Language assessments and the scores from these evaluations often act as gatekeepers in policy decisions related to immigration, education, and employment (McNamara et al., 2019). Test bias can call into question the fairness of the assessment and the decisions that are made based on the test scores (Messick, 1989). In the present study, we explore a potential source of testing bias - namely, cross-linguistic transfer, or the influence of a speaker's first language knowledge on their second language use (Bailey, Lockary, & Higby, 2024) - via the following question: When a language test designed for a linguistically homogenous public is used for the assessment of linguistically diverse speakers, does this lead to an observable bias?

A group of participants (n=929) representing five different first languages (Arabic, Chinese, Farsi, French, and Spanish) took a multiple-choice ESL (English as a second language) vocabulary test designed for use in a French-speaking institution. Given this multiple-choice test's French-based distractors, we hypothesized that when controlling for overall English proficiency, French speakers would obtain lower scores than all other groups. However, our analyses revealed that in addition to the French group, Spanish and Arabic-speakers also performed significantly less well than the other participants. We put forward an explanation for these results and spotlight how unintended consequences can arise when bias is left under-investigated in the test revision cycle. Leveraging these insights, we propose some practices to incorporate in an iterative test validation process to control for bias.

References:

Bailey, L. M., Lockary, K., & Higby, E. (2024). Cross-linguistic influence in the bilingual lexicon: A framework for understanding semantic transfer. Bilingualism: Language and Cognition.

McNamara, T., Knoch, U., & Fan, J. (2019). Fairness, Justice, and Language Assessment. Oxford University Press.

Messick, S. (1989). Validity. In R.L. Linn (Ed.), Educational Measurement (3rd ed., pp. 13-103).

Language assessment has become a key component of immigration policy in many destination countries, shaping eligibility for entry, permanent residence, and citizenship. Although often presented as neutral measures that promote social and economic integration, critical scholarship shows that these tests frequently function as covert mechanisms of governance that regulate mobility and threaten linguistic human rights (Shohamy, 2022). A core problem is that immigration systems increasingly rely on high-stakes language tests without sufficient validity evidence to demonstrate that these assessments are fair, appropriate, or adequate for the consequential decisions they inform (McNamara & Roever, 2006; Saville, 2009).

This study proposes a systematic policy analysis to evaluate how immigration regimes justify and implement language requirements. Through a thematic analysis of publicly available legislative and regulatory texts across Canada, the United States, the United Kingdom, Australia, and New Zealand, the study examined how expectations of linguistic competence are framed, operationalized, and enforced within immigration pathways.

The analysis revealed consistent cross-national patterns: rising proficiency thresholds aligned with dominant-language norms; reliance on standardized tests lacking transparent validation; limited recognition of multilingual repertoires; and minimal procedural safeguards when test scores restrict immigration rights. Despite well-established principles advocating fairness, transparency, and responsible test use (ALTE, 2016), validity arguments remain weak or nonexistent.

The study proposes a reorientation of immigration language policy toward explicitly validated, ethically grounded assessment practices that acknowledge multilingualism as a societal asset. It contributes to language policy and planning by demonstrating how current approaches reproduce inequity and by outlining steps—public validation studies, participatory policymaking, and inclusive assessment design—to align practice with social-justice principles. The paper concludes with a call to action: to move from rhetoric to action, governments must ensure that language assessments in immigration support equitable inclusion rather than reinforce exclusionary control.

This paper examines the direction(s) of Francisation Québec in light of the program’s failure to realize its projected role as a backbone of the current government’s vision for Québec—one of viable potential for sovereignty that is anchored in a culturally, economically, and linguistically aligned society. My project first unravels the interdependent relationships between the place of Francisation in the modern Québec imaginary and the economic stakes of immigration and language. I then delineate actual conditions in the province that go unaddressed by Francisation Québec’s structure, namely employment, education, and access to public services. Finally, I interrogate the potential of Francisation Québec using examples from language policies in South Africa, South Korea, and Spain.

Francisation Québec administers French language classes and a collection of other services intended to support the prosperity across the province. The program, as conceived by the CAQ government, would, ideally, substantiate their overarching vision for Québec—one with viable potential for sovereignty founded on a culturally, economically, and linguistically aligned society.

However, Francisation Québec has continuously incurred numerous issues spanning from its organization to administration: Substantial budget cuts, inconsistent class accessibility, and a lack of infrastructural reinforcement impede its efficacy. Those meant to benefit from Francisation Quebec instead endure a domino effect of diminishing economic opportunities and discouraging prospects of social mobility without the promised support of communicative competency in French. Undoubtedly, the persisting failure of the program will have larger implications for the long-term growth of Québec. Briefly considering comparable situations in South Africa, Spain, and South Korea, I think through how Francisation Québec can effectively consider multi-level approaches that have been effectively implemented in programs in similar contexts.