Variables Affecting Choice of Language Learning Strategies by University Students

Author(s): Rebecca Oxford and Martha Nyikos

Source: The Modern Language Journal, Vol. 73, No. 3 (Autumn, 1989), pp. 291-300
Published by: Wiley on behalf of the National Federation of Modern Language Teachers Associations
Stable URL: http://www.jstor.org/stable/327003 .
Accessed: 16/01/2014 10:20

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Wiley and National Federation of Modern Language Teachers Associations are collaborating with JSTOR to
digitize, preserve and extend access to The Modern Language Journal.

http://www.jstor.org

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
 Choice ofLanguage
VariablesAffecting
LearningStrategies
byUniversityStudents
REBECCA OXFORD and MARTHA NYIKOS
ofAlabama
University IndianaUniversity

IN THIS ARTICLE WE DISCUSS VARIABLES ing and writing)forovercomingdeficienciesin

affectingchoice of learning strategiesused by knowledgeof the language." Appropriatelearn-
1,200 foreignlanguage students in a conven- ing strategieshelp explain the performanceof
tional academic setting,a major universityin good language learners; similarly,inappropri-
the midwesternUSA. In termsof the number ate learningstrategiesaid in understandingthe
of subjects involved, this investigationis prob- frequent failures of poor language learn-
ably one of the largest learning studies to date ers - and even the occasional weaknesses of
in anyinstructionalfield,and is almostcertainly good ones.6
the largest completed studyof language learn- Use of appropriate learning strategies
ing strategies.' enables studentsto take responsibilityfortheir
own learningby enhancing learner autonomy,
RESEARCH BACKGROUND independence, and self-direction.7 These
factorsare importantbecause learners need to
Learning strategiesare operations used by
learnersto aid the acquisition, storage,and re- keep on learningeven when theyare no longer
trievalof information(52). Outside of the lan- in a formalclassroom setting(42). Moreover,
guage learning field, research comparing cognitivepsychologyshowsthatlearningstrate-
experts to novices indicates that experts use gies help learners to assimilate new informa-
more systematic and useful problem-solving tion into their own existingmental structures
and native-language reading comprehension or schemata,thuscreatingincreasinglyrichand
strategies.2A similar findingoccurs withmore complex schemata.8 As theymove toward lan-
successfullanguage learnersas compared to less guage proficiency,language learners develop
successful ones.3 Better language learners their own understandings or models of the
second or foreignlanguage and its surround-
generallyuse strategiesappropriateto theirown
stage of learning, personality,age, purpose for ing culture. Unlike most other characteristics
of the learner, such as aptitude, attitude,moti-
learning the language, and type of language.4
Good language learners use a varietyof learn- vation, personality,and generalcognitivestyle,
ing strategies, including cognitive for
strategies learning strategies are readily teachable.9
Various researchers have studied factors
associating new informationwith existing in-
formationin long-termmemoryand forform- related to choice of language learning strate-
ing and revisinginternalmental models; meta- gies, as shown in a review by Oxford (40).
forexercising"executive con- These factors include: 1) language being
cognitivestrategies
trol" through planning, arranging, focusing, learned; 2) level of language learning, profi-
and evaluatingtheirown learningprocess; social ciency, or course; 3) degree of metacognitive
strategiesfor interactingwith others and man- awareness; 4) sex; 5) affectivevariables such
for directing as attitudes, motivation, and language learn-
aging discourse; affectivestrategies
feelings,motivations, and attitudes related to ing goals; 6) specificpersonalitytraits;7) over-
all personalitytype; 8) learningstyle;9) career
learning; and compensation strategies(such as
orientation or field of specialization; 10)
guessing unknown meanings while listening
and reading, or using circumlocutionin speak- national origin; 11) aptitude; 12) language
teaching methods; 13) task requirements;and,
ifrelevant, 14) typeofstrategytraining.Many
The ModernLanguageJournal, 73, iii (1989) of thesefactors,such as language learninglevel,
0026-7902/89/0003/291'
$1.50/0 national origin,fieldof specialization, and lan-
?1989 The ModernLanguageJournal
guage teachingmethods,have been definitively

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
292 RebeccaOxford& MarthaNyikos
shown to be stronglyrelatedto language learn- out native speakers of the target language as
ers' choice of strategies; but others, such as conversationpartners. Respondents are asked
motivation and sex, have until now not re- to answerin termsofthelanguage theyare cur-
ceived sufficient
researchattentionto allow firm rentlylearning.
conclusions to be reached. For detailed, com- The SILL has been used around the world
prehensiveresearchreviewson language learn- for students of second and foreignlanguages
ing strategies,see Oxford (40, 41). The study in universities,schools, and governmentagen-
reported in the current article investigated a cies. Strategy descriptionson the SILL were
number of the factorslisted above, including drawn froma comprehensivetaxonomy(35) of
some which have been frequentlystudied and language learningstrategiesthat systematically
otherswhichhave been inadequatelyexamined covers the four language skill areas of listen-
in the past. ing, reading,speaking,and writing.The taxon-
omy was based on an extensiveresearchreview
METHODS
(36, 38). Internal consistencyreliabilityusing
Cronbach's alpha is .96 based on a 1,200-
Sample. Slightlymore than 1,200 students, person universitysample (in the currentstudy)
including approximatelyequal proportionsof and .95 based on a 483-person Defense Lan-
men and women, participated in the study. guage Institute (DLI) field test sample (37).
They were almost all (97%) undergraduates, Content validity is .95 using classificatory
and were studyinga total of five differentfor- agreementbetweentwo independentraterswho
eign languages: French (40% of the sample), blindly matched each of the SILL items with
Spanish (28%), German (27 %), Russian (2 %), strategiesin thecomprehensivetaxonomycited
and Italian (2%).1o Almost all (95%) were above (37). Concurrent, and to some extent
native English speakers, and the same propor- construct,validitycan be assumed based on the
tion (95 %) was in the age range seventeen to demonstrationof strongrelationshipsbetween
twenty-three.Half of the studentswere major- SILL factorsand self-ratingsof language pro-
ing in technical fields, such as engineering, ficiencyand language motivation, as reported
computer science, or physical sciences; thirty- in this article. Additional evidence supporting
five percent were majoring in social sciences, validityis foundin a different studyby Ehrman
education, or humanities,while the rest(15 %) & Oxford (16), in which more highlytrained
majored in business and other subjects. linguists, in contrast to less highly trained
The sample consisted of relativelyinexperi- linguists, predictably reported significantly
enced language learners.Two-thirds(66 %) had more frequentand more wide-ranginguse of
studiedno foreignlanguages otherthan the one strategieson the SILL.
theywere currentlylearning at the university, Questions about respondents' truthfulness
while the balance (34 %) had previouslystudied sometimes arise with self-reportinstruments
at least one other foreign language. The like the SILL. To check for truthfulness,the
majority (72%) were in their firstor second developerofthe SILL compared its resultswith
semester of universitylanguage study, while informalinterviewdata gleaned fromthe 483-
twenty-threepercent were in their third or person field test and earlier clinical trials.
fourth semester, and the rest (5%) were Interview data and SILL data tended to be
enrolled in higher level courses. Seven out of mutually supportive, thus lending credibility
every ten were taking the foreignlanguage as (and furtherevidence of validity)to the SILL.
a graduationrequirement,while thirtypercent In addition, the SILL findings themselves,
chose language study as an elective. taken from several samples, were carefully
The main instrumentused in
Instrumentation. scrutinizedto determinewhetherany bias ap-
this study is the Strategy Inventory for Lan- peared, i.e., whether respondents systemati-
guage Learning, or SILL (37).11 This 121-item cally offered"socially desirable" answers. No
instrumentasks learnersto reportthefrequency such bias was evident. In fact, respondents
withwhich theyuse certain language learning seemed determined to rate their strategies
strategies. A typical SILL item asks the re- as honestlyas possible, even ifthese strategies
spondent to indicate, in a multiple-choice were not optimal. The guaranteeofanonymity,
fashion,the frequencyof use (almost always to and the fact that the SILL scores were not to
almost never, on a five-pointscale) of a given be used forperformanceevaluation (grading),
strategy,such as breakingdown an expression probably contributedto the apparent honesty
into parts in order to understand it, or seeking of the respondents.

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
LanguageLearningStrategies 293
In addition to the SILL, we also adminis- Factor One, formalrule-related
practice
strategies,
tered a backgroundquestionnairecoveringsex, received a high level of usage and contained
years of foreignlanguage study,electivevs. re- strategiessuch as using structuralknowledge,
quired course status, self-perceptionsof profi- findingsimilaritiesbetweenlanguages, generat-
ciency and motivation, and other topics. ing and revising rules, and analyzing words.
Research Two keyresearchquestions
Questions. Factor Two, functionalpractice included
strategies,
were addressed in the study. First, what kind the least frequentlyused of all the strategies,
of strategiesdo universityforeignlanguage stu- such as attendingforeignlanguage films,seek-
dents report using? Second, what variables ing native speakers forconversation,imitating
(sex, course status,motivationlevel, and so on) native speakers, initiating foreign language
influencethe use ofthesestrategies?This article conversations, and reading authenticmaterial
summarizesresultsrelatingto the firstquestion in the new language - all strategiesrequiring
but goes into detail concerning the second language practice in natural settingsoutside of
one. 12 the classroom. Factor Three, resourceful, inde-
Data Collectionand Analysis.Data collection pendent strategies,
comprisedrelativelylow-usage
was conducted by the second author, with the strategiesinvolving: 1) independentmanipula-
cooperation of other language teachers and tion of foreignlanguage material in order to
teaching assistants. Respondents received uni- embed it in memory (listing related words,
forminstructionsto fillout the SILL and the making up sentencesand exercises,using mne-
backgroundquestionnaire,and everyeffort was monics, elaborating sentences,using a tape re-
made to cause minimal classroom disruption. corder); and 2) independent use of certain
To understand the data we: 1) calculated de- metacognitive actions (planning, self-testing,
scriptivestatisticssuch as frequenciesto deter- self-reward).Strategiesin Factor Four, general
mine overall patterns;2) discernedthe underly- studystrategies, were reported to be used at
ing factorson the SILL throughfactoranaly- moderate to high frequencylevels. Factor Four
sis; 3) determinedthe variables which had the strategiesincluded such all-purposetechniques
greatest influence on the choice of learning as studyinghard, ignoringdistractions,being
strategies through analysis of variance prepared, organizing, and using time well.
(ANOVA) on SILL factorscores; and 4) when Factor Five, conversationalinputelicitation
strate-
necessary, conducted post-hoc tests to deter- gies, included strategiessuch as the following,
mine the precise contrasts in which the ob- reportedas used moderatelyoften: requesting
served significance occurred.13 slowerspeech, askingforpronunciationcorrec-
tion, and guessing what the speaker will say. 14
RESULTS In brief,these findingsindicate thatthe uni-
versitystudentsfrequentlyreportedemploying
AnswerstoResearch QuestionOne.To answerthe strategies (e.g., formal rule-related practice
firstresearch question, i.e., which kinds of strategiesand general study strategies) likely
strategies are used by university students to to be usefulin a traditional,structure-oriented,
learn a new language, we turnto the SILL fac- discrete-point foreign language instructional
tor analytic findings,summarized very briefly environment geared toward tests and assign-
here. Five main factors emerged from the ments. Strategieswhich involved a concerted,
analysis. In addition to identifyingthe factors, extracurriculareffortto communicate in the
we examined the frequencywith which strate- new language (functionalpracticestrategies)or
gies in each factorwere reportedto be used (see required workingindependentlyon mnemonic
Table I). or metacognitive aspects (resourceful, inde-
TABLE I
Factors and Frequencies

Factor Average Frequency of Use of Rank Order

Number Factor Name Strategies in This Factor of Usage
1 Formal rule-related practice strategies Medium to High 1
2 Functional practice strategies Low to Medium 5
3 Resourceful, independent strategies Low to Medium 4
4 General study strategies Medium to High 2
5 Conversational input elicitation strategies Medium to High 3

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
294 RebeccaOxford& Martha Nyikos

pendentstrategies)were mostlyshunnedby the probabilitiessmallerthan .0001 were common,

students in this sample. In general, the only indicatingan extremelylow likelihoodthatthe
communicative involvement shown with at results could have occurred by chance.'5 We
least moderatefrequencywas in conversational will examine each of the significantvariables
input elicitationstrategies,whichdid not neces- in turn (see Table II for details).
sarilydemand any outside-of-classinvolvement Motivation.The degree of expressed motiva-
with native speakers of the foreignlanguage tion to learn the language was the most power-
and which were used occasionally. For a much fulinfluenceon strategychoice. Motivationhad
more intensivediscussion of these factorana- extremelysignificanteffectson scores forfour
lyticresults,includingtheirinterpretationand (1,2,4,5) of the five factorsand a nearly sig-
a comparison with a SILL factoranalysis on nificanteffecton Factor Three. The more moti-
a very differentgroup of subjects, see Nyikos vated students used learning strategiesof all
and Oxford (30). these kinds more oftenthan did the less moti-
Answersto ResearchQuestionTwo. In answer- vated students.'6
ing the second research question, i.e., which Ratings.Language proficiencyself-
Proficiency
variables affectchoice of language learning ratingsin speaking, reading, and listeningalso
strategies,we examined the influence of self- stronglyaffectedstrategychoice. For instance,
perceptionsof motivationand proficiencyand speakingproficiencyratingswere highly influ-
the effectsof course status (required vs. elec- ential forFactors Two, Three, Four, and Five
tive), years of study, sex, and major. These and had a nearly significanteffecton Factor
variables were taken from the background One. Reading proficiency ratings had very
questionnaire.Results indicatedthatthedegree highlysignificanteffectson Factors One, Two,
of expressed motivation was the single most Four, and Five. Listeningproficiencyratings
powerful influence on the choice of language powerfullyinfluencedlearners'choice of strate-
learning strategies, that sex had a profound gies in two factors(3 & 4). The higherthe stu-
effecton strategychoice, and that all the other dent'sself-perceived proficiencyin each of these
variables listed- and some interactionsamong three skills, the more frequentlythe student
these variables - had significanteffectson the chose to use learning strategiesin the factors
reporteduse of strategies.In thisinvestigation, named here. Greater strategyuse accompanied

TABLE II
Significant Effectsof Background Variables on Factor Scores

Factors*

Variables 1 2 3 4 5

Sex .002 - - .0001 .002

Major - .02 .01 - -
Years of Study - .0001 - - .002
Course Status (Elective vs. Required) - .002 - .04 -

Speaking [1] - .0001 .02 .0002 .001

Listening [1] - - .0009 .004 -

Speaking & Listening [1, 2] - .04 - - -

Reading [1] .0001 .002 - .0001 .008

Motivation [1] .0001 .0001 - .0001 .0001
Course Status & Years of Study [2] - .01 - - -

Major & Motivation [2] - .005 - -

Major & Course Status [2] - - - .007
Course Status & Motivation [2] .01 - - -

Major, Course Status & Motivation [2] - .02 - - -

Sex & Motivation [2] .01 - -

Notes: Significant levels (probabilities) below .05 are shown in this table.
of the skill proficiency(e.g., reading), or of the variable (e.g., motivation).
[1] Self-rating
between the two variables.
[2] Interaction
*1 = formal rule-related practice strategies; 2 = functional practice strategies; 3 = resourceful, independent strategies;
4 = general study strategies; 5 = conversational input elicitation strategies.

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
LanguageLearningStrategies 295

perceptionsof higherproficiency,and a causal interactedforFactor Two (Interaction 6). Not

relationship actually existed between profi- very surprisingly,speakingand listeningprofi-
ciency self-ratingsand strategyuse.17 ciency self-ratingsinteracted for Factor Two
Electivevs. RequiredStatus.Clear differences (Interaction 7). 18
were found for elective vs. required course
statusforFactors Two and Four. For both, stu- DISCUSSION
dents who elected to learn the language rather
than takingit as a graduationrequirementused This discussion explores the reasons behind
these kinds of strategiesmore often. the cause-and-effect relationships just de-
YearsofStudy.Years spent studyingthe for- scribed. Note that causality is by definition
eign language had a very highly significant involved in the use of the ANOVA technique;
effecton two communicatively-oriented factors indeed, that is why researchers use it (when
(2, 5). In general terms,studentswho had been possible) instead ofless explanatorytechniques
studyingthe language for at least four or five like correlation.However, we feelthatthe rela-
years used strategieshere far more oftenthan tionshipsdemonstratedhere are only part of a
did less experienced language learners. More complex picture, which needs to be fullyde-
precisely, students studying the language at scribed in subsequent research. We start our
least fiveyears used Factor Two strategiesmore discussion with motivation,which exerted the
often than did students with less study, and strongestinfluence on strategychoice.
learners studying the language at least four ofMotivation.In this study,motivation
Effects
years used Factor Five strategies more often had a pervasive influenceon the reported useof
than did learners with less longevity. kindsofstrategies,
specific as well as on the degree
Sex. Profoundlysignificantsex differencesin of active involvementin language learning as
strategychoice were also evident for Factors reflectedin the overallfrequency ofstrategyuse in
One, Four, and Five. Females reportedmore general.Our findings about motivation support
frequentstrategyuse than males in these three Gardner's (20) statement:"Attitudesand moti-
factors,while males reportedno more frequent vation are importantbecause they determine
strategyuse than females on any factors. the extentto which the individuals will actively
Major. Universitymajor made a highlysig- involve themselves in learning the lan-
nificant difference for Factor Three with guage. . . . The prime determining factor is
humanities/social science/education majors motivation" (p. 56).
using themmore oftenthan theothertwo broad At firstthe relationshipbetween motivation
categories of majors examined. Humanities/ and strategyuse appears simple: learners who
social science/educationmajors used strategies are highly motivated to learn a language are
in Factor Two significantly more oftenthan did likely to use a variety of strategies. But the
technical majors, but not significantlymore motivationissue quickly becomes highlycom-
often than did their business counterparts. plex. Not only does high motivationlead to sig-
InteractionsbetweenMotivationand OtherVari- nificantuse of language learning strategies(as
ables. Motivation significantlyinteractedwith we found in this study), but high strategyuse
several variables, often in complex ways, probably leads to high motivation as well!
to influencechoice of strategies.For instance, Based on researchfindingson skilldevelopment
in a significanteffecton Factor One, motiva- and self-esteem,we would expect that use of
tion interacted with sex (Interaction 1). appropriate strategiesleads to enhanced actual
Another significantinteractionaffectingFac- and perceivedproficiency, whichin turncreates
tor One was between motivation and elective high self-esteem,which leads to strong moti-
vs. required course status (Interaction 2). In vation, spirallingto stillmore use of strategies,
a significantinteractionaffectingFactor Two, great actual and perceived proficiency,high
motivation interacted with university major self-esteem,improvedmotivation,and so on.19
(Interaction 3). In addition, motivation inter- When viewed in lightof thischain ofvariables,
acted with both course status and major self-perceptionsof language proficiencycan be
togetherto affectFactor Two (Interaction 4). either effects
or causesof strategyuse (depend-
OtherInteractions. A significant interaction ing on which part of the chain is examined),
between university major and elective vs. and they are intimatelyrelated to motivation
requiredcourse statusappeared forFactor Four and self-esteem.20
(Interaction5). Years of studyand course status Our results show how strategy choice is

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
296 RebeccaOxford& Martha Nyikos
affectedby two variables we would expect to studentsthus seemed to take seriouslythe need
be related to motivation: 1) number of years to find extracurricular, communicatively-
of language study; 2) elective vs. required oriented practice opportunitiesin natural set-
course status. The students who studied the tingsand to guide theirown language studyin
same language fora minimum of fiveyears (a an autonomous, independentway, reflecting an
tinypercentageofthe sample), and the students awareness of metacognitive strategies. Cer-
who were taking the language electively(also tainlysuch findingsdisplay the language learn-
a minority)used significantlymore functional ing goal of developing communicative compe-
practice strategies(Factor Two) than did the tence. Goals reflectmotivational orientation,
rest of the students. Moreover, the students a recurringtheme in this study.
who studiedthe language forat least fouryears SexDifferences.
Sex differences in strategyuse,
used more Factor Five strategiesthan did stu- despite theirneglect in most previous strategy
dentswithless yearsof study.Perhaps students research,had a profoundinfluencehere.22For
who made less use of such communicatively example, compared with males, females re-
orientedlearning strategieswould not be likely ported significantlymore frequentuse of con-
to performwell and would drop out oflanguage versational input elicitationstrategies,reflect-
study upon completing the language require- ing social interaction.This resultcoincideswith
ment. On the otherhand, those who employed previousresearchon sex differences in women's
communicativelyorientedstrategiesmay have and men's speech in theirnative language. In
recognizedtheirvalue only in advanced classes, our culturethe two sexes actuallyuse theirown
resultingin continued study. native language differently, withgreaterdirect-
The pervasiveness and complexityof moti- ness, aggression,inputdiscouragement,power,
vational influencesare furtheremphasized by self-assuredness,and dominance-seeking dis-
the interactionsof motivationwith other vari- played in men's speech, and with greater
ables, such as elective vs. required course indirectness, politeness, input elicitation,
status, universitymajor, and sex. These sig- questioning,uncertainty,and harmony-seeking
nificantinteractionsaffectedseveral groups of reflectedin that of women.23One explanation
strategies. forthese differencesis known as the "strategy
Motivation is not just an internal, private model," which suggeststhatbecause of the un-
phenomenon generated by the individual stu- equal division of labor and power in American
dent. A student's motivation is affected by society, men and women use differentspeech
external variables (teaching and testingprac- strategiesto influencepeople and events. Men's
tices, peer interaction, overall task require- influenceis in the public sphere, and women's
ments, and the institutionalenvironment). In is in theprivate;theformersphereis considered
this study, students' language learning goals more important.In the two spheres, speech is
(embodied in the popularity of formal different,with differencesreflectingdominant
rule-learningstrategies)mirroredthe messages and subordinate positions: public speech is
of their instructional setting, which focused assertiveand direct;privatespeech is nurturing
mainly on developing analytic language skills and indirect.24 Of course, thismodel describes
and learning discrete language elements to strategies for native language use, but such
attain success on tests. strategieswould be reflectedin the way women
Like otherstudies,thisone
CareerOrientation. and men transferunconscious discourse strate-
showed that career orientation,reflectedhere gies to a new language. Our resultssuggestthis
in universitymajor, has a strongeffecton selec- to be the case.
tion of language learning strategies.21People In our study women used two additional
withdifferent career interestsseemed to choose typesof strategies- generalstudystrategiesand
differentstrategies. In this study, students formalrule-relatedpractice strategies- signifi-
majoring in humanities/socialscience/educa- cantly more often than men. This fact could
tion used two kindsof strategies-interestingly, be related to women's desire for good grades
the least popular kinds, functional practice and may reflecta need for social approval.25
strategies(Factor Two) and resourceful,inde- Women's greater use of these two kinds of
pendent strategies (Factor Three)--signifi- strategies might also echo their verbal supe-
cantlymore oftenthan did studentswithother riority.26 Another possible explanation is
majors. Humanities/social science/education women's greaterwillingnessthan men to con-

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
LanguageLearningStrategies 297

form to conventional norms.27 oriented strategies.30Students should be en-

social
Women's stronger,or at least different, couraged to experimentwitha great varietyof
orientation might have led us to expect that strategiesand to apply themto taskswhichpro-
theywould choose functionalpractice(authen- mote creative, communicative learning.
tic language use) strategiessignificantlymore Coupled with such an agenda must be a lan-
often than men.28 However, in this sample, guage programwhichtakes into account learn-
everyone's use of such strategiesappeared to ers' needs, including the need to gain self-
be suppressed by the traditional, academic control and autonomy through strategyuse.
environmentof the classroom- a settingwhich Subsequent research should explore all these
promotesand rewardsperformanceon discrete issues further.
tasks rather than interactive, communicative Such research can contributeto an impor-
efforts.29 tant and necessary transformation:changing
language learning classrooms into stimulating
CONCLUSION
places where use of communicatively-oriented
strategies- forboth learningand teaching-will
This study provides many insights about be commonplace. The change may lie not so
variables influencing the choice of learning much in the application of a given method or
strategiesby foreignlanguage studentsin a con- approach, or in the use of a giventextbook,but
ventional classroom setting.We have demon- in promotinga conscious awareness and use of
stratedthe powerfuleffectsof motivation,sex, workable strategieswithin the confinesof the
years of study, and other variables on choice foreignlanguage classroom. This change will
of language learning strategies. We have also not be easy; it will involve modifyingattitudes
suggestedhow the expectationsimposed by the and behaviors of learners and teachers alike.
standard academic approaches to teachingand Nevertheless,thistransformationis essential if
testinglimit the motivation of most language students are to obtain the greatest possible
learners to trynew, creative,communicatively benefit fromlanguage instruction.31

7Dickinson (15), Holec (22).

NOTES 8Anderson (1, 2), Bates (4), Carrell (10, 11), and Slobin
(58) provide background on development of mental struc-
tures or schemata.
1AlthoughReid's (49) studyof the language learningstyle 9Dansereau (12, 13), Derry & Murphy (14), O'Malley
preferences of ESL students contained a slightlygreater and colleagues (31, 32, 33), Oxford (39), Russo & Stewner-
number of subjects than ours (approximately 1,400 vs. Manzanares (56). The teaching of learning strategies is
1,200), neverthelessour studyappears to be the largestcom- sometimes known as "learnertraining,""strategytraining,"
pleted investigation of language learning strategiesas of or "learning-to-learntraining,"and it- along withthe iden-
spring 1988. Learning styles, as compared with learning tificationand diagnosis of learning strategies- changes and
strategies, are much more global and more difficultto expands the teacher's role; see Wenden (63), Oxford (38,
change. Style dimensions include (among many others) 39).
reflectivevs. impulsive; aural, visual, kinestheticvs. tac- 1oThese figuresdo not reflectthe actual enrollment by
tile; leveling vs. sharpening; thinkingvs. feeling; field de- language at the university. The sample was drawn from
pendence vs. fieldindependence. See Shipman & Shipman classes of teachers who volunteered to have their students
(57) for style dimensions. surveyed.
2For example, see (1, 2), Brown, Campione & Day (7), 11TheSILL was originallydeveloped forthe Army Re-
Brown & Palinscar (8), O'Neil (34), Weinstein, Goetz & search Instituteand the Defense Language Instituteforthe
Alexander (61). Language Skill Change Project, which studies variables
3Bialystok (5, 6), Naiman, Fr6hlich & Todesco (29), affectingchange (improvement, maintenance, or loss) of
Oxford (35, 36, 38, 39, 40), Reiss (51), Rubin (53, 54), language skills during and after language training. A
Rubin & Thompson (55), Wenden (62), Wenden & Rubin number of differentformsof the SILL (other than the one
(64). described in this article) have since been developed for
4politzer(45, 46, 47), Politzer & McGroarty (48), Oxford specific populations and uses, including both foreignlan-
(40). guages and ESL/EFL. The SILL is now being translated
5Oxford (39), Rubin (53, 54), Wenden & Rubin (64). into Spanish and Chinese.
6Hosenfeld (23, 24), Reiss (50). 12Asnoted, the main focus of this article is on the second

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
298 RebeccaOxford& Martha Nyikos
question, which concerns influenceson strategyuse. Com- opment, self-esteem,and motivation, see McCombs (28).
plete details regarding the firstquestion, including an in- 20In this study we looked at proficiencyself-perceptions
depth discussion of the factor analytic results, are found as independent (or causal) variables, but it is also certainly
in Nyikos & Oxford (30). possible to examine such perceptions as dependent variables
13SILL factoranalysis involved Promax (oblique) rota- (effectsof strategyuse).
tion with the maximum number of allowable factorsset at 21Politzer (45), Ehrman & Oxford (16).
10 and the minimum eigenvalue set at 1. This procedure 22See Oxford, Nyikos & Ehrman (44) for a review of
was the same used by Oxford (37) in the original SILL field studies of sex differencesin language learning strategyuse.
test studyat the Defense Language Institute; thus it is pos- 23Fishman (17, 18), Kramarae (25), Lakoff (26), and
sible to compare factors across studies (see Nyikos & Tannen (59) all discuss sex differencesin native language
Oxford). The general linear model (GLM) program was use.
used for conducting analyses of variance (ANOVA) on 24The strategymodel, developed by Britishand Ameri-
factorscores (i.e., the scores assigned to individuals on each can anthropologists, is summarized by Kramarae (25).
of the five strongest SILL factors found in the current 25The theme of sex differencesin social approval is elo-
study). The independent variables in the analyses of vari- quently explored by Bardwick (3).
ance were sex, elective vs. required course status, motiva- 26See, for example, Maccoby & Jacklin (27), Gage &
tion and proficiencyself-ratings, universitymajor, and other Berliner (19), Tyler (60) forinformationon sex differences
background variables, while the dependent variables were in verbal ability.
the factorscores. As noted earlier,the use ofANOVA allows 27Bardwick(3) discusses the controversial topic in some
directional statements about causality. detail.
14Fora SILL factoranalysis witha large militarysample, 28Maccoby & Jacklin (27) suggest that females are very
see Oxford (37), and Nyikos & Oxford (30). differentfrom males in many social skills, with females
15This level, p <.0001, means that probability of such showing greater gentleness, cooperation, and nurturance
resultsoccurringby random erroris less than one in 10,000. in social relationships and men displaying greater aggres-
The standard p-value indicating significance is usually sion, competition, and dominance in social relationships.
p <.05, the accepted level for this study. These social aspects are reflectedin speech. Other infor-
16The motivation question resulted in a somewhat mation on sex differencesin social orientation is found in
"flattenedout" normal curve, withthirty-onepercentof the Bardwick (3), Gilligan (21), and Tyler (60).
students expressing low motivation, forty-one percent 29Throughoutthis article the descriptions"conventional"
moderate motivation, and twenty-eightpercenthigh moti- and "traditional" are used to describe programs (such as
vation. The vast majority (82 %) also indicated they used the one under study) where teaching tends to be eclectic
the language less than ten percent of the time, reflecting and textbook-driven,with emphasis on covering practice
absence of strongmotivation (and perhaps of opportunity) exercises and lessons. Consequently, relativelylittle time
among the majority of subjects. is spenton promotingcommunicativeskills.Such situations
171ngeneral, studentsrated themselvesas more proficient are oftenprevalent where close adherence to the textbook
in reading and writingand less proficientin speaking and is used to ensure uniformityof content across classes at a
listening, as might be expected in a traditional, academic given level. Our results can be generalized to similar uni-
language situation. Positive (excellent or good) proficiency versitylanguage programsto the degree that such an orien-
self-ratingswere given by forty-onepercentof the students tation underlies classroom instruction.
in listening,sixtypercent of the studentsin writing,thirty- The designation of "conventional"in the sense described
one percent in speaking, and forty-four percent in writing. above is based on the second author's familiaritywith the
Looking at it the other way around, negative (fair or poor) teaching and testing techniques at several large universi-
proficiencyself-ratingswere given by fifty-nine percent of ties, includingthe one in thisstudy,as well as at her present
the students in listening, fortypercent in reading, sixty- institution. Conventional language programs are typified
nine percent in speaking, and fifty-six percent in writing. by much individual variation in the degree of communica-
Only two to eight percent of the students, depending on tive teaching practices employed.
the skill, described themselves as excellent in any of the 30Yet, paradoxically, even academic programs that
fourlanguage skills. Learners rated theirproficiencycom- promote high functional language use tend to revert to
pared with that of other universityforeignlanguage stu- discrete-pointtestswhich, studentsrealize, set the real cri-
dents. Obviously, much lower proficiencyself-ratings would teria forjudging theiracademic success. Such testingprac-
be expected if learners compared themselves with native tices underminethe potentialuse and transferof interactive,
users of the foreign languages. communicative strategies to realistic, functional settings
18See Oxford, Nyikos & Crookall (43) fordetails and a beyond the classroom.
complete interpretationof the interactions. 3"Originally a paper presented at the ModernLanguage
19Inthiscase, we are talkingabout situationalself-esteem, Journal/OhioState UniversitySymposium on Research Per-
which is not the same as global self-esteem;see Brown (9). spectives in Adult Second Language Learning and Acqui-
For furtherinformationabout linkages among skill devel- sition, Columbus, 22 October 1988.

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
LanguageLearningStrategies 299
21. Gilligan, Carol. In a Different Voice:PsychologicalTheory
BIBLIOGRAPHY and Women's Development. Cambridge, MA: Harvard
Univ. Press, 1982.
22. Holec, Henri. Autonomy and ForeignLanguageLearning.
1. Anderson, J. R. CognitivePsychology and Its Implications. Oxford: Pergamon, 1981.
San Francisco: Freeman, 1980. 23. Hosenfeld, Carol. "Cindy: A Learner in Today's For-
2. - . CognitiveSkills and TheirAcquisition.Hillsdale, eign Language Classroom. The ForeignLanguage
NJ: Erlbaum, 1981. Learnerin Today'sClassroomEnvironment. Ed. Warren
3. Bardwick, Judith M. Psychology of Women:A StudyofBio- Born. Middlebury, VT: Northeast Conference,
culturalConflicts.New York: Harper & Row, 1971. 1979.
4. Bates, Elizabeth. Languageand Context.New York: Aca- 24. -. "A Learning-Teaching View of Second Lan-
demic, 1972. guage Instruction."ForeignLanguageAnnals12(1979):
5. Bialystok, Ellen. "The Role of Conscious Strategies in 51-57.
Second Language Proficiency." Modern Language 25. Kramarae, Cheris. Womenand Men Speaking.Rowley,
Journal,65 (1981): 24-35. MA: Newbury House, 1981.
6. & Maria Fr6hlich. "Variables of Classroom 26. Lakoff, Robin. Languageand Woman'sPlace. New York:
Achievement in Second Language Learning." Harper & Row, 1975.
Modern LanguageJournal62 (1978): 327-66. 27. Maccoby, Eleanor E. & Carol Jacklin. The Psychology
7. Brown, A. L., J. C. Campione & J. D. Day. "Learning of Sex Differences.Stanford, CA: Stanford Univ.
to Learn: On Training Students to Learn from Press, 1974.
Tests." Unpubl. Paper, AERA Annual Meeting, 28. McCombs, Barbara L. "The Role of AffectiveVariables
Boston, 1980. in Autonomous Learning." Unpubl. Paper, AERA
8. - & A. S. Palinscar. "Inducing Strategic Learn- Annual Meeting, New Orleans, 1984.
ing fromTexts by Means of Informed Self-Control 29. Naiman, Neil, Maria Fr6hlich & Angie Todesco. "The
Training." Topics in Learningand LearningDisabili- Good Second Language Learner." TESL Talk
ties 2 (1982): 1-17. 5(1975): 58-75.
9. Brown, H. Douglas. PrinciplesofLanguageLearningand 30. Nyikos, Martha & Rebecca Oxford. "Language Learn-
Teaching.Englewood Cliffs,NJ: Prentice-Hall, 1987. ing Strategy Use: Factor Analytic Comparisons of
10. Carrell, Patricia. "Evidence of a Formal Schema in Sec- Military and UniversitySamples." Unpubl. (1988).
ond Language Comprehension." LanguageLearning 31. O'Malley, J. Michael, Rocco P. Russo & Anna U.
34(1984): 87-112. Chamot. A ReviewoftheLiterature on LearningStrate-
11. - . "Schema Theory and ESL Reading: Classroom gies in theAcquisitionofEnglishas a SecondLanguage:
Implications and Applications." Modern Language The Potential for ResearchApplications.Rosslyn, VA:
Journal 68(1984): 332-43. InterAmerica Research Associates, 1983.
12. Dansereau, Donald F. "The Development of a Learn- 32. - , Gloria Stewner-Manzanares & Lisa Kupper.
ing Strategies Curriculum." LearningStrategies.Ed. LearningStrategies UsedbyHigh SchoolStudents in Learn-
H. F. O'Neil, Jr. New York: Academic, 1978: 1-29. ingEnglishas a SecondLanguage.Rosslyn, VA: Inter-
13. - . "Learning Strategy Research." Thinkingand America Research Associates, 1983.
LearningSkills:RelatingLearningtoBasic Research.Ed. 33. - , Rocco P. Russo, Anna Chamot, Gloria
J. W. Segal, Susan F. Chipman & Robert Glaser. Stewner-Manzanares & Lisa Kupper. "Learning
Hillsdale, NJ: Erlbaum, 1985: 209-40. Strategy Applications of Students of English as a
14. Derry, Sharon J. & Debra A. Murphy. "Designing Sys- Second Language." TESOL Quarterly19(1985):
tems that Train Learning Ability: From Theory to 557-84.
Practice." Review of EducationalResearch43(1986): 34. O'Neil, Harold F., Jr., Ed. LearningStrategies.
New York:
455-67. Academic, 1978.
15. Dickinson, Leslie. Self-Instruction in LanguageLearning. 35. Oxford, Rebecca. A New TaxonomyofSecondLanguage
Cambridge: Cambridge Univ. Press, 1987. LearningStrategies.Washington: ERIC, 1985.
16. Ehrman, Madeline E. & Rebecca L. Oxford. "Effects 36. - . "Second Language Learning Strategies: What
of Sex Differences,Career Choice, and Psychologi- the Research has to Say." ERIC/CLL New Bulletin
cal Type on Adults' Language Learning Strategies." 9(1985): 1, 3-5.
Modern LanguageJournal 73(1989): 1-13. 37. - . Development and PsychometricTestingoftheStrategy
17. Fishman, P. M. "Interactional Shitwork." Heresies:A Inventory forLanguageLearning.Army Research Insti-
Feminist onArtand Politics2(1977): 99-101.
Publication tute Technical Report 728. Alexandria, VA: US
18. - . "Interaction: The Work Women Do." Social ArmyResearch InstituteforBehavioral & Social Sci-
Problems25(1978): 397-406. ences, 1986.
19. Gage, N. L. & David C. Berliner. EducationalPsychol- 38. - . SecondLanguageLearningStrategies:CurrentRe-
ogy. Chicago: Rand McNally, 1975. searchand ImplicationsforPractice.Los Angeles: Center
20. Gardner, Robert C. SocialPsychology
and SecondLanguage for Language Education and Research, Univ. of
Learning:TheRole ofAttitudes
andMotivation.London: California, 1986.
Arnold, 1985. 39. - . LanguageLearningStrategies.: WhatEveryTeacher

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions
300 RebeccaOxford& MarthaNyikos
ShouldKnow. New York: Newbury House/Harper 52. Rigney, J. W. "Learning Strategies:A Theoretical Per-
& Row, forthcoming. Ed. Harold F. O'Neil,
spective." LearningStrategies.
40. - "Use of Language Learning Strategies:A Syn- Jr. New York: Academic, 1978: 165-205.
thesis of Studies with Implications for Strategy 53. Rubin, Joan. "What the 'Good Language Learner' Can
Training." System17 (1989): 1-13. Teach Us." TESOL Quarterly 9 (1975): 41-45.
41. . "Lessons fromResearch on Language Learn- 54. - . "Learner Strategies: Theoretical Assumptions,
ing Strategies." (Submitted.) Research, History, and Typology." LearnerStrategies
42. - & David Crookall. "Learning Strategies." You in Language Learning. Ed. Anita Wenden & Joan
Can Take it With You.:HelpingStudents Maintain Sec- Rubin. Englewood Cliffs,NJ: Prentice-Hall, 1987:
ondLanguageSkills. Ed. Jean Berko-Gleason. Engle- 15-30.
wood Cliffs, NJ: Prentice-Hall, 1988. 55. - & Irene Thompson. How to be a More Successful
43. - , Martha Nyikos & David Crookall. "Learning LanguageLearner.Boston: Heinle & Heinle, 1982.
Strategiesof UniversityForeign Language Students: 56. Russo, Rocco P. & Gloria Stewner-Manzanares. "The
A Large-Scale Study." Unpubl. Paper, TESOL Training and Use of Learning StrategiesforEnglish
Annual Meeting, Miami, 1987. as a Second Language in a Military Context."
44. - , Martha Nyikos & Madeline E. Ehrman. "Vive Unpubl. Paper, AERA Annual Meeting, Chicago,
la Diff6rence?Reflectionson Sex Differencesin Use 1985.
of Language Learning Strategies." ForeignLanguage 57. Shipman, Stephanie & Virginia C. Shipman. "Cogni-
Annals 21(1988): 321-29. tive Styles: Some Conceptual, Methodological, and
45. Politzer, Robert L. "Research Notes: An Exploratory Applied Issues." ReviewofResearchin Education.Ed.
Study of Self-Reported Language Learning Be- E. Gordon. Washington: AERA, 1985.
haviors and Their Relation to Achievement."Studies 58. Slobin, Dan I. "Developmental Psycholinguistics."A
in SecondLanguageAcquisition6 (1983): 54-68. Surveyof LinguisticScience. Ed. W. O. Dingwall.
46. - . "Linguistic and Communicative Competence College Park, Md.: Univ. of Maryland Linguistics
Achieved in Relation to Reported Motivation and Program, 1971.
Learning Behaviors in an Intensive ESL Course." 59. Tannen, Deborah. That'sNot WhatIMeant! New York:
Unpubl. Paper, Stanford Univ. (undated). Morrow, 1986.
47. -. "Motivation, Learning Behavior, and Achieve- 60. Tyler, Leona. ThePsychology ofHuman Differences. New
ment in an Intensive ESL Course." Unpubl. Paper, York: Appleton-Century-Crofts, 1965.
Stanford Univ. (undated). 61. Weinstein, C. E., E. T. Goetz & P. A. Alexander, Eds.
48. - & Mary McGroarty. "An Exploratory Study Learning and Study Strategies.:Issues in Assessment,
of Learning Behaviors and Their Relation to Gains Instruction, and Evaluation. New York: Academic,
in Linguistic and Communicative Competence." 1988.
Unpubl. Paper, Stanford Univ., 1983. 62. Wenden, Anita. "Learner Strategies." TESOL News-
49. Reid, Joy M. "The Learning Style Preferencesof ESL letter19 (1985): 1-7.
Students." TESOL Quarterly 21 (1987): 87-111. 63. - . "IncorporatingLearner Training in the Class-
50. Reiss, Mary-Ann. "Helping the UnsuccessfulLanguage room." System14 (1986): 315-25.
Learner." Forum 21, ii (1983): 2-24. 64. - &Joan Rubin, Eds. LearnerStrategies in Language
51. - . "The Good Language Learner: AnotherLook." Learning.Englewood Cliffs,NJ: Prentice-Hall, 1987.
CanadianModernLanguageReview41 (1985): 511-23.

I HI Research Perspectivesin
Adult Language Learning and Acquisition
ISTATE:
[:NIVFRW)'

November3-4, 1989, Columbus, Ohio

Keynote addresses by:
Vicki Galloway, Editor, ForeignLanguageAnnals (ACTFL)
Earl Stevick, U.S. Foreign Service Institute (ret.)
Renate Schulz, University of Arizona
This annual conferenceis organized by the Foreign Language Center of The Ohio State University,
and is co-sponsored by The Modern Language Journal and the OSU College of Humanities.
Preregistration desirable. Contact the OSU Foreign Language Center for details.
RP-ALLA '89
OSU Foreign Language Center
155 Cunz Hall, 1841 Millikin Road
Columbus, OH 43210-1215

This content downloaded from 132.234.251.230 on Thu, 16 Jan 2014 10:20:52 AM

All use subject to JSTOR Terms and Conditions