The role of linguistic features in science and math comprehension and performance: A systematic review and desiderata for future research

There is a growing body of research on the role of linguistic features (LF) in comprehension and performance in science and math. Of particular interest is which LF facilitate or hinder comprehension. In this systematic review, we provide an overview of findings on LF at the word, sentence and text levels, on comprehension and performance in science and math. Our literature search revealed n = 40 articles included in this review. Overall, the role of LF in comprehension and performance in science and math is complex, with findings varying across the different LF. For each LF, we discuss the findings and uncover remaining questions. In the general discussion, we uncover strengths and weaknesses of previous research and discuss open questions before making eight recommendations and identifying tasks for future research aiming to understand the complex nature of LF.     

What makes or breaks competitive research proposals? A mixed-methods analysis of research grant evaluation reports

The evaluation of grant proposals is an essential aspect of competitive research funding. Funding bodies and agencies rely in many instances on external peer reviewers for grant assessment. Most of the research available is about quantitative aspects of this assessment, and there is little evidence from qualitative studies. We used a combination of machine learning and qualitative analysis methods to analyse the reviewers' comments in evaluation reports from 3667 grant applications to the Initial Training Networks (ITN) of the Marie Curie Actions under the Seventh Framework Programme (FP7). Our results show that the reviewers' comments for each evaluation criterion were aligned with the Action's prespecified criteria and that the evaluation outcome was more influenced by the proposals’ weaknesses than by their strengths.     

Fair and Explainable Depression Detection in Social Media

Detection at an early stage is vital for the diagnosis of the majority of critical illnesses and is the same for identifying people suffering from depression. Nowadays, a number of researches have been done successfully to identify depressed persons based on their social media postings. However, an unexpected bias has been observed in these studies, which can be due to various factors like unequal data distribution. In this paper, the imbalance found in terms of participation in the various age groups and demographics is normalized using the one-shot decision approach. Further, we present an ensemble model combining SVM and KNN with the intrinsic explainability in conjunction with noisy label correction approaches, offering an innovative solution to the problem of distinguishing between depression symptoms and suicidal ideas. We achieved a final classification accuracy of 98.05%, with the proposed ensemble model ensuring that the data classification is not biased in any manner.