Measurement Quality of a Multi-item Scale in Plain Language

To ensure good measurement quality, survey questions should be clear, concise, and easily understandable for all respondents. One way of achieving this is to use plain language, an approach comprising a set of measures relating to the wording, sentence structure, and design of texts. Using a between-subjects design in a web survey, we tested whether a multi-item scale in plain language leads to better measurement quality while having similar scale properties to the original, standard language scale. Our results showed that the plain-language version led to greater response differentiation, fewer midpoint responses, and shorter response times than the original version, and that the two language versions had similar scale properties.