Multilingual assessment through pictures: Guidelines for material selection


Thematic Section: Creating language-assessment tools for North, South and in between 

language assessment, language impairment, cross-cultural, language acquisition, aphasia

Pernille Hansen, Norwegian language at Inland Norway University of Applied Sciences
Monica Norvik, Statped, Oslo; University of Oslo; Norwegian University of Science and Technology
Eva Soroli, University of Lille
Ewa Haman, University of Warsaw
Hanne Gram Simonsen, University of Oslo
Frenette Southwood, Stellenbosch University

Pictures are widely used to elicit language, for instance in studies on the lexical development of children, or on the language processing of neurologically healthy or clinical populations at various ages. However, pictures cannot be seen as neutral representations of the world: They are culture-specific, partial representations of reality, often presented disconnected from their context. Interpreting pictures is an acquired skill, dependent on social, cultural and cognitive factors.
Although these points are important to consider when pictures are used to assess language knowledge, particularly in multilingualism and cross-linguistic research, picture-based tools and experiments are often used uncritically across language groups and cultures. The result could be misleading conclusions about language proficiency, an important issue for both research and clinical contexts. In this paper, we discuss the development of picture material for lexical assessment, suitable for multilingual populations, focusing on an important step of the selection process: naming agreement.
Naming agreement involves empirically validating visual stimuli to ensure that language tasks are indeed appropriate for the intended purposes (e.g. for use across the languages of a multilingual individual). In this paper, we review recent naming agreement studies in several populations and languages. Crucially, we propose guidelines for creating and conducting naming agreement studies for designing assessment tools, including the development of a set of words with matching pictures, participant sampling and data collection. We also consider how naming agreement can be used to achieve comparability across languages, generations and societies.
There is a current lack of guidelines for naming agreement. In an increasingly multilingual world, there is both a growing interest in studying multilinguals and an urgent need for valid material to be used in clinical assessment tools suitable for this group. Our guidelines have applicability for anyone using picture stimuli to study language processing and language acquisition.