Abstract:
The following honours thesis compiles a dataset of reported concussions in the National Hockey League for the 2014-2015, 2015-2016, and 2016-2017 seasons. The main purpose of the thesis is data collection for a further statistical study using survival analysis. Each concussion event is observed on video, and the presence of 16 independent, situational factors is recorded. Relevant situational factors are determined through medical analysis of concussion, and interviews with former National Hockey League players. The length of each observed concussion is recorded to provide a measure of concussion severity. Concussion severity is divided into nine levels, and Chi
– square tests of independence are conducted to determine whether statistically significant relationships exist between the independent, situational factors, and concussion severity. The independent, situational factors of location of body contact, legality of body contact, and recurrent concussion are all determined as significantly related to concussion severity. Further experimental analysis reveals a potential relationship between puck possession and severe concussion. Hit velocity is then identified as a variable that interacts with the independent, situational factors to further impact concussion severity. An expansion of the honours thesis in collaboration with Atul Dar and Yigit Aydede is ultimately outlined.