Browsing by Subject "Iterative methods"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Minnesota Crash Records Audit(Center for Transportation Studies, University of Minnesota, 2020-08) Morris, Nichole L.; Libby, David A.; Peterson, Colleen; Ryan, Andrew; Sheppard, McKenzieIn 2016, Minnesota deployed a state-of-the-art electronic crash reporting system to all law enforcement officers across the state. The updated crash report was created with the officer in mind through an extensive usability and design project led by HumanFIRST. Preliminary testing of the new system suggested high user satisfaction but could not determine the extent of the crash data improvement. This study employed a mixed-methods approach to conduct a series of qualitative analyses of the crash data records collected by the legacy crash reporting system and the new MNCrash reporting system. The first analysis compared 360 serious injury and fatal crash reports from 2015 and 2016 and found a year-to-year decrease in mismatching data but a slight increase in missing data. However, overall, the MNCrash system increased the number of data queries, so data access was increased in 2016. The second analysis of serious and fatal injury crashes compared 996 reports sampled from 2015 and 1,572 reports sampled from 2016. The analysis showed a decrease in missing data and a decrease in erroneous data entry in 2016. Finally, the converted 2015 legacy data was analyzed to reveal few errors in the MNCrash conversion process.Item Minnesota MNCrash Design and Training Research Development(Center for Transportation Studies, University of Minnesota, 2020-09) Morris, Nichole L.; Schwieters, Katelyn R.; Craig, Curtis M.This report includes research activities that focused on the MNCrash interface design and training. A series of usability tests was conducted on the existing MNCrash interface system to document errors, frustrations, or confusion points that could be improved through iterative design and training. Usability testing with MNCrash users revealed that the majority of problems were related to detail and efficiency. Next, design recommendations were developed based on a set of criteria to reduce error and user frustration and to improve efficiency and user satisfaction. The outcome resulted in the implementation of several design change recommendations with a focus on addressing more accurate and complete data. To follow, a decision aid prototype was developed to determine if there were measurable effects of increasing accuracy of injury severity reporting for law enforcement participants. The prototype received high-level support, produced good usability, and increased accuracy in injury severity reporting. Finally, to complement the interface design recommendations, a training was developed to address knowledge gaps and improve accuracy in crash data reporting for law enforcement officers. The training was created on Rise360, an e-learning platform. Several iterations and user testing with law enforcement participants and crash reporting experts resulted in a final training design that consisted of an introductory module, eight core modules, eight quizzes, and a concluding module. The training produced good usability and user satisfaction recommended for implementation.