Explaining self-reported language proficiency gains of immigrant women

Professor Emertius Paul Anisef, York University
Professor Emeritus Robert Sweet, Lakehead University
Professor Maria Adamuti-Trache, University of Texas at Arlington
Dr. Sarah V. Wayland, Wayland Consulting
March 2012

This report uses longitudinal data from the Longitudinal Survey of Immigrants to Canada (LSIC) to 1) explore self-reported language competencies of immigrant women (with baseline comparisons to immigrant men) at the time of arrival and how these develop with time spent in Canada, and 2) examine the relationship between self-reported language competencies of immigrant women, individual characteristics, and the initial integration experience.

The views expressed in this document are those of the authors and do not necessarily represent those of the Department of Citizenship and Immigration Canada or the Government of Canada.

Copies of the full report are available upon request to Research-Recherche@cic.gc.ca.

• Read the executive summary:
  Explaining self-reported language proficiency gains of immigrant women (PDF, 359.89KB)

Executive summary

This report uses longitudinal data from the Longitudinal Survey of Immigrants to Canada (LSIC) to 1) explore self-reported language competencies of immigrant women (with baseline comparisons to immigrant men) at the time of arrival and how these develop with time spent in Canada, and 2) examine the relationship between self-reported language competencies of immigrant women, individual characteristics, and the initial integration experience.

In analyzing self-reported language competencies of immigrant women, the following set of themes and research questions are explored:

Theme 1: Comparative profiles

• Are there gendered differences in self-reported official language competencies at arrival and do these differences persist over time? Are there gendered differences in the settlement experiences of recent immigrants (e.g., difficulties with finding employment, level of satisfaction with life in Canada, etc.) by self-reported official language ability?
• Among immigrant women, are there differences at arrival in self-reported official language speaking ability by age group and other socio-demographic, economic, social and cultural characteristics?
• What are the patterns of self-reported official language competencies within six months, two and four years after arrival in Canada? Who reports gains over time, and who does not? Are there differences by age group, other socio-demographic characteristics, and immigration category?

Theme 2: Learning opportunities and barriers to language training

• Does the presence of perceived opportunities to learn English or French at arrival (6 months after arrival) and the perception of language as a barrier to integration (4 years after arrival) vary by age at arrival and other socio-demographic characteristics?
• Do the strategies employed to learn or build upon existing official language skills vary by age at arrival and other socio-demographic characteristics? Is self-reported official language competence indicative of ease of access to learning opportunities?
• Are there unique integration challenges experienced by recent immigrant females (e.g., finding employment, accessing education or training, finding housing or accessing health care) that respondents attribute to language barriers? Do lower official language skills reported six months, two years or four years after arrival in Canada result in perceived language barriers in finding employment?

Theme 3: Language proficiency gains

• How are the dynamics of the settlement experience shaped by self-reported language ability gains of recent female immigrants? Does self-reported language ability increase or decrease over time? Are there significant demographic, economic, social and cultural differences that distinguish those that report gains from those that do not?

Starting from the premise that language proficiency is one of the most important factors contributing to the successful integration of immigrants (see, for example, Skuterud, 2011, and Goldman, Sweetman, and Warman, 2011), a literature review was conducted and examined dozens of articles on second language acquisition, with a particular focus on Canada and on literature that addressed the situation of immigrant women. In examining whether immigrant women face unique integration challenges related to language proficiency, the review highlighted data showing that – stemming from factors related to the varying objectives of Canadian immigration categories of admission, cultural factors in countries of origin, family dynamics in Canada, and more -- immigrant women often have lower levels of proficiency in English or French relative to both persons born in Canada, and their male counterparts. Based on the literature review, the conclusion is drawn that immigrant men and women do not operate on a “level playing field” when it comes to second language acquisition. This is demonstrated by the findings identified in the review that, in addition to the barriers faced by immigrant men and women alike, women are more likely to experience isolation, lack of information, extensive family obligations, and cultural barriers, as well as to have lower levels of education, each of which can be a barrier to language learning.

Methodology

This study uses the Longitudinal Survey of Immigrants to Canada (LSIC), a comprehensive survey conducted by Statistics Canada and designed to study the process by which new immigrants adapt to or integrate into Canadian society. The population of the LSIC includes all immigrants who (1) arrived in Canada between October 1, 2000 and September 30, 2001; (2) were age 15 or older at the time of landing; and (3) landed from abroad and applied through a Canadian Mission Abroad. Three LSIC questionnaires were given to respondents at six months (Wave 1) after their arrival in Canada, two years (Wave 2) after landing, and four years (Wave 3) after landing. The interviews were conducted by Statistics Canada, with the first wave taking place between April 2001 and May 2002, the second between December 2002 and December 2003 and the last between November 2004 and November 2005. Among respondents in this longitudinal survey group, we include in our research sample only adult immigrants aged 20 to 59 years at arrival who did not live in Canada prior to immigration. By these criteria, the LSIC research sample examined in this report is composed of 6,090 respondents (79% of the LSIC longitudinal respondents) which represents approximately 122,980 of the 250,000 immigrants admitted to Canada between October 1, 2000 and September 30, 2001.

This study explores self-reported language competencies of immigrant women at time of landing and how these develop with time spent in Canada, as well as the relationship between self-reported language competencies of immigrant women, individual characteristics, and the initial integration experience (see Appendices B and C for a detailed explanation of the full methodology employed). The first step in this analysis is to assign each longitudinal respondent (LR) in our LSIC research sample an official language that can be used consistently to assess self-reported language proficiency and gains across the three survey waves. Assigning an official language (English or French) requires choice of criteria for selecting among official languages in Canada and choosing one main language. For example, the selection could be based on: (1) mother tongue; (2) language most spoken at home; (3) self-reported skills at arrival; (4) the language the respondent chooses to use in responding to survey questions in the LSIC; and/or (5) the dominant language used in the province of residence. As a result, a number of measures are used to develop and define the official language variable, with longitudinal respondents receiving ‘scores’ according to the various language proficiency criteria, and an official language is assigned accordingly.

In assigning each longitudinal respondent an official language (English or French), it is important that the outcome variables specific to understanding the self-reported language competencies of immigrant women are not solely based on the official language most commonly used in a province. We thus build upon and refine an approach employed by Adamuti-Trache (2011) (whereby English is the main official language for longitudinal respondents in all provinces except for those in Quebec, where it is the highest of either English or French self-reported speaking levels) and suggest that the official language assignment methodology that we employ (described in Appendix B) is more sensitive and responsive to variations in individual circumstances than a province-based methodology. For example, while the choice of province of destination is indicative of the language immigrants expect to communicate in and are likely to improve upon, the actual choice of language for the LSIC interview (where multiple language options are available) provides a behavioural indicator that perhaps more accurately reflects immigrants’ intention to communicate in a specific official language. This being the case, for immigrants whose mother tongue or language spoken at home is not the most commonly employed official language of the province, it is possible these immigrants may have found a niche for sustaining their preferred official language.

The second step in this analysis is the construction of self-reported language competence profiles in Wave 1 in order to allow us to relate them to social structural factors, immigrant specific factors (e.g., immigration category, region of last permanent residence, etc.), as well as prior level of education and other pre-migration factors. Different aspects of language comprise the self-assessed language proficiency outcome measure, requiring self-reported proficiency be defined to include both expressive and receptive language skills - that is, measures of speaking, writing and reading. Descriptive analysis is used to compare and associate self-reported language profiles with pre-migration factors and to gauge any possible self-reported language disadvantage for immigrant women (men are used as a baseline comparison), as well as identify women’s background characteristics that may relate to any self-reported language disadvantage.

The final step in the approach of our work comprises two tasks, both of which intend to explain gains made by immigrant women in their self-reported language competence through their initial 4 years of settlement (Waves 1, 2, and 3). First, we examine changes in self-reported language proficiency over time (Waves 1, 2, and 3) in order to describe the evolving self-reported language profiles of immigrant women over the initial settlement period in relation to various factors that first include the pre-migration factors examined in the earlier stage of our analysis. The evolution of self-reported language of immigrant women is then examined in relation to further learning obtained in Canada (e.g., LSIC respondents are specifically asked about the importance of learning and improving language proficiency, plans at arrival and steps to improve their skills), and situational constraints posed by the integration experience (e.g., perceived barriers). At this stage, we examine language learning involvement based on learning opportunities in formal (e.g., post-secondary education, language training, etc.) and informal (e.g., friends, on the job, community, etc.) contexts and assess which context may lead to higher gains in self-reported official language proficiency. Second, we develop a model of self-reported language proficiency over the first four years after arrival ( Waves 1, 2, and 3) in relation to individual differences, and situational/contextual factors, while controlling for pre-migration characteristics. We employ a linear mixed model based on official language overall scores to explore issues pertaining to changes in self-reported language proficiency. In using this model to predict self-reported language proficiency change, the relationship of interest is the contextual factors that comprise an ‘opportunity to learn’ notion.

Different aspects of language comprise the self-assessed language proficiency measure constructed, which takes into account both the separate and combined dimensions of speaking, reading and writing. We are thus able to examine change in each self-assessed language component as well as in relation to an aggregated or global measure; this approach also provides an element of cross-checking of self-reported language ability measures.

Findings

Theme 1:Comparative Profiles

Gender Differences

Our analysis reveals distinct gender differences on all four measures of official language proficiency and across all waves, with immigrant women’s self-reported language skill levels being moderately lower than those of immigrant men. All self-reported measures reflected a small increase over time, particularly from Wave 1 (6 months) to Wave 2 (2 years), with reading skill scores somewhat higher than speaking and writing skill scores. While generally, there was greater self-reported language skill variability among women, when comparing these proficiency measures, both men and women reported the most significant gains in speaking skills. Descriptive analysis suggested that immigrant women confront the tasks of settlement and integration in Canada with less developed language skills than immigrant men, as even after four years, about 15% of immigrant women reported they either cannot speak or, manage to speak an official language poorly. Similar patterns are found for immigrant women and men with respect to the impact of self-reported official language skills on employment status, diversity of friendship networks and level of satisfaction with life in Canada.

Effect of pre-migration factors on the self-reported official language skills of immigrant women

The self-reported ability of immigrant women to speak an official language in Wave 1 (6 months) is analyzed in terms of their demographic, economic, social, and cultural characteristics. A number of important variations were identified: (1) the self-reported inability to speak an official language well was centered among older and married immigrant women; (2) the self-reported ability to speak an official language well varied with level of formal education at time of arrival, with almost half of immigrant women with high school education or less self-reporting they cannot speak an official language or speak it poorly; (3) the self-reported ability to speak an official language well varied by immigrant class, with refugee women perceiving the greatest difficulties while economic principal applicant immigrant women reported the strongest facility in speaking an official language; (4) the self-reported ability to speak an official language well varied by region of origin of the immigrant women, with less than 30% of those from Eastern Europe, East/ Southeast Asia and South Asia self-reporting they spoke an official language very well when they first came to Canada; (5) the self-reported ability to speak an official language well related to employment status prior to arriving in Canada, with previously employed immigrant women reporting a greater capacity to speak an official language well; and (6) the self-reported ability to speak an official language well varies by province of destination, with Western provinces having the highest proportion of immigrant women arriving who reported an inability to speak an official language or only capable of speaking it poorly.

Gains in self-reported official language skills by immigrant women

Language gains are examined for each of the four measures of self-reported official language proficiency: speaking, reading, writing and an overall measure (to which special attention is paid). With respect to age at arrival, visible minority status, highest level of education prior to arrival, immigration class, and region of origin, self-reported reading scores were higher than writing and speaking scores across all groups or categories examined. Most striking are the findings in relation to age at arrival, with younger immigrant women reporting the most proficient language abilities while women in excess of 45 years reported negligible gains in official language proficiency. Less pronounced but noteworthy were the lower reported gains in official language proficiency by visible minority immigrant women in comparison with non-visible minority immigrant women; these differences are corroborated in analyzing region of last residence, with immigrant women from Asia reporting overall lower language levels even 4 years after arrival. While immigrant women of all levels of education at arrival reported gains in overall official language proficiency, those with university education reported higher levels of official language competency than immigrant women with high school education or less. In terms of immigrant class, economic principal applicant immigrant women reported the highest official language proficiency. Refugee women reported the strongest challenges in learning an official language but also reported the largest gains within four years of arrival in Canada. Immigrant women that arrived as economic spouses or dependents comprised the largest group (45.7%) and reported lower language skills (and greater variability of skills) in Wave 1 (6 months) than economic principal applicant immigrant women. However, after four years in Canada this group reported improvement, scoring themselves relatively well on most official language competencies.

Theme 2: Opportunities to learn and barriers to language learning

Perceived opportunities and barriers

Perceptions among immigrant women regarding official language learning opportunities and language as a barrier to settlement are investigated. A high proportion of female respondents in Wave 1 (6 months) (most living in provinces where English is the dominant language) indicated an awareness of English learning opportunities, with a much smaller proportion aware of venues to learn French. Women 45 years and older were least likely, and those with higher levels of formal education at arrival most likely, to recognize opportunities to learn official languages. In Wave 3 (4 years), nearly 30% of the immigrant women identified language difficulties as an impediment to settlement, a perception that differed most significantly according to age at arrival, visible minority status, and education level at arrival. Older women, those who were visible minorities (particularly those from Asia), and those with less education tended to perceive language barriers as the greatest impediment to settlement; this perception was not necessarily accompanied by a low level of awareness regarding official language learning opportunities. Refugee women were most likely to view language as a barrier to settlement, and conversely, economic principal applicant women were least likely to hold this perception.

Strategies employed to learn/improve language skills

Strategies to learn an official language employed at least once by female respondents within 4 years of arrival in Canada include: media, family or friends, contacts at work, self-study and language classes. Approximately two thirds of the immigrant women identified using media to improve either English or French language skills, and more than half relied on language classes, while about 40% denoted learning from friends and family, at work, or self-study . Overall, younger immigrant women were more likely to mobilize each of the five strategies than older immigrant women, particularly with respect to language training classes. Level of education was found to positively influence a respondent’s likelihood to employ at work and self-study methods. Differences in type of strategy engaged was also notable with respect to visible minority status, with work, self-study and language classes being relied upon most frequently by immigrant women who were visible minorities to improve language proficiency. Region of origin was also found related to type of learning strategy, with relatively higher proportions of immigrant women from Eastern Europe and the Middle East employing various learning strategies. Furthermore, immigration class was found related to learning strategies employed, with principal economic applicants making relatively stronger use of all five methods, while economic spouses and dependants tended to rely heavily on media, language classes and self-study, and family class immigrants turned to media, family or friends and language classes. Refugee women had the highest participation in language classes of all groups and were less reliant on at work or self-study methods. There is some correlation between official language skill improvement strategies and self-reported language competencies across all waves. Respondents who reported lower scores in official language competency were more likely to employ strategies for improving their language skills (though did not necessarily show proficiency gains over time). When learning strategies and self-reported competencies were cross-referenced, perhaps the most significant finding was that the largest improvements in speaking skills occurred among women who took language classes and employed the media to improve language competency.

Effect of language barriers on settlement

We examine the extent to which language barriers are seen to be challenges to integration, and whether there is a relationship between perceived barriers and low self-reported official language proficiency scores. The majority of female respondents in Wave 1 (6 months) do not find the resolution of settlement tasks to be problematic, and there is a noticeable decrease from Wave 1 (6 months) to Wave 3 (4 years) in the proportion of women who perceived language as a barrier in each settlement area (e.g., pursuing further education and training, finding housing, getting health care, and finding employment). Finding suitable employment and obtaining education and training were areas where language was perceived to present the greatest challenges. Average self-reported official language scores were lower for women who reported facing settlement challenges related to language proficiency; their scores improved over time.

Theme 3: Language proficiency gains

Modeling overall self-reported language proficiency gains

In order to examine gains in self-reported official language overall scores over the three waves of the survey, a linear mixed model is employed to examine the estimated effects of individual, as well as, situational, and contextual factors. In doing so, we distinguished between time-invariant variables (e.g. information collected at arrival, such as age and education level) and time-variant variables (e.g. information on the first four years after arrival in Canada, such as employment status). In the results of this model, we observed the employment situation of immigrant women in the research sample steadily improving over the four-year settlement period, although at four years after arrival, about a quarter of female respondents were non-participants in the labour force (not working and not looking for work). We also observed less than average scores on the ‘ethnic friends’ scale, which suggested to us that these immigrant women were more likely to maintain friendships in their own ethnic community than to ‘bridge’ relationships with other Canadians and ethnic groups. Gains in self-reported official language overall scores occurred largely within the first two years of arrival for the immigrant women in the LSIC immigrant research sample (n=3,090). Our use of a linear mixed model also revealed that time-invariant factors were more predictive than time-variant ones, with age at arrival, prior level of education, employment status before arrival in Canada, and immigrant class proving to be significant indicators of self-reported language proficiency.

Self-reported language proficiency gains among women that reported speaking an official language poorly (SOLP)

In the final section of the report, in order to focus on self-reported language gains among arguably those immigrant women who are most disadvantaged in terms of this competency, we used our linear mixed model to examine gains over time in self-reported official language overall scores for only those who reported speaking an official language poorly in Wave 1. The reduced sample of 1,370 immigrant women included those who self-reported they could not speak, spoke poorly or spoke fairly well an official language. In the results, some predictors (e.g., region of last residence, presence of children) were not found significant with respect to helping to explain self-reported official language scores of the SOLP group, whereas these variables were found significant when estimated for the full immigrant women research sample. Predictors of self-reported official language overall scores included in the model for the SOLP sample that were statistically significant are: survey wave (e.g., between 6 months after landing and 2 and 4 years after landing, respectively), age at arrival, prior level of education, prior work experience, immigration class, citizenship intention, volunteer work, post-secondary education participation, and ethnic diversity of friendship network. Compared to the ‘all immigrant women research sample’, the direction of association between the dependant and independent variables did not change for the SOLP sample. However, the magnitude of some estimates varied. For instance, economic spouses and dependants in the ‘all immigrant women research sample’ evaluated their official language ability approximately 0.2 points less than economic immigrant principal applicants. However, this difference is not evident among immigrant women in the SOLP group. Female immigrants in the SOLP group who reported taking language classes to improve their official language skills self-reported lower gains over time than women who did not take language classes. This counter intuitive finding may suggest that official language classes attract immigrant women with poor official language skills and who remain in the lower language skill range after four years in Canada. Alternatively, the finding may relate to the self-reported nature of data, that is, immigrants who face language barriers and/or are placed in situations of having their skills assessed (e.g., through language training classes) may become more aware of their language skill gaps and may provide a more accurate self-reporting of their language proficiency levels.

Conclusions

This study extends the research literature by examining variations in self-reported official language proficiency among female immigrants, paying particular attention to women who arrive in Canada and report speaking English or French only poorly. Our analysis of LSIC data offers support for the contention that immigrant women in Canada are generally less proficient than immigrant men in either official language upon arrival and, subsequently, may be more constrained in accessing the full range of official language learning opportunities. Pre migration factors were found to play a particularly important role in accounting for the self-reported inability of immigrant women to speak an official language well in Wave 1 (6 months). Specifically, older women who were married at arrival (particularly refugees), with lower levels of education at arrival and who come from specific regions of the world (e.g., Southeast Asia), and who were not actively engaged in the labour force prior to migration, were less likely to report speaking an official language well. We argue that foremost among the incentives to become proficient in an official language is the possibility of obtaining remunerative work or employment that matches the individual’s prior work experience; citizenship is a second incentive that motivates individuals to increase their exposure to one of the official languages. Be that as it may, our analysis confirms the existence of several barriers to women’s development of official language proficiency previously reported by Vanderplatt (2009) and Kilbride and Ali (2010). Opportunities to learn nevertheless exist and the choice of an effective learning strategy becomes important; this involves assessing the relative value of informal and formal approaches to learning.

Our analysis further indicated that most immigrant women reported possessing sufficient language proficiency to manage family and neighbourhood relationships and to engage in community activities. Obtaining work or accessing the further education opportunities that enhance employability, however, were perceived as more difficult challenges for the immigrant women, and success in both appears to depend heavily on official language proficiency. Proficiency in English or French is needed to enroll in a Canadian post-secondary education institution (Adamuti-Trache, 2011) and is also most often essential in obtaining well-remunerated employment and advancing one’s career. Immigrant women that have lower self-reported official language proficiency are more likely to report facing settlement challenges related to language. A comparison of the education level and employment status of immigrant women who reported speaking an official language poorly (SOLP) in Wave 1 (6 months) with those of immigrant women who self-reported possessing greater official language facility in the LSIC research sample underscored the importance to settlement success of differences in personal and situational resources. There is a distinct profile to the women in the SOLP group: they tended to be older at arrival, visible minorities, migrate as family class or refugees, and come from regions in Asia and the Middle East. These factors suggest less exposure to one of Canada’s official languages prior to arrival in Canada and are also associated with difficulties in developing language proficiency after migration. Although the literature suggests that language training contributes to greater official language proficiency, our analysis of the relative effectiveness of formal language instruction on self-reported official language ability among SOLP women indicated that other factors operating in the lives of this particular group of immigrant women may potentially compromise the potential benefits of these programs. How official language programs can be made more responsive to immigrant women who report speaking an official language poorly potential to benefit from such training remains a challenge.