Abstract:
The CCI Method was mathematically and statistically tested and proved as an improved method to the current university selection method, ZScore. Education of Sri Lanka is entirely free and students are selected to limited capacity of the state universities by the GCE(A/L) examination. The CCI method is a simple and easily understood method which selects the top most students in each combination without distorting their earned raw marks at the examination. The raw marks contain values such as ability, efforts made by the student, parents and the society, and the raw marks uniquely determine the level of the student. Two data sets of raw marks for symmetric and more skewed were generated by means of an additive model taking the student intelligent effect and the error measurement as random, and the subject effect as fixed. The generated raw marks of three combinations in each data set were ranked according to the ZScore and CCI method. 3% of the student marks was assumed to be re-scrutinized and increased 5 – 20 marks randomly and ranked. The effect of the re-scrutiny was measured by way of the rank differences made by each selection method for original and re-scrutinized marks and studied descriptively and with Non-Parametric statistical procedures. It was concluded that CCI method is far more robust when compared to ZScore method for re-scrutinizing and it would help to speed up the University selection procedure at least by three months. By introducing CCI method for university selection, Sri Lankan GCE (A/L) examination system can be tuned to select the best set of students for human resource development
