stat 332

Homework 3

This homework assignment has two stages with two due dates. The first stage is due Oct 8, the second stage is due Oct 15.
Please submit assignments electronically to all three instructors at heike, dicook, chkostel@iastate.edu.

User Study

Given two graphical displays, we want to answer the question, which one is better.  One way, among others, to answer that question is to do an empirical study of the display's users.

  1. first stage: design the study
    In the first stage of this homework, you will need to do several steps:
    • identify a hypothesis of interest
      This hypothesis can be anything, as long as it is related to some aspect of data visualization.
    • create two charts that capture the hypothesis
      These two charts need to exhibit aspects of your hypothesis, e.g. with the pie and barchart discussion, one chart would need to be a barchart, one would need to be a pie chart.
    • design a user study
      This needs to discuss, how you will find subjects for your study (be specific, e.g. if you plan on asking people on Campus think about where and when ...).
      What question(s) you will ask them, what answers/demographics you will record and how you will record them.
  2. second stage: perform the study
    Keeping to your sampling scheme, perform the study and collect answers.
    Evaluate the data to answer your initial hypothesis. Make sure to check that other demographics to don't exhibit major differences.

Deliverables

For the first stage, hand in a paragraph with your hypothesis, a description of the study and a copy of the survey/study.
For the second stage you will need to hand in the data (electronically), and a 1-2 page write-up summarizing results of the study, including an answer of the original hypothesis (if the results are inconclusive, that's an answer as well), and a critical assessment evaluation of the study, i.e. what worked well, what (unexpected) problems you encountered, and what you would differently next time.

The best answers will be published anonymously on this website.