Visual Analysis of Multi-Source College Students’ Mental Health Questionnaire Data

A variety of mental health questionnaires are widely used for the diagnosis and treatment of potential mental disorder and illness of college students,such as University Personality Inventory(UPI)and Symptom Checklist90(SCL-90).Different questionnaires get insights into mental health status from different perspectives,the results of which easily conflict with each other,making it challenging to conduct comprehensive investigation,collaborative analysis and precise treatment for mental health of college students.In this paper,we propose a visual analytics system to achieve the correlation of different categories of questionnaires and reduce the uncertainty of mental health analysis.Firstly,a Circos view is designed to visualize the answers and associations of multi-source questionnaires for individuals.Then,a parallel coordinate system is employed to present the distribution questionnaires and the correlation of questionnaires is further evaluated to optimize the arrangement of coordinate axes.According to the answer distribution,a dimensionality reduction method is utilized to present the dissimilarity of students across different questionnaires.The uncertainty generated from multi-source questionnaire analysis is evaluated based on the geometric difference in the low-dimensional space,which is further highlighted through color mapping,enabling users to easily focus on those students with uncertain analysis results.Finally,we implement a visual analytics framework for multi-source questionnaire exploration,with the above-mentioned questionnaire designs and analysis models integrated.A rich set of user interactions are further provided,allowing users to achieve comprehensive judgment and tracking analysis of mental health status of college students.The effectiveness and practicability are demonstrated through case studies with real-world multi-source questionnaire datasets and the feedback of domain experts.