Abstract: Whether to enter the ESI subject index is usually used to evaluate the academic research level of a certain scientific research institution. In order to evaluate and predict the subject competitiveness of scientific research institutions, and to provide decision support for scientific research management departments, the paper proposed the leading variable index RFV based on the standardized value of the influence of the reference journals produced by the institution and builds linear Regression model to predict the citation frequency of disciplines at the institution level. The histogram, scatter plot and P-P plot were used to verify the validity of the normal distribution and the model, and the Pearson correlation coefficient was used to analyze the correlation between the two. The empirical research results show that there is a statistically strong correlation between the institution’s citation frequency TCo and its RFVo, which verifies the feasibility and superiority of the model.

Key words: Citing-side variable, Times cited, Pearson correlation coefficient, Citation analysis, Linear regression