A Guide to Using Statistics in BCS Lab Courses

This Stats Wiki is a guide to the use of statistical tests and software tools in BCS lab courses. If you are an undergraduate student who has been enrolled in a lab course, you are encouraged to skim through the example problems in this wiki before class begins. In the process of designing experiments and conducting statistical tests, this wiki can also help you by providing toy problems and their solutions, from which you can work by analogy.

If you are a graduate TA, please feel free to edit this wiki to make it more useful to students. To request editing permissions to this wiki (an existing wiki account is required), please contact Ting at tqian@bcs.rochester.edu .

Table of Contents

• Experimental Design Basics:

  • Random Variables

    • Continuous vs. Discrete

      • Dependent vs. Independent

    • Within vs. Across Subjects

  • Samples and Summary Statistics

  • Estimators

  • Choose the Right Statistical Test

  • Interpret a Significant Result

• Common Scenarios (Parametric Tests):

  • When should you NOT use the following tests?

  • There is one variable of interest:

    • Is the value of the variable different from X?

    • Is the value of the variable in population A different from that in population B?

    • Does the value of the variable differ among Groups A, B, C, D...? If so, between which two groups?

  • There are two variables of interest:

    • Do they correlate with each other?

    • Does one of the variables predict the other well?

  • There are more than two variables of interest:

    • What are the typical situations with more than two variables involved?

    • Two-way ANOVA

    • Multiple regression

• Statistical Software Programs:

  • Prism 4.0

  • Vassar Stats

  • R

• Homework Sets: