Scribd
Upload
48 views5 pages

Statistics Homework 2 Problem 1.: ST RD

This document appears to be statistics homework that analyzes the calorie content of chocolate and vanilla ice cream. It tests the hypothesis that there is no difference in calorie content between chocolate and vanilla ice cream flavors. Data on the calorie content of 79 ice cream brands was collected and analyzed. Statistical tests found no significant difference in the average calorie content between chocolate and vanilla ice cream flavors, failing to reject the null hypothesis. Therefore, calorie content does not appear to impact consumer preference between chocolate and vanilla ice cream.

Uploaded by

Nick Ashley
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
48 views5 pages

Statistics Homework 2 Problem 1.: ST RD

This document appears to be statistics homework that analyzes the calorie content of chocolate and vanilla ice cream. It tests the hypothesis that there is no difference in calorie content between chocolate and vanilla ice cream flavors. Data on the calorie content of 79 ice cream brands was collected and analyzed. Statistical tests found no significant difference in the average calorie content between chocolate and vanilla ice cream flavors, failing to reject the null hypothesis. Therefore, calorie content does not appear to impact consumer preference between chocolate and vanilla ice cream.

Uploaded by

Nick Ashley
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

Statistics homework 2 problem 1.

The feud between vanilla and chocolate has created enemies in ice
and rain alike. Just how to make the decision between two scrumptious
delights is very perplexing to the uneducated consumer. In considering
potential factors resulting in consumer preference, caloric intake may
contribute to the preference curve of a chocolate or vanilla consumer.
It was hypothesized that caloric content of chocolate vs. Vanilla icecream would have a null effect on consumer utility derived from said
product. To test that hypothesis, a population of 79 icecream brands
both chocolate and vanilla was assembled and calorie content was
noted.
Initially looking at the data, the variable calories was separated by
flavor to compare chocolate vs. vanilla. The subset of chocolate icecream was found to have the following characteristics:
Min.
120

1st Qu
140

Median
170

Mean
198.7

3rd Qu
260

Max
320

In addition, the standard deviation of the chocolate subset was 63.06.


This value represents a maximum standard deviation equivalent to
31.73% of the mean. Therefore the data is varied around the mean.
In the case of the Vanilla subset, the same approach was repeated
save isolating vanilla as the flavor of interest. The subset of vanilla was
found to have the following characteristics.
Min.
120

1st Qu
140

Median
160

Mean
191.4

3rd Qu
240

Max
307

In addition, the standard deviation of the chocolate subset was 58.644


This value represents a maximum standard deviation equivalent to
30.63% of the mean. Therefore the data is varied around the mean.
The boxplot of both subsets is as follows:
As shown, the mean for vanilla is slightly lower than the mean for
chocolate. There also appears to be a larger sample distribution for
chocolate than there does for vanilla. But how can we know for certain?

A random distribution analysis of the mean difference for the data was
performed. The sample size was set to n=491 and R was repeated
1000 times. It was hypothesized that a null hypothesis would return a
mean difference of 0, since that would indicate no difference in Icecream means. The alternative hypothesis was that Vanilla was not
equal to Chocolate. When the data was tested, the left tailed P value
was calculated at 0.305. Since the randomization distribution is not
100% normal, we must multiply that value by 2, giving a P value of
0.61. This P value is too large to reject the null hypothesis. Therefore,
there is not a statistically significant different in calories between
chocolate and vanilla ice-creams.

Theres a terrible movie from the 90s called War Games where a
computer narrowly avoids total nuclear warfare by being programmed

to overload itself in a game of tic-tac toe against itself. Case in point,


tic-tac toe is a solved game. The only way anyone wins/loses in tic tack
toe is in the case of someone making a mistake. Therefore, in our
simulation, the maximum number of times the computer lets us win is
the maximum amount of times one could win at tic-tac toe assuming
total tic tack toe proficiency. Upon inspecting the flash applet, it was
determined that the maximum number of times the tic-tac toe game
would let anyone win was 93% of the time. Therefore, our stat of 14/15
wins makes sense for the program but does it make sense in the ever
changing world of randomization?
To find the accuracy of our null hypothesis, H0: P=0.9, and our
alternative hypothesis HA: P>0.9, a right tailed P test was conducted
based on a randomization distribution that was created from R=10000
runs of our original ratio 0.933. The distribution and corresponding P
test is pictured below:

The calculated P value based on the null hypothesis of HO: P=0.9 is


shown to be 0.549. Based on that value, it would be unreasonable to
reject the null hypothesis because it happens frequently. Therefore,
there is not enough evidence to reject the null hypothesis and I win
90% of the time.

You might also like