This project allows you to test hypotheses about the relationship of two variables using SMS based on their shared frequency. A common use case is health symptoms that you may suspect are being caused by something.

• OS: Ubuntu 14.04 LTS
• language: Python 3.4.3
• application server: Flask 1.0, Jinja2 2.9.6
• database: MySQL Ver 14.14 Distrib 5.7.18

This project comes with various setup scripts to support automation, especially during maintanence or to scale the entire project. The following files are the setupfiles along with a brief explanation:

• Enrollment: Initial enrollment happens when you text to the assigned Twilio number where you will be asked if you want to sign up as well as being asked your name. A link will be generated with an access code that will redirect you to the website

  • Access code: The access code needs to match the number it was paired with and doesn't have an expiration date.

• Registration: Once at the website you will be prompted to choose a password so that you can complete the registration

  • Password: The password needs to be at least 8 characters long

• Variables set up: In this form you will set up the predictor and outcome variables as well as a session duration. The words that you choose are the same words that the system will expect you to text to match them up with the right variables.

  • Session: The session duration determines how long you want to wait in hours before a predictor variable cannot be paired with an outcome variable. The session duration cannot be less than one hour.

• Data collection phase: Once you've completed the steps above you are ready to collect data. Begin texting the variable words that you picked.

  • Errors: If you sent a word that is not linked to one of your variables, you will get an text telling you it is an error. This data will not be saved. If you text a predictor when an outcome is expected (e.g., when a session is still open), you will also get an error message.

• aucontraire

MIT License