Language test scores are difficult to compare across different languages, even when they use identical instruction and visual stimulus materials. For vocabulary items, cultural differences may yield differences in lexical usage frequencies. At the sentence level, complexity differences account for varying cognitive load. Dual language learning might jumble item difficulty rankings due to cross-linguistic influences. Still, equivalent language tests for different groups of single and dual language learners are urgently needed. In multilingual countries, language disorder diagnostics lack objectivity, language research findings are barely comparable cross-nationally, and, dual language learners’ language balance and/or dominance (i.e., potential predictors of cognitive development) cannot be validly measured. In this paper, we scrutinize design principles that are deemed to drive construction of culturally and linguistically fair oral language tests for single and dual language learning children with different societal and home languages. Moreover, preliminary validation analyses compare and contrast estimates of language dominance within dual language learners, which are obtained through aforementioned language tests, with parental assessments conventionally employed in such contexts. One-hundred-and-three single and 89 dual language learning preschoolers living in Switzerland and Germany (Age: M(SD) = 50.02(7.61) months) were tested for the societal languages French (26%) or (Swiss-)German (74%). Dual language learners’ home languages were Turkish (58%) or Italian (42%). Language skills for a child’s spoken language(s) were assessed via productive and receptive vocabulary plus sentence comprehension tests, each entailing 32 items. Eight items per language were taken from established language scales, which were subsequently translated either into the other societal or home language(s). To construct equivalent home and societal language tests, 16 semantic and linguistic item-twins (matched in complexity, frequency, etc.) were created based on the 16 selected and translated original items. We tested for measurement invariance between and within groups via (multiple-group) factor analyses for categorical items. The translated test items showed to be largely invariant between the societal and between the home languages. However, within dual language learners, measurement invariance of societal and home language tests was only observable for a limited selection, i.e., more simple sentence comprehension items. For these items, children’s language dominance estimates correlated only moderately with the parental assessment (phi = .34; p < .001). Generally, parents who rated their child’s home language skills as dominant tended to overestimate these or, alternatively, rather underestimated their child’s societal language skills. Implications of our research for dual language learning assessments in research and practice are discussed.

References are not included in the abstract