Statistic question

The Assignment

• Simulate data for which assumptions of the t-test are met
• Simulate data for which assumptions of the t-test are not met
• Simulate data with outliers
• Analyze all three types of data mentioned above with both the t-test and binomial test
• Calculate the Type I error in each situation
• Which test is more resistant and/or robust, at least in terms of Type I error?

Submission Guidelines

• The results should be typed like a paper, with clearly defined sections for Introduction, Methods, Results, and Conclusions.
• Your name must be at the top of the first page and on each successive page.
• There should be an informative title ("Simulation Challenge 1" is not informative).
• If you use MS Word to type homework, use an easy to read variable-width font (I like Ariel, Helvetica, and Geneva fonts) with a minimum of 12 point font.
• Tables and graphs should be included in the body of the paper, as close as possible to the text that describes them.
• Any graphics from R must be electronically cut-and-pasted in-line at the appropriate point of the write-up. Resize the graphics window to optimize the readability. I will show you how to do this in class.
• Any mathematical notation must be provided with appropriate use of subscripts, superscripts and symbols. Use MS Equation or another equation editor if you submit your work in Word.
• Captions for tables and graphs should be contain a description of the variable on the x-axis, a description of the variable on the y-axis, a description of any plotting symbols, and a brief summary of the point that the graph or table is meant to convey.
• Code for the simulation challenge should be uploaded IN A SEPARATE FILE in plain text format. The code should be working code (no mistakes). Part of your grade is determined by how easily I can copy and paste your code from the text file and reproduce your results.
• Maximum of 7 pages, including tables and graphs. There is no minimum. The code file is not included in the page count.