lanalytics

Analysis of student data from quizzes

Files:

1. parsequizzes.R

• First data parsing step - this script is designed to be run exaclty once at the beginning of the data analysis process
• Takes xlsx file for each quiz as downloaded from Learning Catalytics
• Assigns each student a unique random identifier
• Creates 3 tables, all of which have one row per student ID, and one column per
  quiz question. Data in those tables are
  • Absolute time of answer
  • Content of answer
  • Correctness of answer (1 or 0) All three tables are saved into xlsx files
• Creates an additional (confidential) table of student IDs and names, to be stored in a safe place

2. order_time.R

• Depends on parsequizzes.R
• Takes absolute time stamp data as computed using parsequizzes.R
• For each quiz and each student, computes
  • the order in which questions were answered
  • the time it took to answer each question

3. easiness_time.R

• Depends on order_time.R, parsequizzes.R
• Takes time per question and correctness information
• For each quiz question, compute
  • item easiness (% of sdtudents who g)
  • time to answeer item (median across all students)
• scatter plots easiness vs time

4. guessers.R

• TODO: This function is still buggy and incomplete
• Depends on order_time.R, parsequizzes.R
• Assumes that the first two and last two questions of each quiz are easy (in our case, because they are about study style and preparaation)
• Uses time for those easy question as a threshold (i.e. fastest possible)
• Computes questions answered below this threshold and classifies them as
  • cheating, if the answer was correct
  • guessing, if the answer was wrong
• Plots overall occurence of guessing and cheating across quizzes and students