Premium
Prediction of cadet basic training physical performance outcomes
Author(s) -
Ploch Christopher,
Thomas Diana M,
Adams Bryan,
Turner Dusty,
Hug Adam,
Suby Jason,
Palmer John E.,
Bigelman Kevin
Publication year - 2019
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2019.33.1_supplement.693.17
Subject(s) - cadet , mile , multi stage fitness test , test (biology) , psychology , physical fitness , basketball , cluster (spacecraft) , physical therapy , computer science , medicine , geography , paleontology , archaeology , geodesy , biology , programming language
Background Entry to the United States Military Academy requires submission of scores on the Cadet Fitness Assessment. The cadet fitness assessment consists of push‐ups, crunches, kneeling basketball throw, shuttle run, and 1‐mile run. Upon arrival for Cadet Basic Training, cadets are assessed on the components of the Army Physical Fitness Test; push‐ups, sit‐ups and a 2‐mile run. Additionally all cadets are assessed on a swim test. Records of stress‐fractures incurred during Cadet Basic Training are also retained. Objective To apply supervised and unsupervised machine learning to describe Cadet Basic Training and Cadet Fitness Assessment data compiled from the graduating classes of 2012 to 2021. Methods Simple data‐analysis plots, linear regression models, and unsupervised k‐means cluster analysis were developed to predict Cadet Basic Training performance. Models were generated to identify the relationship between body mass index (BMI), age and gender on 2‐mile run performance to replicate past Army results. Results A total N=8886 (82% male) cadets were retained for analysis. The relationship between 2‐mile run performance and BMI was weak (R 2 =0.07), with age, BMI, and gender representing significant predictors (p<0.01). Similar results were found for all performance outcomes except pull‐ups. The relationship between pull‐up performance and BMI was stronger (R2=0.47). The cluster analysis revealed four clusters. One cluster was identified to poor 2‐mile run performance while a second cluster was identified with poor swim, push‐up, and sit‐up performance. Scatterplots did not indicate any relationship between cluster assignments and BMI. Conclusions BMI and even waist circumference may be poor predictors of performance on military physical fitness test outcomes. Machine learning could offer more precision for identifying poor performers early and provide novel opportunities for remediation. Support or Funding Information None This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .