levin (Eds),
Cognitive strategy research: Educational applications (pp. 267-287). New
York: Springer.

Pressley, M. (1977). Children’s use of the keyword method to learn simple
Spanish vocabulary words. Journal of Educational Psychology, 69, 465-
472.

Pressley, M. (1987). Are keyword method effects limited to slow presentation
rates? An empirically based reply to Hall and Fuson (1986). Journal of
Educational Psychology, 79, 333-335.

Pressley, M., Borkowski, J. G., 8 O’Sullivan, J. T. (1984). Memory strategy
instruction is made of this: Metamemory and durable strategy use.
Educational Psychologist, 19, 94-107.

Pressley, M., 8 Dennis-Rounds, J. (1980). Transfer of a mnemonic keyword
strategy at two age levels. Journal of Educational Psychology, 72, 575-
582.

Pressley, M., 8 Levin, J. R. (1977). Developmental differences in subjects’
associative learning strategies and performance: Assessing a hypothesis.
Journal of Experimental Child Psychology, 24, 431 -439.

142

Pressley, M., 8 Levin, J. R. (1978). Developmental constraints associated with
children’s use of the keyword method of foreign language vocabulary
Ieaming. Journal of Experimental Child Psychology, 28, 359-372.

Pressley, M., Levin, J. R., 8 Chatala, E. S. (1984). Memory strategy monitoring
in adults and children. Journal of Verbal Ieaming and Verbal Behavior, 23,
270-288.

Pressley, M., Levin, J. R., 8 Delaney, H. D. (1982). The mnemonic keyword
method. Re view of Educational Research, 52, 61 -92.

Pressley, M., Levin, J. R., Kuiper, N. A., Bryant, S. L., 8 Michener, S. (1982).
Mnemonic versus nonmnemonic vocabulary-Ieaming strategies: Addition-
al comparisons. Journal of Educational Psychology, 74, 693-707.

Pressley, M., Levin, J. R., 8 McCormick, C. B. (1980). Young children’s learning
of foreign language vocabulary: A sentence variation of the keyword
method. Contemporary Educational Psychology, 5, 22-29.

Pressley, M., Levin, J. R., 8 McDaniel, M. A. (1987). Remembering versus
inferring what a word means: Mnemonic and contextual approaches. In
M. C. McKeown 8 M. E. Curtis (Eds), Nature of vocabulary acquisition
(pp. 107-127). Hillsdale, NJ: Erlbaum.

Pressley, M., McDaniel, M. A., Turnure, J. E., Wood, E., 8 Ahmad, M. (1987).
Generation and precision of elaboration: Effects on intentional and
Incidental Ieaming. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 13, 291 -300.

Pressley, M., Ross, K. A., Levin, J. R., 8 Ghatala, E. S. (1984). The role of
strategy utility knowledge in children’s strategy decision making. Journal
of Experimental Child Psychology, .38, 491 -504.

Pressley, M., Samuel, J., Hershey, M. M., Bishop, 8. L., 8 Dickinson, D. (1981).
Use of a mnemonic technique to teach young children foreign language
vocabulary. Contemporary Educational Psychology, 8, 110-116.

Pressley, M., Symons, S., McDaniel, M. A., Snyder, B. L., 8 Turnure, J. E.
(1988). Elaborative interrogation facilitates acquisition of confusing facts.
Journal of Educational Psychology, 80, 268-278.

143

Pressley, M., Woloshyn, V., Lysynchuk, L. M., Martin, V., Wood, E.. 8
Willoughby, T. (1990). A primer of research on cognitive strategy
instruction: The important issues and how to address them. Educational
Psychology Re view, 2, 1-58.

Raugh, M. R., 8 Atkinson, R. C. (1975). A mnemonic method for learning a
second language vocabulary. Journal of Educational Psychology, 87, 1-
16.

Richards, I. A. (1943). Basic English and its use. London: Kegan Paul.

Rohwer, W. 0., Jr. (1973). Elaboration and Ieaming in childhood and
adolescence. In H. W. Reese (Ed), Advances in child development and
behavior (Vol. 8). New York: Academic Press.

Rosenheck, M. B., Levin, M. E., 8 Levin, J. R. (1989). Learning botany concepts
mnemonically: Seeing the forest and the trees. Journal of Educational
Psychology, 81, 196-203.

Sapir, E. (1921). Language. New York: Harcourt, Brace. 8 Co.
Sayer, T. A. (1877). Aids to memory. London: Daldy, Isbister 8 Co.

Scruggs, T. E., 8 Mastropieri, M. A. (1989). Reconstructive elaborations: A
model for content area learning. American Educational Research Journal,
28: 311-327.

Shepard, R. N. (1967). Recognition memory for words, sentences and pictures.
Journal of Verbal Learning and Verbal Behavior, 8, 156-163.

Shriberg, L. K., Levin, J. R., McCormick, C. B., 8 Pressley, M. (1982). Learning
about “famous people” via the keyword method. Journal ofEducational
Psychology, 74(2), 238-247.

Standing, L., Conezio, J., 8 Haber, R. N. (1970). Perception and memory for
pictures: Single trial learning of 2560 visual stimuli. Psychonomic
Science, 19, 73-74.

Stevick, E. W. (1976). Memory, meaning and method. Rowley, MA: Newbury
House.

Stevick, E. W. (1986). Images and Options in the language classroom.
Cambridge, UK: Cambridge University Press.

144

Stevick, E. W. (1989). Success with foreign languages: Seven who achieved it
and what worked for them. Englewood Cliffs. NJ: Prentice-Hall.

Thorndike, R. L. (1973). Reading comprehension education in fifieen countries.
New York: Wiley.

Tulving, E., 8 Thompson, D. M. (1973). Encoding specificity and retrieval.
processes in episodic memory. Journal of Experimental Psychology, 80,
352-373.

West, M. P. (1953). A general service list of English words. London: Longman.

Wilkins, D. A. (1972). Linguistics and language teaching. London: Edward
Arnold.

Willerman, B., 8 Melvin, B. (1979). Reservations about keyword mnemonic.
Canadian Modern Language Re view, 35, 443-453.

Xue-G-Y., 8 Nation, P. (1984). A university word list. Language Learning and
Communication, 3(2), 215-229.

Yates, F. A. (1966). The art ofmemory. Chicago: University of Chicago Press.