Table of Contents

• How big is my population?
• 1: Counting big populations
• 2: Sampling on a grid
• 3: Scaling it up
• 4: Keeping it random
• 5: When size matters
• 6: Bare bones sampling: P/A
• 7: Collecting P/A data
• 8: Impervious Surface
• 9: Getting fancy: transect sampling
• 10: Try a transect
• How do my populations interact?
• 11: Making connections
• 12: Pond pollution paradox
• 13: Proving pond pollution
• 14: Pond pollution the elegant way
• 15: Killer housecats
• 16: Housecats exposed
• 17: Honeybee preferences
• 18: Interpreting honeybees
• 19: Review

 

Keeping it random

 

You used an applet to randomly select the 10 grid segments. Why is it important to randomly select segments to count?
 

It's something called "experimenter’s bias"! You can’t help but be a little biased. You as the sampler might have a tendency to choose the most conveniently located segment, or the fullest segment, or the least full segment, or the most interesting looking segment, or the segment the farthest from the beehive…..

 

 

BUT - for the random sample to be really, really, random, all members of the population need to have an equal chance of being selected. To overcome experimenter’s bias, there are all sorts of different ways to generate a random selection. In the olden days, we used random number tables (like the one on the left, titled "Ten Thousand Random Digits") to come up with numbers. The tables would be conveniently placed in the back of your text book, but using them was painful.

Thankfully, in this modern age, computers can efficiently generate random numbers for us and we can now leave the random number tables with the abacuses and slide rules. Try this random number generator.