Class Fifteen: Let's Build Better Surveys
Group work
Select one of the four surveys.
SURVEYOR:
describe briefly what you are attempting to do with the survey
discuss what you plan to do with the results
what hypotheses do you have?
what graphs can you make
what information is necessary
TAKERS OF SURVEY:
take the survey.
outline any issues or questions you have about the survey
address any problems with the data
We will choose a few if we really need to work through them as a class.
IF WE HAVE TIME
probabilities of the same birthday
probabilities of rolling a 1
probabilities of rolling boxcars
----
Homework:
By Tuesday at noon, have your survey completed and edited. Make sure to clear out any old information from previous people taking it.
By class, have taken all 17 surveys (including your own).
Read this paper
Perspective: I came across this paper while looking for bad graphs to turn good. It's an interesting twist on what I've been saying thus far; in it, they argue that the bad graphs are good for remembering or assigning value to data. But I think there are some specific flaws in the way that they did this paper.
Goal: Read the paper and do the following:
Look for statistical numbers, ideas and uses directly in the paper. For example, can you figure out what t_19 and t_9 are used for in the paper?
What methods are unknown or strange to you?
In methodology or other aspects, are there things you agree with or disagree with?
You should be able to discuss this paper as a group or individually on Thursday.
Resources: I've found a thing that might be helpful to you. Read and think about the paper on your own before you read this:
https://www.perceptualedge.com/articles/visual_business_intelligence/the_chartjunk_debate.pdf