Data Collection Flow

Production

0. requisites

   • download Heroku CLI
   • login to Heroku by running heroku login from project repository

1. add initial data w/o quizzes & attempts & student acc: heroku run python main.py initial-data --no-quizzes=1 --secure-first-users=1 --app inquizitor-prod

   • NOTE: this resets the database and should only be ran once

2. add mock exam: heroku run python main.py add-exam mock-exam.txt --has-attempts=0 --blanks-case-sensitive=0 --app inquizitor-prod

3. create accounts from exported responses file:

   • clone this repository
   • add csv files inside the clone repository
   • for the cheating group, run python main.py create-accounts cheaters.csv inquizitor-prod --is-cheater-dataset=1
   • for the non-cheating group, run python main.py create-accounts noncheaters.csv inquizitor-prod --is-cheater-dataset=0
   • alternatively, generating accounts directly through the database url would be faster: main.py create-accounts-on-db respondents.csv DB_URL

4. update cheater dataset:

   • main.py toggle-cheater-dataset test.txt <IS_CHEATER_DATASET> <DB_URL>

---

Quick Testing

initial data w/o attempts: heroku run python main.py initial-data --use-realistic-data=1 --has-attempts=0 --app inquizitor-test

mock exam: heroku run python main.py add-exam mock-exam.txt --has-attempts=0 --blanks-case-sensitive=0 --app inquizitor-test

accounts: python main.py create-accounts team.csv inquizitor-test

test case:

• parallel/ concurrent quiz answering
• web app check up

---

QUIZ APP BRIEFING

• make sure form has been answered
• access to application
  • check email for account credentials
  • use chrome
• nature of quiz app & cheating detection (? discuss quiz actions?)
• instructions for answering quiz
  • for cheating group:
    • incentives to get high score thru cheating means
    • without using other device (?)
  • for non-cheating group:
    • incentives to not cheat and get over the quiz peacefully, hassle-free