Demographic Variables and English Proficiency of Adult Language Learners: A Correlational Study

This study examined the influence of demographic variables on adult language learners’ English proficiency development. Specifically, it tried to ascertain the level of English proficiency of the respondents when grouped according to demographic variables such as gender, age, civil status, length of work experience, highest educational attainment, specialization, and designation. It also tried to find out if there are significant differences in the respondents’ level of proficiency in the four macro-linguistic skills and if there are significant correlations between the respondents’ English proficiency and the identified demographic variables. Using descriptivecorrelational research design with academic professionals working in a higher education institution in the Philippines as respondents, this study found that the respondents have intermediate level of English proficiency (B1, CEFR), lower than what is expected from academicians in a higher learning institution. Findings further reveal that there are significant differences in the respondents’ level of proficiency in the identified macrolinguistic skills and that age, educational attainment, and length of service are positively correlated with English proficiency, while gender, civil status, specialization, and designation are negatively correlated with the same. Implications of the findings are discussed in relation to curriculum development, language teaching and assessment, and language policy and planning.

an individual. In its 8 th English Proficiency Index (EF EPI) report, it revealed that there are correlations between English proficiency, higher incomes, greater connectivity, and innovation (EF Education First, 2018). In the Philippines, English proficiency provides competitive edge on career prospects among graduates and economic competitiveness for the whole country. Filipinos who have high English proficiency easily get hired locally and abroad. With 70 percent of its population being able to speak English (Hernandez, 2015), the country is also the preference of foreign investors for business process outsourcing (BPO). In recent decades, the country has benefited economically from its human capital's proficiency in the English language.
To a certain extent, English proficiency influences the Philippine's economy. It is therefore important that Filipinos are highly proficient with the English language. However, getting highly proficient is perceived to be long and difficult process, influenced and controlled by various internal and external factors (Nallaya, 2012). In the case of Filipino adult language learners who make up the country's labor force, what are these factors? This study aimed to examine the factors or variables influencing adult language learners' (ALL) English proficiency development. Specifically, it tried to answer the following specific questions: 1) What is the level of English proficiency of the respondents when they are grouped according to demographic variables such as gender, age, civil status, length of work experience, highest educational attainment, specialization, and designation?; 2) Are there significant differences in the respondents' level of proficiency in the identified macro-linguistic skills?; and 3) Is there a significant correlation between the respondents' English proficiency and their age, gender, civil status, length of service, educational attainment, specialization, and designation?
The current study was also guided by the following hypotheses: 1. There are no significant differences in the respondents' level of proficiency in the identified macrolinguistic skills.
2. There are no significant correlations between the respondents' English proficiency and their age, gender, civil status, length of service, educational attainment, specialization, and designation.
Investigating the correlations between English proficiency and demographic variables (factors) is useful in understanding language learners, curriculum development and design, language assessment, and language policy and planning. This study is useful to teachers and curriculum developers as this provides them insights that can help them design engaging activities and materials that suit the nature and level of adult language learners. It is also valuable to test developers as results can reveal some points to consider in designing appropriate language assessment tools. Lastly, it is beneficial to school leaders and decision makers as results can help them formulate informed decisions and policies geared towards holistic, meaningful, and transformative English language education. Hymes (1974) defines language proficiency as the ability to use a language effectively in authentic communicative situations. Harley, Cummins, Swain, and Allen (1990) also describe language proficiency as the ability to use a language accurately (linguistic knowledge) and appropriately in various contexts (sociolinguistic competence) and to organize one's thoughts through language (discourse competence). More comprehensively, according to Canagarajah (2006), language proficiency does not only include ability to use a language correctly in appropriate context for a functional purpose but also awareness of different norms of interaction attached in the language and in its different varieties. In the case of English being the world's lingua franca, English proficiency is currently used as 'a benchmark to assess an individual's inclusion or exclusion with regard to politics, commerce, the economy, society, and education' (Nallaya, 2012, p. 149).

English Proficiency
There are many widely recognized systems and frameworks used to define and measure language proficiency. One of which is the American Council on the Teaching of Foreign Languages (ACTFL) Proficiency Guidelines. Using a scale with descriptors, it categorizes language learners into novice, intermediate, and advanced (ACTFL, 2018). In Europe and some parts of the globe, the Common European Framework of Reference for Language Learning, Teaching, and Assessment (CEFR) is a popular guideline used to describe proficiency of foreign language learners. Widely used in language teaching worldwide, it provides a method of learning, teaching, and assessing foreign languages (Council of Europe, 2001). Can understand and use familiar everyday expressions and very basic phrases aimed at the satisfaction of needs of a concrete type. Can introduce him/herself and others and can ask and answer questions about personal details such as where he/she lives, people he/she knows and things he/she has. Can interact in a simple way provided the other person talks slowly and clearly and is prepared to help. (Council of Europe, 2001, p. 25) In measuring proficiency of individuals in the English language, globally recognized high-stakes tests such as the International English Language Testing System (IELTS), the Test of English as a Foreign Language (TOEFL), and the Test of English for International Communication (TOEIC) are used. These standardized English tests are commonly used by academic, government, and corporate organizations for admission, hiring, job promotion, professional certification, and migration purposes. The EF Education First's Standard English Test (EFSET), an online adaptive English test of reading and listening skills, is also used to draw more comprehensive picture of English proficiency across countries and geographical regions. Educators, officials, and business leaders use results for organizational strategic planning and policymaking.  Education First, 2018). These data are relevant to the current study since its main purpose is to examine the proficiency level of a small group of English language learners in the Philippines working in a specific context where English is used as the medium of communication.

Demographic Variables Influencing Language Proficiency
Learning the English language is directly and indirectly influenced by several factors. Age, socio-economic status, educational attainment, aptitude, intelligence, learning styles, personality, learner's mastery of his first language, and opportunities or exposure to English are believed to be influential in the rate of learning and can predict learner's success in gaining proficiency with the language (Fromkin, Rodman, & Hyams, 2014;Khasinah, 2014;Ellis, 1986;Orillos, 1998, Latu, 1994, Frakfurt International School, 2018. In addition, motivation and socio-cultural background can also influence language learning outcomes (Nallaya, 2012).
In the study of Prapphal and Oller (1982) investigating the relationship between demographic variables and English proficiency of university students in Thailand, it was found that demographic variables were significantly related to English proficiency. El-Omari (2016) also found the same among Jordanian secondary school students in English. Specifically, he found that social and socio-economic factors have direct impacts on students' achievement in English language learning.
In terms of gender as demographic variable, females tend to be more proficient and communicative in English than their male counterparts (EF Education First, 2018). This was supported by the 2017 TOEIC test takers performance report released by Educational Testing Service (2018b). Based on the TOEIC listening and reading comprehension scores across gender, females scored higher than male test takers. In terms of age, younger learners learn a (new) second language faster and easier (Roherick, 1983). This was found true in the study of Wang (1999) where younger people who migrated to Canada tend to develop higher proficiency faster and easier than those who arrived at an older age.
On the other hand, in Morocco, El Ghouati, Koumachi, and Khoumich (2018) investigated the potential impact of students' demographic variables (i.e. gender, age, computer experience, and computer use) on achievement in English among Moroccan university students. They found that there is no significant relationship between gender, age, and computer experience and frequency of use on students' level of achievement. Similarly, Phon (2017) also examined the factors affecting the English language proficiency of students majoring in English at a rural university in Cambodia. Findings revealed that socio-economic status as a demographic factor does not significantly affect English proficiency. The study of Hines (2015) also confirmed that there is no significant relationship between demographic variables and English language learning.

Research Design
This study utilized the descriptive-correlational research design to describe the demographic variables and existing English proficiency of the respondents and to establish patterns of relationship of both. According to Grove, Gray, and Burns (2014), this method is appropriate when a study's aim is to describe variables and examine relationships among these variables. This method can help identify many interrelationships of variables that have already occurred or are currently occurring. As a process, according to Wragg (2012), it begins describing what happens and then it examines systematically the correlation between process and outcome.

Respondents of the Study
The respondents of the study were academic professionals in one of the higher education institutions (HEI) in the Philippines where English is the main medium of instruction. With their ages ranging from 20 -69, they are considered adult language learners. They have diverse backgrounds with the English language considering their age when they started learning the language, amount of exposure to the language, academic positions, and socioeconomic circumstances. All faculty members in the undergraduate programs of the subject HEI were involved in the study.

Data Gathering Instrument
To gather data on demographic variables such as age, gender, civil status, length of service, educational attainment, specialization, and designation, a survey questionnaire was used. To determine the English proficiency levels of the respondents, an English proficiency test from the Official IELTS Practice Materials Volume 2 published by the University of Cambridge Local Examinations Syndicate [UCLES] (2010) was administered by the researcher. Containing items from the previously administered tests, it resembles the actual IELTS test in terms of format, content, and level of difficulty. The testing material used was secured in Dubai, UAE from IDP IELTS Exhibition in TESOL Arabia Conference in 2012; hence, its non-exposure to the respondents in the Philippines was highly assured. The IELTS Academic Module was particularly used as an assessment tool for this study since the respondents were working in an academic context.

Data Gathering Procedure
Prior to the administration of the survey questionnaire and the English proficiency test, permission to conduct the study was sought. After approval and endorsement from the school management, the researcher visited different colleges / departments and communicated to the respondents. Having first-hand experience in taking the test and in giving IELTS preparation courses in the Middle East, the researcher personally administered the English proficiency test by college/department. He administered the Listening, Reading, and Writing sections of the test for one week covering all colleges and the Speaking section (Interview) for another week. The researcher systematically ensured the consistency and correctness of the data for each respondent. Names or codes were written in the profile form and answer sheet so that respondents could personally access the result of their test at a later time from the researcher. The Speaking test was also recorded and videotaped for researcher's reference during scoring and for respondents' future use. The speaking and writing parts were scored using the prescribed rubrics. The collected data were then tallied, analyzed, and correlated in order to answer the specific questions and confirm/reject the hypotheses

Statistical Treatment of Data
Descriptive statistics was employed to compute means, standard deviations, frequency counts, and percentages of data on demographic variables and English proficiency. The one-way analysis of variance (ANOVA) for repeated measures was also used to determine significant differences between the means of the four macrolinguistic skills (listening, reading, writing, speaking).
The following formula was used: To correlate the respondents' language proficiency and their age, the Pearson-r was used with the following formula: To compute correlation between language proficiency and respondents' gender, length of service, and educational attainment, the z-test was used with the following formula: The Spearman -Rho Rank Difference Coefficient of Correlation was used to find correlation between respondents' language proficiency and their civil status, specialization, and designation.
The formula of which is as follows: P = 1 -6 zD 2 n(n 2 -1) The null hypotheses were tested at .05 level of significance.
To analyze the English proficiency performance of the respondents, the following table of band descriptors was used. Has a fully operational command of the language with only occasional unsystematic inaccuracies and inappropriate usage; may misunderstand some things in unfamiliar situations; handles complex detailed argumentation well.

Good User
Has an operational command of the language, though with occasional inaccuracies, inappropriate usage and misunderstandings in some situations; generally handles complex language well and understands detailed reasoning.

Competent User
Has a generally effective command of the language despite some inaccuracies, inappropriate usage and misunderstandings; can use and understand fairly complex language, particularly in familiar situations.

Modest User
Has a partial command of the language and copes with overall meaning in most situations, although likely to make many mistakes; should be able to handle basic communication in own field.

Limited User
Basic competence is limited to familiar situations; frequently shows problems in understanding and expression; not able to use complex language.

Extremely Limited User
Conveys and understands only general meaning in very familiar situations. Frequent breakdown in communication occurs.

Intermittent User
Has great difficulty understanding spoken and written English.
1 Non-user Has no ability to use the language except a few isolated words.

0
Did not attempt the test Did not answer the questions.    Table 4 reflects the level of proficiency of the respondents when grouped according to gender. As shown above, the male respondents have slightly higher score/level than female with a difference of 0.12. Overall, the male group is considered "modest user" while the female group as "limited user." The data refute the common belief that females are more linguistically inclined than males (Orillos, 1998).  Table 5 shows that singles have the highest mean score among the respondents with 5.06 (Modest User), followed by the married ones with 4.93 (Limited User), and the widow with 3.00 (extremely limited user). The data illustrate that single respondents tend to be at the advantage in gaining higher English proficiency.  Table 6 shows that respondents with the length of service of 26 -30 years have the highest English proficiency level with a mean score of 5.50 (Modest User), followed by those with 6 -10 years in the service with a mean score of 5.21 (Modest User), 0 -5 years with 4.94 (Limited User), 11 -15 years with 4.57, and 36 -40 years with 3.00 (Extremely Limited User). The data reveal that having longer experience at work where English is used as the medium of communication does not always guarantee development of higher English language proficiency.  Table 7 shows that respondents who have masteral degree have higher level of English proficiency. As reflected above, masteral graduates have an overall score of 5.20 (Modest User) while the non-masteral degree holders only have a score of 4.42 (Limited User). The data suggest that undergoing higher educational qualification can help develop English language proficiency.   Table 10 shows the level of English proficiency of the respondents when grouped according to designation. The respondent with the designation of Department Head has the highest level of English proficiency with the mean score of 7.0 (Good User), followed by Program Coordinator with the mean score of 5.16 (Modest User), and Instructor with a mean score of 4.86 (Limited User). As reflected, the level of English proficiency progresses as the designation goes higher.    Table 13 shows the correlation between the respondents' English language proficiency and their gender. As shown in the table, the male respondents (X 1 ) have an overall average proficiency of 5.00 with a Standard Deviation (SD) of 0.97 and a frequency of 19. The female respondents (X 2 ) on the other hand have an overall average score of 4.89 with the Standard Deviation (SD) of 1.15 and a frequency of 22. Using the z-test, the value of z is 0.34. Based on the result of the z-test vis-à-vis 0.05 level of significance, there exists no significant correlation between the respondents' gender and proficiency; hence, null hypothesis (H 0 ) is accepted. ∑D 2 =2 0.5 Table 14 shows the correlation between the respondents' English proficiency and their civil status. As reflected, there were 16 single respondents with a mean proficiency score of 5.06, 24 married respondents with a mean proficiency score of 4.93, and 1 widow with a mean proficiency score of 3.00. Applying the Spearman Rho Rank Difference Coefficient of Correlation, the value of P is 0.5. Comparing the value of P against 0.05 level of significance, it is revealed that there is no significant correlation between civil status and level of English language proficiency; hence, null hypothesis (H 0 ) is accepted.  Table 15 shows the correlation between the respondents' level of English proficiency and their length of service. As shown in the table, the length of service of the respondents has a sum of 304.5, a mean of 7.43, and a standard deviation of 7.86. Their overall English proficiency score, on the other hand, has a sum of 202.5, a mean of 4.94, and a standard deviation of 1.07. Applying z-test computation, the value of z came out at 5.38, indicating that there is a very significant correlation between the respondents' English proficiency and length of service. This means that as one grows older in the teaching profession, the more he develops his proficiency in the English language.  Table 16 shows the correlation between the respondents' English proficiency level and their educational attainment. As reflected in the table, there were 27 masteral graduates (X 1 ) with a mean proficiency score of 5.2 and a standard deviation of 1.03. Bachelor graduates (X 2 ) on the other hand were registered at 14 with mean proficiency score of 4.43 and a standard deviation of 0.96. Applying the z-test, the value of z came out at 2.36. The z value indicates that there is a significant correlation between educational attainment and English proficiency; hence, the null hypothesis is rejected. The finding further implies that as the respondents' qualifications go up, their level of English proficiency also increases.  Table 17 presents the correlation between the respondents' level of English proficiency and their specialization. As shown in the table, respondents specializing in Nursing have a proficiency mean score of 5.78, followed by those specializing in English with a mean score of 5.4. Respondents with specializations in Midwifery and Finance have the lowest proficiency score of 3.00 each. Overall, the sum of the squared standard deviation of all the specializations is 691.12. Applying the Spearman Rank Difference Coefficient of Correlation (P), the value of P came out at 0.39. The P value of 0.39 indicates that there is no correlation between the respondents' level of English proficiency and their specialization basing at .05 level of significance. Therefore, the null hypothesis (H 0 ) is accepted.  Table 18 shows the correlation between the respondents' level of English proficiency and their designation. As reflected in the table, Department Head has the highest mean score of 7.00 and D 2 of 4, while Coordinator has a mean score of 5.16 and D 2 of 0. Instructor has a mean score of 4.86 and D 2 of 4. After applying Spearman Rank Difference Coefficient of Correlation, the value of P came out at -1, which means that there is a perfect negative correlation between the respondents' level of English proficiency and their designation; hence, the null hypothesis (H 0 ) is accepted.

Discussion
Results show that the respondents' level of English proficiency varies depending on the specific demographic variable being investigated. When the mean band scores from each demographic variable are averaged, the general English proficiency level of the respondents is 4.12 (Limited User), which means that the respondents' basic competence in English is limited to familiar situations, that they frequently exhibit difficulties in understanding and expression, and are unable to use complex language (British Council, 2018 Looking closely at the mean scores when grouped according to demographic variables, there are some patterns and tendencies that can be established. For example, as educational attainment and designation progress, the level of English proficiency also increases. In terms of civil status, single respondents seem to be at the advantage in gaining higher proficiency in English. In terms of specialization and length of work experience, English proficiency is not exclusive to those who studied English language and to those who are frequently exposed to it. The results also interestingly reveal that English proficiency is not only for females and that it chooses no specific age. When the relationship of the respondents' levels of proficiency in the four (4) macro-linguistic skills was examined, it was found that there are significant differences in their level of proficiency in listening, reading, writing, and speaking; hence, the null hypothesis (H 0 ) is rejected. It is a common belief that language skills are related and influence one another. However, in this study, results reveal that there is no uniform level of the respondents' English language proficiency in the four macro-linguistic skills. While one may perform high in one skill, he/she may also perform low in other skills. Further, while all the skills make up one's overall English proficiency level, it is also implied in the findings of the study that one skill does not necessarily affect or influence the other skill/s.
Investigating the relationship of the respondents' demographic variables and their levels of English proficiency, it was found that there are positive correlations between age, educational attainment, and length of service; hence the null hypotheses for these demographic variables are rejected. The positive correlation between age and English proficiency supports the claim that age affects the rate of language proficiency development (Fromkin, Rodman, & Hyams, 2014) and that younger learners tend to learn faster and easier (Roherick, 1983;Wang, 1999) than adult learners. The strong correlations of educational attainment and length of service with English proficiency may be due to the fact that when learners pursue post-graduate studies, they are exposed to many opportunities to use the English language. Similarly, when learners work in an English-rich environment at a longer time, they tend to develop higher proficiency with the language.
While the abovementioned variables positively correlate with English proficiency, gender, civil status, specialization, and designation seem otherwise; hence, the null hypotheses for these demographic variables are rejected. The results confirm the finding of El Ghouati, Koumachi, and Khoumich (2018) that gender does not influence English proficiency. The strong negative correlations between civil status, specialization, and designation suggest that an individual can be proficient in English regardless of what s/he is and what s/he does in life. These negative correlations also suggest that there might be more important unexplored factors beyond demographic variables that should be looked into in order to find out what propel and what hinder English proficiency development among adult language learners.

Conclusion
Findings reveal that the respondents of this study have intermediate level of English proficiency, lower than what is expected from adult language learners working in a higher education institution. Findings further disclose that there are significant differences in the respondents' level of proficiency in the identified macrolinguistic skills. Lastly, results show that demographic variables such as age, educational attainment, and length of service are positively correlated with English proficiency, while gender, civil status, specialization, and designation are negatively correlated with the same.
The results imply that there is a need for an English language enhancement program designed to increase the respondents' English proficiency level. Since results of this study also show that there are significant differences in scores in the four macro-linguistic skills, there is a need for an integrative approach in teaching, curriculum design, and instructional materials development to ensure that no language skill is left behind. The positive correlations between English proficiency and age, educational attainment, and length of service also demand for more varied opportunities for personal growth and professional development to keep more qualified and more seasoned academic professionals in the academe. The non-correlations of other variables such as gender, civil status, specialization, and designation with English proficiency also suggest that there are more significant factors beyond demographic variables influencing the respondents' English proficiency development, hence the need for more comprehensive investigation along this line of inquiry.
While the findings of this study indicate that the English proficiency of the respondents is lower than the prescribed level, it should be noted that such findings are based only on an official practice test that is not administered in an official test condition. The results can provide general insights on understanding adult language learners, designing curriculum, language testing, and formulating policies on training and development but may lack reliability and generalizability considering the limitations of the instrument used and the size of the population. Future researchers interested in investigating the same line of inquiry should use official data from relevant testing agencies in order to render more valid and reliable conclusions.