to predict the national outcome, you need to have data first of all.
I am not familiar with NGAT but I think it is a test so predicting the national outcome happens by:
1) collecting the results of a few previous years and creating a dataset.
2) build a model to predict it in the future.
the type of the model depends on what you want to predict, for example:
- if you want to understand the behaviour and the progress of the test results in general for all candidates, you can use a regression model. and feed it the whole data processed and cleaned, and get a line or a curve that will help you understand where the nation is heading and what to expect from all the participants in the future.
- if you want to understand how an individual candidate might perform in the test, you need to collect features and include them in the dataset and not only the results, so for example for each candidate, you record the age, place of examination, number of exam trials, and the exam result for each trial, and other features as many relevant ones as you can find, then by feeding these features into a classification model to train it on, you can use the model later by giving it the information of the candidate that you used as features before and the model will predict the result for that single candidate.
I hope this answer helps, and please if you find ambiguity in it or looking for more details do not hesitate to ask, I am always happy to help.