Statement of Problem:
What we are going to do: We are going to develop a psychological survey mobile application for Android.
This application will pull user data and survey questions from a database. The user will then be able to
take this survey, which will be based around a variety of questions. After the user finishes answering the
questions, they will be presented with an overview of their answers. Their answers will then be sent to the database.
These answers will then be analyzed for possible psychological problems.
Why we are doing this: We want to aggregate massive amounts of psychological data about the general population at
KSU Marietta campus. We want to use this data to gauge the overall mental health of the campus. We also want to
provide people with feedback for self reflection and improvement.
The new features provided by our app:
Questions that involve vocal feedback
Statistical feedback that will allow users to make comparisons with themselves and the general population
Utilization of machine learning to create psychological evaluations of anonymous users
Mobile Features:
Microphone usage: Users will have the ability to provide vocal feedback using the device’s microphone for recording;
this data will not be fed into the machine learning algorithm or the foreseeable future.
GPS functionality: Users will also be able to optionally give their current location to the app to be stored into a database.
Database Queries: Questions that the user answer will also be pulled from a database. Answers will also be pushed back to that database.
Functionality and Overview
Basic functionality:
Users will be able to take a survey that is pulled down from a database
Each question will be on its own page.
The user will be able to overview their answers before submitting.
The results will be pushed back up to the database, when the user is done
The users will be able to use multiple choice, true false, and verbal questions.
Expected functionality:
A machine learning algorithm will analyze the survey results for possible psychological disorders.
Users will be able to look at results and predictions determined by the algorithm.
Users will be able to compare their answers anonymously with the averages of other users and see how many standard
deviations away from the norm they are.
An account will be required for pushing their results to the database.
The application will consist of multiple surveys, the user can choose which one to take.
Users will be able to choose from more than one survey.
Stretch goals:
A picture can be a response for some of the questions. The picture will be provided by what the users feels best answers the questions.
A Convolutional Neural Network(CNN) will be used to analyze these results, and predict future answers from the users.
Users will also be allowed to submit their own questions, which will be reviewed and added to the database if such questions are useful.
Users will also be allowed to submit their own questions, which will be reviewed and added to the database if such questions are useful.