Biomedical Statistics Course ANAT 597 Examination One: Due Tuesday, April 13th.

Table of Contents...Hide

Examination One Consists of Two Projects

Project One - Discrete Random Variables

• Draw on information from each of the first 6 chapters of our textbook.
• The dataset(s) can be from your laboratory, from a public repository, or you can make it up.
• The dataset needs to satisfy the Binomial Distribution.
• Provide R scripts used to describe the sample(s).
  • Provide a printed table.
  • Provide a printed graphic (bar, line, etc).
• Provide R scripts used to test the null hypothesis.
  • Is the null rejected at the .05 level?
  • Binomial test

Project Two - Continuous Random Variables

• Draw on information from each of the first 6 chapters of our textbook.
• The dataset(s) can be from your laboratory, from a public repository, or you can make it up.
• The dataset needs to satisfy the Normal Distribution.
• Provide R scripts used to describe the sample(s).
  • Provide a printed table.
  • Provide a printed graphic (line, histogram, etc).
• Provide R scripts used to test the null hypothesis.
  • Is the null rejected at the .05 level?
  • T-test - no more than two groups. Can be independent or dependent.

Suggestions for Each Project

• Restrict each project to one or two groups. Do not plan on doing an ANOVA. For the continuous random variable project you can do a one-group (test an outlier) or two-group t-test with independent or dependent data.
  • Do not use a large sample size unless you are working with something meaningful. Do not spend excessive time at the keyboard with data entry. In fact, import existing data; or manually type it in; either way.
• Discuss these projects with anyone including each other and/or your PI; or work alone.
  • Although two group designs are rare in laboratory research; a t-test can be used as a post hoc test. Thus, you could use real data from you laboratory to compare two conditions from a much larger design.
  • Please test the sample for assumptions of variance. This can be easily done in R.
• Treat this is an opportunity to "dry run" a sample of your dissertation data and design (no more than two groups).
• Please do not use the same datasets among you.
• The outline of your projects can essentially follow the outline of the book (below) as relevant.