## Abstract

Structural equation modelling (SEM) is analogous to multiple regression but is a more powerful statistical method as it can evaluate the modelling of more complex interactions, nonlinearities, correlated independent variables, measurement error, correlated and uncorrelated error terms and multiple latent independents or factors, each measured by multiple indicators (1, 2). SEM may be used as a more powerful alternative to correlation, multiple regression, path analysis, factor analysis, time series analysis, and analysis of covariance.

Research has established some of the relationships between strength, power and speed (3, 4, 5, 6), usually the measures of muscular power being more predictive of sprint acceleration and sprint speed. Some multivariate research utilising the multivariate statistical technique of factor analysis have indicated peak torque-strength and power constructs share some common variance but a significant amount of the variance is unique to each construct or factor (7). Motor fitness assessment tests are readily available and provide differences between genders in terms of strength, power and speed and apply norm referenced assessments; as well as IAAF World Athletic Rankings for the genders competing in identical athletic events.

However, research that has examined relationships of strength, torque and power with sprint acceleration and sprint speed have relied on statistical correlation and regression, which are predominantly univariate measures of analysis and regression analysis is based on a direct effect statistical model. These statistical approaches do not consider mediating variables that can be analysed by quantifying indirect statistical effects using structural equation modelling and more complex path analysis.

The two research aims of this investigation were to:

1. Evaluate using structural equation modelling and path analysis the more complex relationships between the constructs of strength-torque, power and sprint acceleration in male and female athletes.2. Evaluating if there are differences between males and females by comparing the derived path analyses using identical conceptual path diagrams for strength-torque, power and sprint acceleration.

Research has established some of the relationships between strength, power and speed (3, 4, 5, 6), usually the measures of muscular power being more predictive of sprint acceleration and sprint speed. Some multivariate research utilising the multivariate statistical technique of factor analysis have indicated peak torque-strength and power constructs share some common variance but a significant amount of the variance is unique to each construct or factor (7). Motor fitness assessment tests are readily available and provide differences between genders in terms of strength, power and speed and apply norm referenced assessments; as well as IAAF World Athletic Rankings for the genders competing in identical athletic events.

However, research that has examined relationships of strength, torque and power with sprint acceleration and sprint speed have relied on statistical correlation and regression, which are predominantly univariate measures of analysis and regression analysis is based on a direct effect statistical model. These statistical approaches do not consider mediating variables that can be analysed by quantifying indirect statistical effects using structural equation modelling and more complex path analysis.

The two research aims of this investigation were to:

1. Evaluate using structural equation modelling and path analysis the more complex relationships between the constructs of strength-torque, power and sprint acceleration in male and female athletes.2. Evaluating if there are differences between males and females by comparing the derived path analyses using identical conceptual path diagrams for strength-torque, power and sprint acceleration.

Original language | English |
---|---|

Pages (from-to) | 167-169 |

Number of pages | 3 |

Journal | Journal of Australian Strength and Conditioning |

Volume | 22 |

Issue number | 5 |

Publication status | Published - 2014 |