Hello! And happy Groundhog Day to you all –
Tomorrow morning (week five of the 10 week, introductory course - first course in a two course series) brings me to the start of a required unit on probability.
One of the nice things about teaching at MHCC is that ALL math classes are taught with students in table groups of 2-4 students, so group activities are a natural
and expected part of students’ experience.
As an introduction, I use a modification of an activity stolen from you guys, Allan and Beth. Table groups are each given a sheet with four rectangles on it,
and the names here Hermione Hearts, Cathy Clubs, Delila Diamonds, and Sally Spades each in one of the rectangles. Then we go through the story of the staff at a hospital handing babies back to their mothers at random, and what is the probability that all
four mothers will go home with the correct baby.
Each table repeats the experiment 10 times, for a total “probability sample” of 90 trials. We use this to introduce the idea of probability as long-term relative
frequency - then we go back and list all possible outcomes of the experiment as permutations of 1234 and calculate classical probabilities as well. However we like to emphasize that relative frequency trumps the classical model - that's how we know that the
probability of a woman having a baby boy is not exactly 50%.
From this activity, we can also develop the ideas of probability distributions based on relative frequency histograms.
I soft pedal the algebra of probability, instead use Roxy Peck's idea of a hypothetical sample in a two-way table. I find it helps reinforce the idea of a categorical
variable being separate from the values that variable can take – eg Dorm Type as the variable as opposed to Co-ed.
Conditional probability and Independence can be taught as concepts springing from the sample itself, rather than coming from algebraic definitions.
Rob Hauss
Mathematics Instructor
Mt Hood Community College
503 491 7383
From: sbi-bounces@causeweb.org [mailto:sbi-bounces@causeweb.org]
On Behalf Of Joseph Oechsle
Sent: Monday, February 02, 2015 8:38 AM
To: Simulation-Based Inference
Subject: Re: [SBI] Happy Groundhog Day! What happened in your introductory statistics class today?
This is my plan for tomorrow morning's Intro Stats class: (8:10-9:50)
I put four Excel spreadsheets on Blackboard: one with data re colleges and universities, one with a cell phone survey data, and two re automobiles. I will allow ten minutes for groups to create numeric and graphic analyses of the data
sets and present them briefly to the class. (review of types of variables and appropriate numeric and graphic analysis)
I'll discuss the computation of standard deviation and Z scores and then do an interactive application of these tools to four worksheets I will have put on Blackboard and we will complete in class.
After a five minute break, I will introduce a "Friendly Observers" scenario (24 people randomly assigned to 2 groups; one group performs a skilled task and is told it will be observed and rewarded monetarily if it exceeds its prior threshold
performance; the other group does not have observers who will also be rewarded if the performers exceed the threshold and only the participants will receive a reward. Three persons in the first group exceed the threshold, and eight in the second group do.
The question is: do vested observers have an effect on performance?)
One group in class will simulate the scenario using playing cards, one with a random number table, one with a calculator random number generator, and one with Excel's random generator. Groups will present a graphic of their results, and
we will discuss them without using the language of inference.
(We did Dolphin Therapy two weeks ago on the first day of class)
On Mon, Feb 2, 2015 at 12:34 AM, Allan Rossman <arossman@calpoly.edu> wrote: