Monitoring emotional information is highly complex: it is difficult to accurately register it due to subjectivity and technical complexities; and it is difficult to provide reliability and incorporate contextual information. However, it is an important problem in healthcare, since it is useful to monitor people, especially if they are at a high risk of depression or other mental illnesses. Research in affective computing seeks to generate new methodologies to help store, analyze and share this information. Several techniques have been proposed to monitor emotions. One of them is self-report, which is a subjective method of measuring emotions from the perspective of the individual. This work presents a new tangible interface to self-report emotions, called EmoBall, specifically designed for people with low digital competences, since it requires practically no previous knowledge of technology. We evaluated this interface and analyze the results of the evaluation, considering the digital skills of the interviewed users. We found EmoBall to be a promising first step towards a tangible interface to self-report emotions; however, we did not find evidence of digital competences affecting user perceptions of the device. This paper discusses our insights regarding the reasons for these results, as well as directions for future research.