Adaptation, translation, and validation of information literacy assessment instrument

Hani Syazillah, Nordin and Kiran, K. and Chowdhury, G. (2018) Adaptation, translation, and validation of information literacy assessment instrument. Journal of the Association for Information Science and Technology, 69 (8). pp. 996-1006. ISSN 2330-1635 (https://doi.org/10.1002/asi.24023)

[thumbnail of Syazillah-etal-JAIST-2018-Adaptation-translation-and-validation-of-information-literacy-assessment]
Preview
 Text. Filename: Syazillah_etal_JAIST_2018_Adaptation_translation_and_validation_of_information_literacy_assessment.pdf
Accepted Author Manuscript
Download (222kB)| Preview

Abstract

The assessment of information literacy (IL) at the school level is mainly dependent on the measurement tools developed by the Western world. These tools need to be efficiently adapted and in most cases translated to allow them to be utilized in other cultures, languages, and countries. To date, there have been no standard guidelines to adapt these tools; hence, the results may be cross-culturally generalized to a certain extent. Furthermore, most data analyses produce generic outcomes without taking into account the ability of the students, including the difficulty of the test items. The present study proposes a systematic approach for context adaptation and language translation of the preexisting IL assessment tool known as TRAILS-9 to be used in different languages and context, particularly a Malaysian public secondary school. This study further administers a less common psychometric approach, the Rasch analysis, to validate the adapted instrument. This technique produces a hierarchy of item difficulty within the assessment domain that enables the ability level of the students to be differentiated based on item difficulty. The recommended scale adaptation guidelines are able to reduce the misinterpretation of scores from instruments in multiple languages as well as contribute to parallel development of IL assessment among secondary school students from different populations.

CORE (COnnecting REpositories)