Improving Satisfaction through the Analysis of Decision-Making Criteria of Meteorological Information User Kim In-Gyum1, Kim Hye-Min1, Lim Byunghwan2 1Researcher, Future Strategy Research Team, National Institute of Meteorological Sciences, KMA, Korea 2Deputy Director, Observation and Forecast Research Division, National Institute of Meteorological Sciences, KMA, Korea Online published on 4 June, 2019. Abstract It is important to understand how forecast information is recognized and used in order to effectively facilitate service delivery to the meteorological community. This study was conducted to quantitatively estimate users’ dissatisfaction level with forecasting errors and to evaluate whether the distribution of user's decision criteria of probabilistic forecasts is appropriate. Correlation analysis was conducted between the satisfaction value of forecasts and the user-perceived accuracy to estimate the level of dissatisfaction regarding inaccurate forecasts. Further, satisfaction value was compared between an actual and a virtual group to investigate how forecast users are effectively utilizing the information. The analysis showed that the users who responded to the questionnaire generally did not recognize the effective probability threshold as being able to provide a better value. These users were estimated to be very dissatisfied when adverse weather was not predicted in advance. Meanwhile, the value of the forecast to younger users (19–39 years)can be increased by adjusting their current user threshold. However, those over 60 years did not benefit from this adjustment. To improve user satisfaction with forecasts, it is deemed necessary to educate users regarding the meaning of probabilistic forecasts and the probability threshold pattern based on age group. By quantitatively analyzing user dissatisfaction, it is possible to estimate which forecast elements and forecasting errors garner user dissatisfaction. Through such analysis, moreover, we can help the meteorological community to develop better services by considering the characteristics of the user group. Top Keywords Perceived accuracy, probability threshold, value score, 22 model, satisfaction. Top |