Empirical Investigation: A UVicinspired Proposal

1
Empirical InvestigationA UVic-inspired Proposal

• Mike Godfrey
• SWAG, University of Waterloo

2
The initial exploration

• Try this
• Time yourself walking around the windowed area in
  this building.
• Count the number of rabbits you see
• Calculate
• bunnies / time MTBB
• (Mean time between bunnies)

3
Terminology issues

• Need to establish common working vocabulary to
  avoid confusion
• e.g, What does density mean?

4
What is the density of rabbits at Uvic?

• Experiment 1
• For a given patch of Uvic land
• Count the bunnies seen
• Bunnies move quickly, may need to issue each with
  a RFID (Rabbit Fur Identifier)
• Tagging may also work Storey 06 (to appear)
• Measure area of patch
• Perform appropriate math
• Ask a grad student if unsure

5
What is the density of rabbits at Uvic?

• Experiment 2
• Grab several bunnies
• Measure the volume of each
• Then immerse in water, and weigh the displacement
  of water
• The EUREKA project Archimedes -215
• Remove bunnies from water OR YOU WILL HAVE
  PROBLEMS REPLICATING THE RESULTS

6
What is the density of rabbits at Uvic?

• Experiment 3
• Grab several bunnies
• Given them each an IQ test
• Compare to average UVic undergrad as a baseline
• Hire the best as RAs

7
A new CSER project!

• Emergent Applications for
• Rabbit-oriented Services
• aka EARS
• Sponsor
• PETA (People for Empirical Testing of Animals)

8
Why rabbits are appropriate for CSER

• Rabbits are good at replication!
• Rabbits already have a custom protocol for
  network communication
• They convert incoming data into small packages
  for redistribution
• AKA pellet switching
• Industry has good experience with rabbit
  consolidation
• e.g., Schneiders meats
• Rabbits can be converted into soft wear

9
Is this a good empirical study?

• What about the Heisenberg effect?
• Does the act of observing bunnies interfere with
  their presence?
• Perhaps they like being observed.
• Does the act of observing and counting bunnies
  interfere with career path of a computer science
  professor?
• Perhaps they will be put on medical leave.

10
Threats to validity

• Can we reliably identify a bunny?
• False positives

11
Threats to validity

• False negatives

12
Threats to bunnies

• Hawks
• Foxes
• Hungry grad students
• Perhaps we can explicitly provision for them in
  the next round of CSER applications if NSERC
  approves