Skip to content

My final exam for Design and Analysis of Sample Surveys

We had 28 class periods, so I wrote an exam with an approximate correspondence of one question per class. Rather than dumping the exam in your lap all at once, I’ll post the questions once per day. Then each day I’ll post the answer to yesterday’s questions. So it will be 29 days in all. I’ll post them to appear late in the day so as not to interfere with our main daily posts (which are currently backed up to early June).

The course was offered in the political science department and covered a mix of statistical and political topics. Followers of our recent discussion on test questions won’t be surprised to learn that some of the questions are ambiguous. This wasn’t on purpose. I tried my best, but good questions are hard to write.

Question 1 will appear tomorrow.


  1. Louis says:

    Great initiative, though I probably won’t be able to make all the questions.
    Could you post the syllabus in the beginning too? That way, the readers of the blog know what is coming and what materials were covered in the class.

  2. Kaiser says:

    I deliberately set open-ended questions, to mimic the real world where the analyst has to make some judgement calls. However, I find that students are extremely uncomfortable with the uncertainty of not knowing what the target is.

    • DK says:

      Open-ended questions are great when posed by good and competent teachers. They are a total disaster when used by dimwits with limited imagination – grading becomes a punishment. Most professors are not competent teachers. Students know it and correctly view open-ended questions with suspicion.

      • Kaiser says:

        Actually, it’s the opposite of what you say: open-ended questions are a self-imposed punishment by professors who set them. Professors who grade them have to read through a lot of text, some of which making little sense, while if we set multiple choice questions, the exams could be graded by a machine.
        Besides, I’d argue that all real problems are open-ended: if students are smart enough, they can argue that their answer is just as valid as the “answer key” for the non-open-ended questions. Unless the question does not test understanding but only test the ability to follow a fixed set of rules associated with a given methodology as described in lecture.

        • Andrew says:

          Or, to put it another way: It’s easy to write a test that’s hard to grade. It’s hard to write a test that’s easy to grade.

  3. DK says:

    What is expected background of the students? Are they supposed to know rudimentary statistics? Any math?

  4. […] Gelman has been posting one question per day from the (28-question) final exam for his course in Design and Analysis of Sample Surveys. Here’s Question 1. Here’s Question 2, and the solution to Question 1. By editing the […]