Imperical studies on the use of Principal Component Analysis in determining principal factors that influences financial sustainability in Higher Education
I like the paper what Hamit uploaded. It is a top-down approach from the categories perspective.
I think you can apply a bottom-up approach what provide more freedom and openness, also more customized to your area. First I would collect the important indicators from the stakeholders or a group of experts. Then, I would reduce a list of 10-40 and conduct a survey about their importance in financial sustainability among the stakeholders. I would upload all the responses to a statistical software (e.g. SPSS) and would do a PCA. The main components could be identified by the incorporated indicators.
Example below. Article Public perception of bioenergy in North Carolina and Tennessee
Justice,
Here is my suggestion -
1) PCA is used to reduce the original number of your indicators.
2) Develop a benchmark indicator for first screening by correlating it with other indicators.
3) You may do second screening of indicators using component matrix to interpret results and refine the model
4) The quality of the data used for the PCA is inspected using the Kaiser-Meyer-Olkin (KMO) measure and the Bartlett Sphericity test.
5) The variances in the PCA explained by each component are called eigenvalues. Estimate the number of factors with initial eigenvalues exceeding one (eigenvalue ≥1).
6) Determine variances in the indicators and proceed to build a model or index specifically relevant to your topic).
(see copy of attached file)
- Badges
- Science topic