Banner-volume - Copy copy.png 

Annual Review of Education, Communication and Language Sciences, Volume 1, 2004

 

THE EFFECTS OF AGE, SEX, AND LANGUAGE PROFICIENCY ON THE SELF-EFFICACY OF ENGLISH LANGUAGE LEARNERS

 

CHRISTOPHER J. JENKS

 

 

Abstract

 

The current paper investigated the self-efficacy of English language learners from different age, sex, and language proficiency backgrounds.  Self-efficacy, or the self-perception of succeeding in a given task, is believed to be a critical element in creating the necessary attitudes for successful learning outcomes (Bandura 1986, 1991).  Second language acquisition studies examining this psychological construct have additionally maintained that self-efficacy can shape the learning strategies employed by language learners (Yang 1999).  As a point of departure, this investigation will juxtapose the self-efficacy construct with three demographic domains; consequently, the statistical analysis that follows will only test associations through chi-square procedures.  These results indicate that there is a significant association between self-efficacy and language proficiency, but not for age and sex.      

 

 

Introduction

    The inherent complexities occupied within the process of learning an additional language are symbolized in the divergent yet sometimes conflicting disciplinary fields of linguistics, education, sociology, psychology, philosophy, and neurology, just to name a few.  It is invariably no surprise that linguists concerned with second language acquisition (SLA) in classroom settings have borrowed a multitude of theoretical and empirical models from various behavioral and psychological fields (see Larsen-Freeman & Long 1991, for an introductory account of SLA research models).  Yet, the common analytical SLA goal, however dissimilar their particular empirical objectives are (e.g., longitudinal, cross-sectional, children, adults, or classroom settings), is to unravel the individual (e.g., innate abilities) and social characteristics (e.g., teaching methods) that facilitate efficient and effective language acquisition (Doughty & Long 2003).  This pervasive and prominent investigatory conundrum is part and parcel of the individual (although in many ways, social) psychological studies investigated within the disciplinary interchange of affective variables (i.e., emotions) and language acquisition (Horwitz, Horwitz, & Cope 1986; Pappamihiel 2002; Hackett 1995; Dornyei & Kormos 2000; Gardner, Smythe, Clement, & Glicksman 1976; Ehrman, Leaver, & Oxford 2003).  The present investigation will follow this psycholinguistic tradition and set forth three fundamental analytical points of interest.

 

1.  What is the effect of age on the self-efficacy of English language learners?

 

2.  What is the effect of sex on the self-efficacy of English language learners?

 

3.  What is the effect of language proficiency on the self-efficacy of English language learners?

 

   As a result, the subsequent section will introduce the requisite literature within the stratum of psycholinguistics, an explication of the methodology in which the data was collected will follow, and a detailed examination and discussion of the data will conclude.

 

 

Self-efficacy

   Before establishing a link between self-efficacy and language acquisition, it would be sensible to establish how self-efficacy is used in the present investigation.  Self-efficacy is a language learner’s perception of how well she will do in a particular academic or linguistic task (not to be confused with self-esteem, which is related to perceptions of self-worth).  For example, if Student A is required to complete task Z, and a scale from 1-10 corresponds to the student’s perception of how well she will do in that task, the pre-task selected number would accordingly represent Student A’s self-efficacy level in relation to task Z.  In short, self-efficacy embodies the confidence to complete any particular endeavor.

    The notion of self-efficacy and its role in child psychological development has been forwarded and theorized in the seminal works of Albert Bandura (1986, 1991).  Bandura (1986) suggests that the ability to successfully complete a pedagogically oriented task is related to a person’s affective (or emotional) make-up; in particular, the self-efficacy (or confidence) to productively complete an academic objective.  Within the contexts of learning a language, the ability to accomplish and achieve is conclusively linked with existing linguistic capacities (Gass & Selinker 2001), but these capacities are unquestionably confined within an affective framework (Bandura 1991).  This affective framework can be seen as an interpretive lens through which language learners “…undertake and perform confidently activities that they judge themselves capable of managing, but they avoid those they believe exceed their ability.  These judgments also help people to determine how much effort they will spend and how long they will persist in the face of obstacles or difficulties” (Yang 1999, p. 517).  With this in mind, what are the practical, classroom implications of these theoretical proposals?

 

 

Self-efficacy & language success

    If language learners are inclined to accomplish classroom tasks according to relative levels of self-efficacy, classroom performance (i.e., grades, peer assessments, and language development) would ostensibly appear to be partially contingent on the perceptual confidence of language learners.  Horwitz et al. (1986) has developed an investigatory model to establish an empirical link between affective characteristics and language success.  Although the Foreign Language Classroom Anxiety Scale (FLCAS) was not implemented to specifically examine self-efficacy, the lack of confidence inherent in feelings of anxiety nevertheless provides ample justification to highlight their linguistic effects (for instance, FLCAS questions such as,  ‘I start to panic when I have to speak without preparation in language class’, relates to the self-perception or confidence of performing well in a given academic situation; see Horwitz et al. 1986, p. 129).  In fact, Ehrman et al. (2003) state, “highly motivated, successful learners…possess self-efficacy” (p. 321) and “L2 performance anxiety…is often highly related to motivation.” (p. 323).  In the Horwitz et al. (1986) study, the difficulties in verifying this affective variable-foreign language achievement link was acknowledged, but the self-perception reports did indicate that students with high (or higher) levels of anxiety would be less inclined to perform particular academic tasks; in effect, parallel with how language learners of various degrees of self-efficacy select only the tasks that they feel most confident in completing (see Yang 1999; Allwright, & Bailey 1991).  What, then, are the developmental effects of predisposed beliefs vis-à-vis the selection and successful completion of difficult academic tasks?

    By applying Bandura’s (1986) theoretical framework of social-psychology, Hackett (1995) determined that the self-efficacies of students are in fact a significant source of variability in language attainment (approximately 300 participants demonstrated that their success or failure in linguistic tasks were influenced by their responses on a self-efficacy measuring model).  On a smaller scale, Pappamihiel (2002) revealed that language learners’ willingness to communicate was associated with their respective self-efficacy levels; in this case, the nonnative language learners of English were less inclined to communicate when they were put into classrooms with their native English-speaking counterparts.  Again, this fastidious selectivity underscores the pedagogical effects of anxiety (and self-efficacy), and the subsequent successful completion (or attempt) of all required academic tasks (Horwitz 1991).  Failure to complete or undertake compulsory tasks, as illuminated in Gardner et al. (1976), may result in diminishing oral development, and a propensity to harbor negative perceptions of language success.
     In sum, two notions of self-efficacy have been established.  The first notion set forth attempts to bridge between anxiety (its association with self-efficacy) and language success; this is reasonably the weak(er) argument.  The second notion provides a stronger argument for self-efficacy by specifically amalgamating language success with self-efficacy.  In order to avoid the complexities and ambiguities involved in establishing an anxiety-self-efficacy link, only the stronger argument will be applied to any subsequent claims regarding language success.  In both cases, however, the pedagogical implications of accounting for affective influences are a common ground (Dornyei, & Kormos 2000).  That is to say, language learners possess distinct characteristics that must be socially and academically adaptable to a larger community (e.g., the language classroom).  The subsequent section will briefly introduce some of these learner characteristics, and demonstrate how individual differences shape pedagogical strategies.

 

 

Affective & individual factors

    The former explications regarding self-efficacy characterize one form of individual learner variation (other, more common cerebral factors include inductive-deductive thinking, aptitudes, and multiple intelligences).  Notwithstanding the importance of other, cognitively situated language learner differences, the present investigation will situate its analytical focus on the demographics of age, sex, and proficiency level.  The question of age and its affect on language proficiency has historically been an issue of great debate (see Hyltenstam & Abrahamsson 2003, for a comprehensive account of maturational constraints).  On a less controversial issue, Oxford and Ehrman (1995) have asserted that the type of strategies employed in language learning episodes is an age-associated process (e.g., older learners are likely to think more abstractly and grammatically, while younger learners often exploit the communicative strategies essential in daily interactions).  Sex, on the other hand, has seen less empirical attention or corpus data; nevertheless, the female-male dichotomy has resulted in studies indicating that females make more use of strategic practices, including more robust listening capacities (Larsen-Freeman & Long 1991).  So, if age and sex can shape learning processes, what role does proficiency levels have on the task-selection strategies of self-efficacy (recall the avoidance strategies established in the previous section)?  In other words, is it the success of achieving higher proficiency levels that promote self-efficacy, or is it the confidence inherent in self-efficacy that promotes the willingness to achieve high levels of proficiency?  As part of its analytical program, the current investigation will detail the perceptual self-efficacies of language learners, and its dependencies, if any, on proficiency levels.  However, it is important to note that this associative process will only furnish a rough, context-free answer to the self-efficacy and proficiency level question.  The fundamental objective is to simply see if highly proficient language learners possess higher levels of self-efficacy than their lower proficient language-learning counterparts (the age and sex variable will be implemented and disseminated in a similar analytical fashion).  The following methodological section will describe the framework used to examine these investigatory concerns.

 

 

Methodology

 

Participants

    A process of random stratification was used to select the schools in which the participants studied English as a Second Language (ESL).  The participants’ consent was established at the discretion of each participating institution (3 public schools in the County of Fairfax, Virginia).  The self-efficacy corpus encompassed 133 ESL students between the ages of 11 and 20 (M = 14.63, SD = 2.05).  Table 1 presents the remaining investigated demographics.  Students’ proficiency levels were grouped according to the language proficiency guidelines set forth by the Fairfax County (Fairfax ESOL, 2004).   

                       

 

Self-efficacy test

    The self-efficacy scores were collected by administering a 5-point Likert scale questionnaire (from strongly agree to strongly disagree).  The validity of the self-efficacy measuring model, which was initially conceptualized in Hackett’s (1995) work, was validated by eight external specialists; this validating process is necessary when analyses are centered on self-reporting constructs (Pedhazur & Schmelkin 1991).  However, the application of self-reporting questionnaires may weaken the control of internal validity (Larsen-Freeman & Long 1991), and offer respondents the opportunity to answer with veiled objectives (Oller & Perkins 1978).  Nevertheless, self-reporting protocols give students a chance to share their feelings and encourage meta-behaviors (Neito 1994).

    To further strengthen the management of internal validity, the self-efficacy questionnaire was translated from English to Korean and Spanish to reflect the linguistic majority (81 ESL students spoke Korean or Spanish as their native language).  By providing more time and English language support, additional comprehension was facilitated to those ESL students whose native language did not benefit from the translations.  Accordingly, each participating ESL departments were instructed to omit any self-efficacy questionnaires that were too difficult for the students to comprehend (only 7, or 5% of the respondents’ answers were omitted).  All of the self-efficacy respondents were randomly selected by their respective ESL instructors.  The questionnaire was conducted during classroom hours, and with the help of research staff.

 

Statistical procedures

    In the interest of space and simplicity, only the self-efficacy questions relating to the four language skills (writing, listening, reading, and speaking) were combined to form one overall self-efficacy language score representing three self-efficacy subdivisions (i.e., low self-efficacy [score of 1-2], medium self-efficacy [score of 3], and high self-efficacy [score of 4-5]).  Cross-tabulations with Pearson chi-square computations were utilized for each research question.  Significant levels were set at the .05 level.  The subsequent section will illustrate these results.

 

 

Self-efficacy results

 

Age & self-efficacy

    The first analytical focus is concerned with the relationship between age and self-efficacy.  Although Table 2 shows that a large number of students between the ages of 13 and 15 (n = 44) reported high levels of self-efficacy, the chi-square result was not statistically significant (p = .179).

 

    Consequently, the participants’ age in this investigation did not play a significant effect in the reported self-efficacy levels (this may be due to the fact that the age scale was too broad for any considerable trends to occur).  It should be noted, however, that none of the 19- and 20-year old respondents were beginner or intermediate-leveled students (whereas none of the 11- and 12-year olds were advance-leveled students).  With this in mind (i.e., the possibility of a correlation between age and proficiency levels), if proficiency levels have any role in self-efficacy (see the last analytical section), then the statistically insignificant results in this section may have resulted from the small or unevenly distributed sample size (N.B., most of the respondents were between the ages of 13 and 16).  In regard to this uneven distribution, the students were not grouped into ‘older’ and ‘younger’ subgroups because further ambiguities may emerge vis-à-vis differentiating between what ages may be young or old.

 

Sex & self-efficacy

    The second analytical point deals with the correspondence of sex and self-efficacy.  In this comparison, no statistically significant chi-square relationship was borne out (p = .971).  Table 3 illustrates the descriptive results.

    The near even distribution of male and female students in each self-efficacy category supports the statistically insignificant result (note the near even distribution in each column of Table 3).  However, a larger sample size is clearly needed before any conclusive claims can be made regarding the effects of sex on self-efficacy levels.

 

Proficiency & self-efficacy

    The last analytical focus set forth the assumed association between English language proficiency levels and self-efficacy.  Table 4 catalogues the composite results.

    Despite the similar aggregate self-efficacy scores amongst beginner and advance-leveled respondents (33 and 29, respectively), low self-efficacy scores corresponded more to beginner-leveled students (see the low self-efficacy column in Table 4), while advance-leveled students almost doubled their beginner-leveled counterparts in the high self-efficacy category (21:12).  These results were in fact statistically significant at the p = .028 level (again, Pearson chi-square calculations).  Working within this association, the tendency for advance-leveled students to possess high self-efficacy would suggest that students’ affective dispositions are more associated with their respective proficiency levels than their age or sex.  Yet, because there was a significant association between proficiency levels and self-efficacy, the previous age and self-efficacy association needs to be reinvestigated using a more even (and larger) distribution of students (correlations between age and proficiency levels may also help disambiguate this issue).

Implications

    As noted in the beginning of this paper, language learners may only select those academic tasks that are in line with their perceptual confidence or self-efficacy (Yang 1999).  Moreover, the potential linguistic and academic repercussions from avoiding demanding but stimulating tasks are hazardous to the overall development of language learners (Horwitz et al. 1986; Horwitz 1991; Ehrman et al. 2003; Hackett 1995; Pappamihiel 2002; Gardner et al. 1976).  Although these studies conclusively and prominently position affective variables within the successful process of learning an additional language, little attempt has been made to account for the individual factors that may affect self-efficacy levels.  For instance, most research that has attempted to account for fluctuating affective variables were primarily concerned with teaching strategies (see Foss & Reitzel 1988, for strategies educators can use to promote favorable language attitudes).

    Hansen’s (2001) investigation, however, explored writing self-efficacy differences of native English speaking male and female students.  Although the male students possessed less motivation to engage in writing activities, there was no significant female-male difference for writing self-efficacies.  These results mirror the insignificant results attained in the sex and self-efficacy section of the current investigation.  Notwithstanding the methodological differences (i.e., while Hansen’s 2001 study was occupied with native speakers of English and their respective writing self-efficacies, the current analysis juxtaposed the overall language self-efficacies with nonnative English speakers), both studies suggest that sex differences may have less influence on self-efficacies than the different female-male learning strategies identified by Larsen-Freeman and Long (1991).  That is to say, confidence in academic and linguistic achievements may be more contingent on learning strategies (or other unidentified individual or social factors) than the biological differences between male and female.

    The age of language learners played an equally negligible part in the self-efficacy results.  As previously noted, the long-standing discussions regarding maturational periods and language acquisition are still a ubiquitous debate.  Similar to the ambiguities involved in the issue of sex and language learners, a large number of demographic variability research within the age debate have centered on the dissemination of strategies used by students (Oxford and Ehrman 1995).  That is, investigations concerning the affective practices (as opposed to learning practices) of language learners have seen little empirical attention.  However, MacIntyre, Baker, Clement, and Donovan (2002) have found female students less willing to communicate as they advance grade levels, but this linguistic apprehension was due to language anxiety; as formerly clarified, the anxiety-self-efficacy connection is a vulnerable argument, and therefore introduced only as a point of reference.  The results of the survey did indicate, however, that the variability of self-efficacy in this study had less to do with age and sex, and more to do with proficiency levels.

    A question was introduced earlier in this paper pertaining to the prominence of self-efficacy and proficiency levels.  That is, are the positive attitudes associated with self-efficacy, or is the success associated with gaining proficiency a prevailing catalyst for language acquisition?  Though the answer to the ‘what comes first’ question may prove to be an elusive endeavor, proficiency levels did have a salient function in the self-efficacy experienced by the investigated participants.  Although this significant association cannot be unequivocally interpreted as a ‘what came first’ answer, the data highlights the need for educators to allocate some instructional time to promote self-efficacy for beginner-leveled language learners (half of the low self-efficacy level respondents were beginner-leveled students; see Table 4); these students are continually facing new and challenging linguistic tasks that justifiably promote feelings of anxiety, hesitation, and trepidation (Matsuda & Gobel 2004).  In order to become a successful language learner, the motivation and confidence needed to overcome such affective barriers must be identified and developed by students and teachers alike (Gardner & MacIntyre, 1993; Horwitz et al. 1986).  Thus, part of what is needed to provide efficient and effective learning and teaching processes (i.e., the common SLA goal; see Doughty & Long 2003) is the awareness, concern, and promotion of language success.   

 

 

Conclusion

    So, what are the characteristics of a successful language learner?  As was highlighted in the preceding sections, self-efficacy (or the confidence to complete any given task) was an attribute more common in advance-leveled students.  That is not to say language learners cannot succeed if they posses lower levels of self-efficacy.  To the contrary, the characteristics of a successful language learner encompass many individual traits including the motivation to integrate into a new community, the responsibility and maturity to apply good study habits, an overall favorable attitude towards a target language, and both introverted and extroverted personalities (Gardner 1985).  Needless to say, the list can go on; yet the objective of the current analytical foci was to simply juxtapose three demographic variables (age, sex, and proficiency levels) with one successful language learner characteristic (i.e., self-efficacy).  The distribution of self-efficacy was (to a degree not explored in this study) contingent on the proficiency levels of language learners.  In a limited attempt to provide an answer to the ‘what came first’ question, Gardner (1985) states that language attitudes are independent of aptitude and intelligence, but “the attitude measures relate to achievement because of important affective components and not simply because the attitudes covary with ability” (p. 45).  The positive language attitudes (e.g., motivation or confidence) invariably needed to become a successful language learner are therefore achievable if “important affective components” (e.g., self-efficacy) are identified and supported.  In other words, it is the attitudes and affective components that help create the willingness to success in language development.  This study has selected self-efficacy as an analytical objective, identified three individual language learner characteristics (age, sex, proficiency level), and discovered that proficiency levels are a more reliable indicator for self-efficacy.  Future studies may wish to seek out any potential correlations between age and proficiency levels.  In any case, affective variables in language development are an area of investigation that should not be neglected.

 

 

About the author

Christopher Jenks has taught EFL in Seoul, South Korea and ESL in Northern Virginia, USA.  He possesses a M.Ed. in TESOL from George Mason University and is currently a PhD student of Educational Linguistics at the University of Newcastle upon Tyne.  Correspondence: cjjenks@hotmail.com

 

 

References

ALLWRIGHT, D., AND BAILEY, K.M., 1991.  Focus on the language classroom.  Cambridge, UK: Cambridge University Press.

 

BANDURA, A., 1986.  Social foundations of thought and action: A social cognitive theory.  Englewood Cliffs, NJ: Prentice-Hall.

 

BANDURA, A., 1991.  Self-efficacy conception of anxiety.  In: R. SCHWARZER AND R.A. WICKLUND, eds.  Anxiety and self-focused attention.  Chur, Switzerland: Harwood Academic Publishers, 89-110.

 

DORNYEI, Z., AND KORMOS, J., 2000.  The role of individual and social variables in oral task performance.  Language teaching research, 4 (3), 275-300.

 

DOUGHTY, C.J., AND LONG, M.H., 2003.  The scope of inquiry and goals of SLA.  In: C.J. DOUGHTY AND M.H. LONG, eds.  The handbook of second language acquisition.  Oxford: Blackwell Publishing, 3-16.

 

EHRMAN, M.E., LEAVER, B.L., AND OXFORD, R.L., 2003.  A brief overview of individual differences in second language learning.  System, 31 (3), 313-330.

 

FAIRFAX ESOL, 2004.  Office of English for Speakers of Other Languages.  Available from: http://www.fcps.edu/DIS/OESOL/index.htm [Accessed 17 January 2004].

 

FOSS, K.A., AND REITZEL, A.C., 1988.  A relational model for managing second language anxiety.  TESOL quarterly, 22 (3), 437-454.

 

GARDNER, R.C., 1985.  Social psychology and second language learning: The role of attitudes and motivation.  London: Edward Arnold.

 

GARDNER, R.C., AND MACINTYRE, P.D., 1993.  A student’s contributions to second language learning. Part II: Affective variables. Language teaching, 26 (1), 1–11.

 

GARDNER, R.C., SMYTHE, P.C., CLEMENT, R., AND GLICKSMAN, L., 1976.  Second language learning: A social-psychological perspective.  Canadian modern language journal, 32, 198-213.

 

GASS, S.M., AND SELINKER, L., 2001.  Second language acquisition: An introductory course.  Mahwah, NJ: Lawrence Erlbaum Associates.

 

HACKETT, S.L., 1995.  Toward a causal model of affect, cognition, and achievement in English as a second language. Dissertation (PhD). The University of Connecticut.

 

HANSEN, S., 2001.  Boys and writing: Reluctance? Reticence? Or Rebellion?  In: 8th annual meeting of the education research network, July 4-8, 2001.  Washington, DC: (ERIC Document Reproduction Service No., ED 456449).

 

HORWITZ, E.K., 1991.  Language anxiety: From theory and research to classroom implications.  Englewood Cliffs, NJ: Prentice Hall.

 

HORWITZ, M.B., HORWITZ, E.K., & COPE, J.A., 1986.  Foreign language classroom anxiety.  The modern language journal, 70 (2), 125-132.

 

HYLTENSTAM, K., AND ABRAHAMSSON, N., 2003.  Maturational constraints in SLA.  In: C.J. DOUGHTY AND M.H. LONG, eds.  The handbook of second language acquisition.  Oxford: Blackwell Publishing, 539-588.

 

LARSEN-FREEMAN, D., AND LONG, M.H., 1991.  An introduction to second language acquisition research.  Essex, UK: Longman Group.

 

MATSUDA, S., AND GOBEL, P., 2004.  Anxiety and predictors of performance in the foreign language classroom.  System, 32 (1), 21-36.

 

MACINTYRE, P.D., BAKER, S.C., CLEMENT, R., AND DONOVAN, L.A., 2002.  Sex and age effects on willingness to communicate, anxiety, perceived competence, and L2 motivation among junior high school French immersion students.  Language learning, 52 (3), 537-564.

 

NIETO, S., 1994.  Lessons from students on creating a chance to dream.  Harvard educational review, 64 (4), 392-426.

 

OLLER, J.W., AND PERKINS, K., 1978.  Language proficiency as a source of variance in self-reported affective variables.  In: J.W. OLLER AND K. PERKINS, eds.  Language in education: Testing the test.  Rowley, MA: Newbury House Publishers, 103-127.

 

OXFORD, R.L., AND EHRMAN, M.E., 1995.  Adults’ language learning strategies in an intensive foreign language program in the United States.  System, 23 (3), 359-386.

 

PAPPAMIHIEL, E.N., 2002.  English as a second language students and English language anxiety: Issues in the mainstream classroom.  Research in the teaching of English, 36 (3), 327-355.

 

PEDHAZUR, E.J., AND SCHMELKIN, L.P., 1991.  Measurement, design, and analysis: An integrated approach.  Hillsdale, NJ: Lawrence Erlbaum Associates.           

 

YANG, N., 1999.  The relationship between EFL learners’ beliefs and learning strategy use.  System, 27 (4), 515-535.