Glogster

M&M Lab

Concepts of Chi-Square distributions and "Goodness of Fit" test using M&M's
By: Aidan, Brett, Chelsey, My-Linh

In 1995, the M&M manufactuer's company claims that on average a new mis of colors of m&m's plain chocolate candies will contain 30% Browns, 20% yellows and reds, and 10% of oranges, greens, and blues. This has been know change every few years. Our task is to see if the current mix of m&m's matches that of the company's claim. It has also been known that the smaller packages we will use vary more than normal size packages.

In order to test this claim we will use a Chi-Square "Goodness of fit" Test.

First we divided the colors up in our group with the four smaller packages of M&M's we used. Next we recorded our counts of blue, green, red, yellow, brown,and orange and the expected values of these colors based on the company's claim on the chart below.

GRAND TOTAL: 86

ASSUMPTIONS: All E's exceed 5


H0 : The population proportion of M&M's are equal to those claimed by the manufacturer

Ha : They are not equal

On the Calculator.....
Enter observed values in chart into ---L1
Enter expected values in chart into ---L2
Enter ((O-E)^2)/E values into---L3

CONCLUSION:
We reject the null since pvalue less than alpha=.05.
There is statistical evidence to suggest that the population proportions of M&M colors is not equal to the manufacturers claim.

M&M Goodness of Fit

By: Kayla, Karina, Leighton, and Molly
In 1995 m&m/mars company claimed the mix of plain chocolate candies is 30% brown, 20% yellow, 20% reds, and 10% each of oranges, greens, and blues. In order to determine whether or not the current mix of colors matches that of 1995 by using a chi-square goodness of fit test. In each m&m sample the colors will be counted and reported on tables. Based on these counts, we will be able to determine whether the sample is true to the Mars company claim.


Ho: The actual percentage of m&ms colors is equal to the percentages stated by the Mars company.

Ha: The actual percentage of m&ms colors is not equal to the percentages stated by the Mars company.


chi-squared: 30.0715
p-value: 1.4278x10-5

Reject null. p-value < .05
Statistical evidence suggesting the actual percentage of m&ms does not equal the percentages stated by the mars company in 1995.

Group Projects

We will be conducting various experiments for the rest of the semester. You will receive half a test grade for an experiment designed by me that your group will lead. Your group will be responsible for conducting the experiment, collecting data, and leading the class through the analysis of the data. Each class member will be responsible for analyzing the class data and submitting a formal statistical write up that will count as a daily grade. The group leading the experiment will be responsible for posting the results on this blog. The group in charge will write a short description of the activity, provide the data collected, and perform a significance test or confidence interval for the data.

The culminating project for the year is a group designed experiment. Each group will choose a topic that they would like to investigate and test. The topic and experiment must receive approval by me. Students in our class will serve as subjects for testing. Presentation of the topic and analysis of results will be presented at the end of the year. More information on this project will be presented later. In the meantime, be thinking of a topic you would like to investigate.

Dan Ariely on Predictably Irrational

This is a wonderful video in which Dan Ariely, author of Predictably Irrational discusses social experiments that he conducted on pain management and cheating. Watch the 18 minute video and provide a short comment in response to share with the class.

Another Quiz for AP Review

This is a seven question quiz on which significance test to use for a given situation. You will need your username and password to take it. The passcode is again twhs. Click Here

Quiz on Planning a Study

Take this seven question quiz on planning a study. Use your username and password given previously in class. The passcode for the quiz is twhs. Click here

Countdown to the Test - Review of Exploring Data

How do you describe a distribution?

• mention center, shape, and spread in context

What about comparing two distributions?

• describe the shape of both distributions
• compare the centers and spreads - example the median of the distribution of length of signature for the females is larger than the signature length for males. The IQR of the male signature lenghts is larger than the IQR of female signature lengths. No credit is given for just listing measures of center and spread.

What kinds of graphs can be used to describe univariate data?

• dotplot
• stem & leaf plot
• histogram
• boxplot
• normal probability plot

Quiz Review for Ch 23-25 Test

A portion of your test on Tuesday will be questions regarding which significant test should be performed. This quiz will give you a few practice questions as well as other information on the test. This quiz is by no means all you need to know for the test, but it is a good start. To take the quiz, you will need a username and password, which will be provided to you on Monday. If you are not in this class and would also like to participate, please use the username: twhsguest and password: guest. Test passcode: twhs Click here to take the quiz.