In classrooms shaped by migration, language diversity and competing cultural expectations, a teacher’s confidence can determine whether difference becomes a resource for learning or a source of persistent tension. A new cross-national study involving Japan and Colombia argues that the tools used to measure this confidence may be too blunt for the classrooms they are intended to describe. Published in Large-scale Assessments in Education, the research proposes a more discriminative and culturally responsive approach to measuring teacher self-efficacy in multicultural settings, using an advanced statistical technique known as the generalized partial credit model.
Teacher self-efficacy refers to educators’ beliefs about their ability to organize and carry out the actions needed to achieve instructional goals. It is not identical to teaching skill, formal qualification or classroom performance. Instead, it captures how capable teachers believe they are when confronting demanding professional situations, from adapting lessons for students with different linguistic backgrounds to managing cultural misunderstandings and creating an inclusive learning environment. Research has repeatedly linked stronger self-efficacy with persistence, instructional innovation, classroom management and willingness to support students facing educational disadvantage.
The study focuses on self-efficacy in multicultural classrooms, an area that has become increasingly important as schools serve students whose languages, cultural identities and migration experiences may differ substantially from those of their teachers. Yet measuring this form of confidence across countries is difficult. A survey item that appears straightforward in one national context may carry a different meaning in another. “Managing culturally diverse classrooms,” for example, may refer in Japan to one set of challenges involving linguistic minority students, regional diversity or internationalization, while in Colombia it may be interpreted through the experiences of Indigenous, Afro-Colombian, rural or migrant communities. Comparing scores without examining these differences can produce misleading conclusions.
The researchers therefore revisited the teacher self-efficacy component of the Teaching and Learning International Survey, or TALIS, an international program conducted by the Organisation for Economic Co-operation and Development. TALIS asks teachers about their professional practices, working conditions and beliefs, creating a valuable source of cross-national evidence. However, the study suggests that conventional scale construction may overlook important distinctions between response categories and may not fully account for the cultural contexts in which teachers answer. A scale can appear statistically reliable while still failing to distinguish among teachers who possess moderately different levels of confidence.
To address that problem, the researchers applied the generalized partial credit model, or GPCM, a model from item response theory designed for ordered survey responses. In a typical self-efficacy questionnaire, teachers may select options ranging from “not at all” to “a lot,” or from low to high confidence. The GPCM does not treat those categories as merely interchangeable points on a simple total score. Instead, it estimates how likely teachers with different underlying levels of self-efficacy are to choose each response category for each item. This makes it possible to examine whether an item is informative, whether its response thresholds are ordered and whether it can discriminate effectively between teachers at different points on the confidence continuum.
That technical distinction matters because a poorly functioning item can distort international comparisons. If nearly every teacher selects the highest category, the question provides little information, even if it contributes to a high reliability coefficient. Conversely, if teachers from different countries interpret the same wording differently, a difference in average scores may reflect cultural response patterns rather than a genuine difference in self-efficacy. The GPCM allows investigators to inspect these response processes more closely, identifying items that offer useful information and those that may require revision. It can also reveal whether the scale functions consistently across groups, a central concern in cross-cultural measurement.
Japan and Colombia provide an especially revealing comparison. The countries differ in educational traditions, demographic patterns, historical experiences and the ways schools encounter cultural and linguistic diversity. Japan is often associated with a relatively homogeneous national population, but its classrooms include growing numbers of students with migrant backgrounds, multilingual families and varied cultural identities. Colombia, meanwhile, encompasses marked regional, ethnic and socioeconomic diversity, including Indigenous and Afro-descendant communities as well as students affected by internal displacement and international migration. The meaning of inclusion, adaptation and classroom difference cannot be assumed to be identical in the two systems.
By placing the countries within the same analytical framework, the study seeks to separate genuine variation in teacher self-efficacy from variation caused by the measurement instrument itself. Its central contribution is not simply a new ranking of national scores. Rather, it is an effort to determine which survey items provide the clearest evidence about teachers’ confidence and whether those items operate fairly across cultural settings. The analysis moves beyond the assumption that a single global scale can be transferred unchanged from one education system to another. It treats measurement as part of the educational phenomenon, not as a neutral step that occurs after the important work is finished.
The findings point toward a more differentiated understanding of teacher self-efficacy in multicultural classrooms. A teacher may feel highly confident communicating with families from different cultural backgrounds but less confident adapting assessment, addressing culturally rooted conflict or supporting students who are still developing proficiency in the language of instruction. Treating all these tasks as one undivided ability can conceal meaningful professional needs. A more discriminative scale can provide a profile of strengths and weaknesses, offering researchers and education authorities a sharper view of where teachers require training, mentoring or institutional support.
The cultural dimension is equally important. Teacher confidence does not develop in isolation from school policies, curriculum expectations, community relationships or the resources available to educators. A teacher may report low self-efficacy not because of limited commitment or competence, but because the school lacks interpreters, specialist staff, culturally relevant materials or clear procedures for working with families. Conversely, high confidence may reflect a supportive institutional environment rather than individual ability alone. A culturally responsive measure should therefore be used to understand context, not to label teachers or hold them personally responsible for structural problems.
The researchers’ approach also carries implications for future international surveys. Large-scale assessments depend on instruments that can travel across languages and education systems without losing their meaning. The study demonstrates why translation, statistical calibration and cultural interpretation must work together. Advanced models such as the GPCM can help identify where a scale is precise and where it becomes vague, but statistical refinement cannot replace consultation with teachers and communities. The strongest instrument will combine psychometric evidence with careful attention to how educators actually understand the situations described in survey questions.
For schools, the practical message is immediate: measuring teacher confidence more accurately can make professional development more targeted and more equitable. Instead of offering generic workshops on diversity, education systems could identify whether teachers need support with multilingual instruction, culturally responsive assessment, family engagement, conflict mediation or inclusive classroom management. More precise evidence may also help policymakers distinguish between a lack of confidence and a lack of resources. That distinction is crucial. Training can strengthen professional knowledge, but sustainable inclusion also requires time, staffing, leadership and learning materials that reflect students’ lives.
The study arrives as multicultural education is moving from a specialist concern to a defining challenge for school systems worldwide. International migration, urbanization and shifting patterns of identity are bringing greater diversity into classrooms that were often designed around assumptions of linguistic and cultural uniformity. In that environment, teacher self-efficacy is not a soft psychological variable at the margins of education policy. It is a potential lever for how schools respond to difference, whether students feel recognized and whether teachers are prepared to turn diversity into an academic and social advantage.
The broader lesson is that better measurement can change what education systems are able to see. A single score may suggest that teachers in two countries are equally confident, or that one group is more prepared than another. A carefully designed, culturally responsive scale can reveal a more complex picture: which classroom tasks generate confidence, where uncertainty is concentrated and how national context shapes professional judgment. By applying the generalized partial credit model to TALIS data from Japan and Colombia, the researchers offer a path toward international comparisons that are not only more statistically sophisticated, but also more attentive to the lived realities of teachers and students. In an era when classroom diversity is expanding faster than many systems can adapt, that precision could help turn inclusion from an aspiration into measurable, improvable practice.
Subject of Research: Teacher self-efficacy in multicultural classrooms and the cross-national validation of a culturally responsive TALIS scale.
Article Title: Toward a more discriminative and culturally responsive TALIS teacher self-efficacy in multicultural classrooms scale: cross-national evidence from Japan and Colombia using the generalized partial credit model
Article References: Published in Large-scale Assessments in Education; study based on TALIS teacher survey data from Japan and Colombia.
Image Credits: AI Generated
DOI: 10.1186/s40536-025-00272-9
Keywords: Teacher self-efficacy, multicultural classrooms, culturally responsive measurement, TALIS, generalized partial credit model, item response theory, cross-national comparison, Japan, Colombia, psychometrics.
