Workshop Description

Through this seminar, attendees will walk away knowing when and how to run modern versions of traditional statistical analysis. These tests and the underlying bioinformatical lesson about resampling will be of use to most scientific disciplines. The course makes no assumptions about familiarity with traditional statistics – we will simply go through relatable experimental examples and ask how to test various hypotheses, introducing the relevant methods along the way. There will be homework assignments each night to solidify the concepts from class. These are short and optional, but will allow advanced students to gain more of the class.

As we introduce modern methods, we will also address how to best deal with common questions in experimental analysis:

  • What tests can be applied to small sample sizes?
  • When can one assume linearity in the data?
  • What to do with non-Gaussian data / residuals?
  • How does one best detect and justify removing outliers?
  • Which estimators are appropriate with highly non-continuous samples?

Day 1

  • Bootstrapping
  • Hypothesis testing

To access the workshop slides for day 1 – 3, click here.

Day 2

  • Paired data
  • Correlations

To access the workshop slides for day 1 – 3, click here.

Day 3

  • ANOVAs
  • Multiple comparison corrections

To access the workshop slides for day 1 – 3, click here.

Workshop Materials

1) Helpful to have A basic understanding of either MATLAB or R.

2) We have 10 brand new mac computers with all relevant software installed. If you would prefer to use your own laptop you may do so if you have the following installed PRIOR to class: You need to have MATLAB or R already installed.


Dr. Don Vaughn is a neuroscience postdoctoral fellow at UCLA. His research has included functional Magnetic Resonance Imaging (fMRI), perceptual psychophysics, and sensory substitution. Don uses fMRI to research neural correlates of freewill and how social dynamics modulate empathy. In psychophysics, Don investigated how information, before and after an event has occurred, influences perception of the event – an effect dubbed peri-diction. Don now applies multivariate classification and non-parametric statistics to bioinformatics datasets.


Education: BS Physics Stanford 2008; BA Economics Stanford 2008

Workshop Details

Prerequisites: None
Length: 3 days, 2.5hrs per day
Level: Introductory
Location: Collaboratory Classroom  (Boyer Hall, 529)
Seats Available: 28

Fall 2018 Dates

October 9, 10, and 11, 2018
1:00 PM – 3:30 PM