For class on Monday, please bring 5 copies of your one sheet and survey/experiment plan.
Grading of this project:
This project shall be graded by the following designations:
10% on complete one sheet, survey, revised based upon conversations on 11/19/2012
20% on correct collection of data
35% on the presentation of the data (what you bring to class December 13)
35% on the paper you turn in.
Each piece is outlined below for expectations.
One Sheet (due 11/19/2012, revised 11/26/2012):
Your one sheet should consist of the following information:
Goal: What is the desired goal of the project? What is the thing you want to investigate or measure?
Hypothesis: What do you believe the outcome of your project will be? what patterns do you think you will see in your data?
Cautions: What problems do you foresee in your data, data collection, experiment? What will you have to watch out for? What issues do you need to make sure you avoid?
Procedure: From the one sheet to the final project, you should have everything planned out.
Tests expected: What tests do you expect to do with your data? What tools make the most sense for the data you will be using? Be specific (e.g. "We expect that a 95% CI will be the best way for us to express the average feeling people have towards a Kanye West record. We will also use t.tests when comparing his albums to Beatles records.")
Tools (due 11/19/2012, revised 11/26/2012):
Your survey should be complete and ready OR all of the materials and participants should be ready for your experiment by this date. Make sure you have all forms printed and all pieces ready.
There will be a peer review section for this, which means that you will make changes they suggest. I will also look at each and make suggestions.
Data Collection:
You should have at least 50 data points if a survey, more or less depending on the type of experiment. Your data should be usable, and you should have a clear idea of what you can use compared to what you will have to discard and why.
Final Paper:
Things that must be in the paper:
Recap of the one sheet (think of a five paragraph essay--this is your introduction. "What is it we are trying to say")
Explanation of the written procedure compared to the actual procedure (think of this as a "what went wrong")
Your data
if you used a survey, include a copy (this may be included as a footnote)
if you used data from somewhere else, give the source as well as any discussion of how they collected the data. Make sure to find any faults with how the data has been collected.
make sure your data is attached in some form (.csv file, a table, whatever)
A comparison of the cautions to actual problems
Graphical information that shows your data.
A conclusion based upon the data and using mathematical tests we have learned in the classroom.
Things to consider: correlation, margin of error or confidence intervals, outliers, five points, population means, or any other important areas that work for your data.
Walk me through the important pieces and explain why you did the analysis that you did.
If necessary, include a null and alternative hypothesis.
Include all r code and graphs that you find that are relevant. As a suggestion use either end notes or foot notes for this piece of the information.
Changes that you would have made to the procedure if you were to do this again.
Questions that arose from the information that you collected.
Presentation:
Your visual in class can be a poster, an interactive online thing, an object, etc. etc. This is pretty open ended. It should, however:
At least one graph that makes the data look pretty (though perhaps not as useful as the graphs above)
this does not inherently have to be a 2d representation.
This part is the meat of the "Presentation of Statistics". The other pieces are to give some context.
Pieces of your one sheet, visible and easy to read/explain
Raw data, processed into a table or some other usable form
Graphs that depict the information in useful ways (at least 2)
Your data conclusions, as well as any relevant statistical tests
note: you can just include the results and not the math (for example, just give the df and the p-value)
YOUR PRESENTATION DOES NOT HAVE TO INCLUDE EVERYTHING FROM THE PAPER. IT CAN FOCUS ON THE INTERESTING BITS.