In this article we propose to relate the quality of writing with a wide range of linguistic and discursive features present in the introductions of the final works of the degree in Civil Engineering in Computer Science. For this purpose, an evaluation of 101 introductions is carried out, based on a rubric designed for this purpose. A correlation study is carried out between the evaluation of quality and 203 variables identified in the texts. Then, a linear regression analysis was carried out in order to identify predictive variables of the quality of the introductions. Among our findings we highlight that the length (average of words) presents a strong negative relationship with the quality of the introduction, in turn the use of conclusive connectors and syntactic complexity are positively correlated with the quality of the introductions. Also, the fulfillment of communicative purposes, in particular the indication of the gap, presents mean positive correlations with the quality of the writing. With the regression analysis it was possible to identify that the first three variables have significant statistical values to predict the quality of the introduction according to the average evaluation values assigned by the evaluators. These results are of interest for computational applications to support academic writing.
