Comparison of German Translations of the System Usability Scale – Which to Take?

Abstract

The System Usability Scale (SUS) is a reliable tool for usability measurement and evaluation. Since its original language is English, a translation is required before a target group can answer it in their native language. The challenge of translating questionnaires lies in the preservation of its original properties. Different versions of a German SUS have been proposed and are currently in use. Objective of this work is to find and compare available German translations. Four versions were found and compared in terms of the translation process and the exact wording of the translation. Only the version of Gao et al. has been systematically validated, but has an unnatural wording. Although not validated yet, the proposed version of Rummel et al. is a good compromise between wording and methodically clean development. The version of Lohmann and Schäffer is the close runner up, as it may improve the wording at the expense of methodological accuracy. Since the version of Rauer gives no information about its translation process, it is considered least preferred of the four compared translations.

Development and Validation of a Positively Worded German Version of the System Usability Scale: The SUS-DE-Pos

“…In addition, all four versions of the SUS investigated in the present work provided ratings for the stimuli websites, which were very close to each other. Furthermore, the SUS version by Rummel (2016) has been deemed to use more natural wording compared to the one by Gao et al (2020) in a prior review comparing different German SUS versions (Brix et al, 2023).…”

“…Recently, a scoping review sought to clarify the landscape of varying German SUS versions. In their work, Brix et al (2023) aimed to identify and compare all currently available German translations of the SUS. They found four versions and examined them regarding the translation process and wording.…”

“…The first version identified was by Rummel (2016), but while its wording is promising, this SUS variant has not yet undergone proper validation. Conversely, the second version recommended by Brix et al (2023) is the SUS translation by Gao et al (2020). However, while this version has been systematically validated, it was criticized for containing unnatural wording.…”

“…After the language center review, the new set of items was investigated in a first preregistered online study alongside the German SUS versions proposed by Gao et al (2020) and Rummel (2016) to identify those positively worded items psychometrically closest to the negative counterparts. The two German SUS versions were chosen based on the recommendations by Brix et al (2023). As a next step, the new positive versions of the German SUS, containing only positive items (SUS-DE-Gao-Pos, SUS-DE-Rummel-Pos), were compared to the other two German versions of the scale (SUS-DE-Gao, SUS-DE-Rummel) in a second preregistered online study, situated in both a desktop/laptop and mobile setting, following current best practices for scale quality investigation.…”

“…In summary, the lack of an agreed-upon German version of the SUS motivated us to investigate and compare two promising versions identified in past research (Brix et al, 2023). Furthermore, the problems caused by reverse-coded items and the apparent benefits of a positive-item version of a survey scale encouraged the development of an alternative version of the SUS without negatively worded items.…”