You must type this and all other homework assignments. Do not e-mail the assignment to me; turn it in early (at 322 Wolf) for a foreseeable absence, or turn it in late after an unexpected absence from class.
1. Collect data suitable for linear regression/correlation. You must collect the data yourself, by measuring something, counting something, or surveying people. You must collect biological data--data on living or dead organisms of some kind (and no, M&Ms don't count). You must have at least 20 observations.
For your answer to question 1, just describe what you measured; you'll include the data in your SAS program.
2. Draw a computer-generated graph of your data.
3. Based on looking at the graph, do you think linear regression/correlation, polynomial regression, or Spearman's rank correlation, would be most appropriate? Explain why you think that.
4. Analyze your data using linear regression/correlation. Report the P-value and the r2. Use both my spreadsheet and SAS. Print your SAS program. You only need to print the log file and output if the SAS results don't agree with my spreadsheet.
5. Analyze your data using polynomial regression, using my spreadsheet. Report the P-value and r2 for the linear, quadratic, and cubic equations. You don't need to do the quartic or quintic, they're stupid. Also report the P-values for the increase in r2 between linear and quadratic, and between quadratic and cubic. Print the graphs with the quadratic and cubic lines fitted to the points.
6. Analyze your data with Spearman's rank correlation, using my spreadsheet. Report the P-value.
7. Now that you've analzyed your data three different ways, which technique do you think is best for your data? Why?
Return to the Biological Data Analysis syllabus
Return to John McDonald's home page
This page was last revised October 28, 2015. Its URL is http://udel.edu/~mcdonald/stathw9.html