Development of a Chatbot for Counseling to Reduce Stress Levels in Teenage

Abstract

This research aims to 1) develop a chatbot for providing counseling to reduce stress levels in teenagers, and 2) evaluate the effectiveness of the chatbot for counseling teenagers to reduce stress levels. The sample consists of 134 teenagers selected using accidental sampling. The research tools include a semi-structured interview, BOTNOI AI CHATBOT program, Line Bot Design, Google Sheet, and a questionnaire to assess the chatbot’s efficiency. The questionnaire's validity was determined using IOC, and its reliability was tested using Cronbach's alpha coefficient, with values ranging from 0.63-0.81, indicating high reliability. Data analysis was performed using mean and standard deviation statistics. The result showed that 1) For analysis and development’s chatbot system, the data from field interviews were analyzed and presented using the UML model. Additionally, the chatbot system includes three assessment system: stress level assessment, depression level assessment, and health level assessment. It also features a question-answer system, sending encouraging articles, and recommending treatment locations. 2) Evaluation of the chatbot’s efficiency in providing counseling to reduce stress levels in teenagers across three aspects revealed that the system’s convenience of use was at a high level (xത = 3.62, S.D. = 0.62), the system’s speed of use was at a high level (xത = 3.98, S.D. = 0.60), and the system’s ease of use was at a high level (xത = 4.12, S.D. = 0.62).