Part One: Find a survey that was conducted. Locally, statewide, nationally, internationally, whichever, but the survey you choose must have actually been conducted. A good
starting point would be looking at the Census Bureau. I want you to write a 1 page essay
critiquing this survey. I want you to comment on things such as
sample size used
question ordering
question wording
would changing either of the above have led to clearer understanding of the question?
how was the survey conducted (phone interviews, in person interviews, mail in, etc)
discuss why or why not this was the best method to conduct this survey.
what was the intended population
did they have a sampling frame when conducting this survey?
what are potential errors that would crop up? Are the errors observational or nonobservational?
Could they have reduced the errors in this survey? How?
Did they use a sampling scheme that we are familiar with (simple random, stratified random, cluster, ratio/difference estimation)? Was this scheme the best
choice? Why?
What are the results of the survey? What parameters are the researchers estimating?
How big is the bound on the error of estimation? Could you have improved it? How?
Make sure to include a copy of the survey chosen so that I can see what you are commenting
about. If the survey is too large simply include a link at the end of your essay.
Part Two: I want you to identify a set of 3-4 research questions that you would like
answered if you had the resources to conduct a survey. I want you to design a single survey
that will answer these questions. You should identify things such as:
target population
sample units and sampling frame
how accurate you would like your estimates to be
appropriate sample size based on how accurate you want to be
how the survey would be conducted
sampling scheme
Then I want you to actually write out a survey that will get you the answers you seek if
you had the resources to distribute it and collect the data.
Note that the survey you write
should probably be longer than just the 3-4 research questions you are interested in. You
should pay attention to things such as
question order
question wording
any potential biases that may crop up based on the above
The post Title: Critique and Design of a Survey: Understanding Methodology and Improving Accuracy appeared first on academic aid express.
