Conversational Interface for Screening

Abstract

There are many adults who lives with ADHD without getting a diagnosis. When being evaluated for ADHD the first step is often to complete what is called the Adult ADHD Self-Report Scale (ASRS). ASRS is a symptom-check questionnaire built by the World Health Organization for screening adults for symptoms of ADHD. In the study presented in this thesis, a prototype for a chatbot has been designed in order to explore how the ASRS test could be designed to a conversational interface. Having the ASRS in a conversational interface, users can answer questions from the ASRS with a more open language and supply answers with information that may be of interest for domain experts. The prototype was evaluated amongst users by conducting a comparative experiment with two objectives. To find out how the results from the conversational interface differed from the results from the paper-based modality, and to find out how the participants perceived the prototype. The results from the experiment revealed an indication that the result differences were of non-significant and that most participants preferred the conversational interface to the paper-based modality. The results support that chatbots can be a useful technological utility for screening in the domain of mental health.