Date of Award

Spring 2019

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

School

Education

Committee Chair

Richard Mohn

Committee Chair School

Education

Committee Member 2

Thomas O'Brien

Committee Member 2 School

Education

Committee Member 3

Kenneth Thompson

Committee Member 3 School

Education

Committee Member 4

Mark Walsh

Abstract

The prevalence of noncommunicable diseases, due to a lack of physical activity, is a significant health issue. Physical educators and coaches have an important role in the prevention of such diseases. Instilling a physical lifestyle at a young age, increasing the likelihood of a having physically active life in the future, is a preventative measure. Thus, physical educators and coaches must help students to become familiar with a variety of physical activities in which they can continue to participate in throughout their life, such as the sport of powerlifting.

Research suggests that for an individual to stay interested in an activity they must be able to set and reach performance goals. This is because goal achievement leads to self-satisfaction which encourages continued participation in the activity. Locke’s Goal Setting Theory outlines the steps in goal setting and achievement. Of particular interest, Locke’s theory describes the importance of feedback information to aid in an individuals progression towards meeting one’s goals in particularly complex tasks, such as powerlifting. Thus, the purpose of the study is to create a transfer of training model for powerlifting. The model can be used in aiding in providing feedback and strategizing how to meet one’s powerlifting goals. This study can be used to encourage continual involvement in the sport of powerlifting and to help decrease the prevalence of noncommunicable diseases.

Unlike the transfer of training models used in other sports, this study used web scraped data to create a Bayesian network. This approach allowed for a larger population of individuals to be included in the study while allowing the structure of the network to be learned based on conditional independencies and dependencies. This allowed for highly correlated data to be efficiently networked, as well as undiscovered relationships to be exposed.

Future research can use this work as a base for incorporating new data and variables into the model. Additionally, the learned structure can be used to help create other forms of Bayesian networks such as Dynamic and Hierarchical Bayesian nets.

Available for download on Wednesday, May 10, 2169

Share

COinS