Apr. 8

Answer questions. Finish check activity (started Monday).

Discuss Carey chapter 1, go through one or two exercises …

Sampling Activities:
To start, help students distinguish between a sample and a population. Provide some examples, use as an introduction to the concept of figuring out exactly what population is being sampled (give a survey to 100 people in the area outside the bookstore, measure 10 rays off Newport Beach, find magnesium ion concentration in 100 mL water from L. Superior) , and thinking critically about the extent to which a sample represents the desired population.

1st Sampling Activity: Give students a sample of beads. Have students randomly sample 10 beads from population, records results, put beads back into population, repeat 5X. Then have group count how many are present for a color—RED (15%). As a class, we=ll discuss the data that students created, accenting natural variations (i.e., different sample outcomes). By definition, no one will be exactly correct. This leads to an issue: how confident can you be that the data from your sample accurately represents the overall population, if there=s absolutely no way to check? [If sample data says 29% of the people floss regularly, how many people in the U.S. actually floss regularly? And how sure are you of that figure?]

2nd Sampling Activity: Give students a very large number of some item, e.g., candy sprinkles or beans…much too large a number for them to count every item in the population. Have students sample the population, generating a number representing some attribute (everyone in the class is looking for the same attribute for their sample). Then, ask the overarching question about the fraction of the total population possessing the sampled attribute. Compare data, noting variation (and comparing variation with this larger sample size to that generated with the smaller bead samples).

Use this as an introduction to sampling concepts:
$ As sample size grows, the variation in outcomes decreases
$ As sample size grows, chances increase of getting data that is very close to the ratio in the population.
$ Size of the population is relatively unimportant.
If data permits, show students how the interval containing only 50% of all possible sample outcomes [.50, or 50% confidence interval] is smaller … the more confident you need to be, the larger your interval gets … THIS POINT IS OFTEN CONFUSING; may want to wait to introduce it formally until later. [NEXT TIME: Jelly Blubbers]
[See also Rutledge, M.L. (2001). An activity to demonstrate the concept of sampling error for the introductory biology classroom. Bioscene 27 (1), 3-6. on-line at http://acube.org/volume_27/v27-1p3-6.pdf]

[On-line applet for sampling This is an interactive applet that lets you run sampling tests. You vary sample proportion, sample size, etc. http://www.rossmanchance.com/applets/]

Set up plants, part II … Decide on Q to investigate, how to measure plant growth, key variables to consider, how to share data … Probably begin on Monday and carry through to end of class. I’ll ask about results and their interpretation on final quiz.

For Monday: Swanson (pp. 17-32) … tell students about the book and how the chapter they’re reading is almost just a checklist.