VQR Scores Estimation Based on PCA Analysis

The “Research Quality Assessment” is a procedure promoted by ANVUR in the academic field, aimed at evaluating the paper quality in academics. Even if the main evaluation criteria are public, the final score of each paper is not directly obtained through an algorithm based on quantitative data, but can be affected by the peer review process and qualitative considerations. Thus, it is not possible to a priori know the outcome of each specific publication. In the present research activity, the VQR results of an Italian Department in the industrial engineering field were analyzed through a PCA approach. Additionally, a score estimation algorithm was developed, based on numerical evaluation parameters, which can quantitatively describe the paper’s quality. Even if the actual final score is not deterministic because of the peer review process, a strong correlation between the predicted and the actual scores was found in the analyzed data. The estimation for the specific paper can be faulty, nevertheless, the developed algorithm demonstrated consistency in terms of overall Department performance, especially for the higher scores.