Weekly report 5 | statistics for research methods II
The given data is suitable for the application of linear discriminant analysis, as it contains two discrete groupings (admission/non-admission) and several numerical independent variables. This allows for the creation of a linear discriminant function which can predict admission decisions on the basis of GRE scores, GPA and other data. Analyzing the data allows us to determine how accurate our model is at predicting admissions. We would expect a 95% accuracy rate if 100 observations are analyzed and 95 of those are classified accurately according to the model.
Linear discriminant analysis can produce a linear discriminant function to predict admission decisions. This is possible using GRE score, GPA and other input variables. You can make more complex models with logistic regression analysis by including additional input variables like work experience and extracurricular activities. This could improve the classification accuracy and predictive power.
One can expect an admission decision when one has a GRE score of 690 and a GPA equal to 3.2. This is based upon our linear discriminant classification model. However, the final classification accuracy depends on both methods’ length data. Therefore we will not know until running their respective analyses.