Of the multitude of survey instruments used for assessing mental health conditions, most are proprietary, expensive, redundant, onerous, diagnosis-centric, in paper form and static in the set of questions and in their scoring, based on dated data sets derived from small populations. These surveys are usually filled out in a clinic, which by itself means that the vast majority of people do not have access to them or to any insights derived from filling them out. Why can’t mental health assessment be engaging, dynamic, informative, and more inclusive, like online tools for building a personality profile?
To address this problem, we propose to build a mental health assessment framework called “My Mind Matters Quiz” (M3Q) that anyone will be able to use to build a personalized mental health profile through a web browser or a mobile app. The M3Q will be free, open source, online, adaptive, and engaging enough to entice users to come back and learn more about themselves and about other people. The Child Mind Institute (CMI) has experience creating questionnaires and administering them to, for example, 10,000 children as part of CMI’s ongoing Healthy Brain Network (HBN) study. The M3Q’s database currently includes symptoms and diagnostic categories from psychiatric statistical manuals such as the DSM-V, thousands of questions from over 80 mental health questionnaires, and information about mental health resources. We will use this database to test whether generating questions from symptoms in a principled bottom-up manner will lead to more informative and helpful answers.
Most mental health questionnaires contain ambiguous questions pertaining to a limited set of symptoms. Moreover, these questions are often too extreme to allow for gradations in response from the population at large. To generate a more rigorous and inclusive set of questions, we will:
We will establish a framework for easy, intuitive and efficient navigation of questions. To determine an efficient, personalized path through the many possible questions, we will model the question space to encode the distribution of high- to low-predictor questions, and optimize efficient traversal and coverage of the mental health space. The user will be encouraged to complete their profile with game-like feedback and rewards.
The above may constitute the first framework to not only generate mental health assessment questions, but also to crowdsource their evaluation. Creating an empirically-driven, dynamically updated assessment instrument was the original intent of the DSM-V, and this project would realize that goal, to the benefit of all mental health stakeholders.