Preference learning and Homework 2

by Max Yi Ren

See the lecture notes, the Qualtrics tutorial, and this engineer-friendly tutorial (Part 1, Part 2) on discrete choice analysis.

Activity 1: Test the survey

Please participate in a mock survey regarding your preference on cars. Once we collect the data, we will go through a data analysis procedure called discrete choice analysis, from where quantified preference on attribute levels can be revealed for the participant group as a whole and for individual participants.

Homework 2: Conduct a market survey and report your analysis results

The report should contain the following elements:

  1. A description of your survey goal, i.e., what do you want to learn from your consumer group? This could include qualitative goals, and quantitative hypotheses.

  2. A description of design attributes and attribute levels

  3. A complete sample survey and a discussion on how questions are designed

  4. A description of how data is collected
    • Who do you collect data from?
    • When is the data collected?
    • Are your collected data valid? How do you validate?
  5. A description of how the data is analyzed
    • What preprocess steps are necessary for your data?
    • For discrete choice analysis, how do you set up your model and what assumptions do you use?
    • What results do you get from your analysis?
    • What insights do you get from your analysis?
    • From these insights, how will you improve your survey?
    • Please provide a visualization of your results. Simply pasting MATLAB tables to your report will not be accepted.
  6. The code and raw data you used for the analysis
    • Make sure that by directly running your code with your data, others can reproduce your results.

Notes: