Research Supervisor Connect
A retaining military personal model will be developed using Partial Least Squares Path Modelling (PLSPM) techniques and other statistical modelling techniques. PLS-PM is a well-known statistical technique which is based on using structural equation modelling, in determining the variables and their contribution to the changes that may occur in a specific topic (for example, the turnover).
The developed model will be able to measure the influential variables and their contribution to the turnover of the military personal. The organisational commitment, such as incentives in salary, home loan (or similar) can be implemented as ‘what if’ scenarios. The selection of what type of incentives to introduce will be based on the contribution of each variable to the turnover. The model’s adequacy will be tested using the popular statistical method of bootstrap and its robustness will be tested via different scenarios.
The most important initial step is to determine the variables that the Department of Defence believes are important to be included. This may be achieved through team collaboration with the Workforce Planning Branch at the Department of Defence.
School of Mathematics and Statistics
In collaboration with the Department of Defence, the project will provide professional analysis of the current military attrition rate and suggest the construction of an index that measures military turnover drivers and their contribution to turnover using PLS-PM techniques. Then a sub model, an economical, will be appended with the PLS-PM model to calculate the loss on drivers. The input data to the model (and the sub model) will be via two different Excel sheets. The input data can be collected from previous surveys and any other resources within the WPB. The most important initial step is to determine the variables that the DoD want to be implement. As the number of variables increase the complexity of the model and its maintenance will follow. WPB may consider for example, high and low cost in occupations, trainings, ranks, … etc. to reduce the number of variables. We are intending to apply machine learning tools to reduce the manual work and model’s maintenance.
This section may contain information about:
- other research project opportunities you may have listed
- associated scholarship/funding opportunities – link to scholarship url/advert
- specifics about the research location e.g. Camden, Narrabri, One Tree Island, ATP etc.
- Inherent requirements of the project e.g. current vaccination, current Australian driver’s license, scuba diving license etc.
Project keywords: Bootstrap, PLS-PM, Structure equation modelling, Turnover Index
Addtional supervisor(s): Shelton Peiris
The opportunity ID for this research opportunity is 3549