IE 2030. Project 1b, Cashier study, Validation

Due in lab, Feb 28 (Monday lab students); March 1 (Wednesday lab students)

This project should be done in groups of size 2 to 4, preferably the same groups as for project 1a. All students must be in the same section.

This is a continuation of project 1a.
1) Collect more data of the type needed for part (2) of project 1a. If you did not collect the right kind(s) of data for project 1a, this is a chance for you to make up part of that deficiency.

2) Compare your second batch of data with the data you had collected previously. There are many ways to do this. Your project is apt to be better if you only use multiple methods. Here are some possibilities:

  • compare means, variances, and other descriptive statistics
  • same as above but also compute confidence intervals
  • compare visual representations of the two data sets e.g. histograms, cumulative distributions.
  • fit the data from the first batch based on a hypothesis, or using best fit or other software; then compare the second batch with the fitted distribution. This could include a goodness-of-fit test. How well does the first batch data fit the distribution? How about the second batch data?
  • compare the second batch of data with the empirical distribution from the first batch, including a goodness-of-fit test if possible
  • use visualizations that directly compare the data
  • compute statistics that compare the data (e.g. based on ordered pairs)
  • think of a comparative method that is more specific to the type of data collected

    3) What can you learn from your comparisons? What insights or experience have you acquired? If you suspected that some of your data from the first batch would be atypical, because of the Superbowl or ice storm, what does your analysis indicate?

    NOTE: This project will overlap significantly with a project in 2028. You should be able to use the same data, and much of the same analysis, for both. However, you will have to prepare separate reports. Be aware that the focus of the project in 2028 may be a bit different.