4.2.1 Context data

Context data correspond to all information which doesn’t change during the study but helps explain the observations or document their values. They may be directly collected, generated for the purpose of the experiment, already exist, or retrieved from external data sources.

They contain, for instance, background information – such as infrastructure characteristics (e.g., map data) and vehicle /driver characteristics and roles during automated driving, including questionnaire results.

Questionnaires collect qualitative and quantitative data reported by each individual participant. They typically cover basic data, such as age, gender, and general attitudes about driving. They can also cover more specific aspects, such as personality traits (e.g., sensation-seeking, introverted). Quantitative data is obtained by means of closed questions (e.g., multiple choice, scales) whereas qualitative data is obtained when specific questions are open for rich text information, often of a more interpretive nature. There is a grey zone where some elements may be considered ‘contextual’ data, such as participants’ characteristics or weather, traffic and driving conditions, and thus part of the dataset, whereas in other datasets this may be considered metadata.