You are on page 1of 7

Skittles Term Project

Chelsey Heward
Math 1040

For this project I will be comparing the amount of each color that is in each
bag of skittles. I have my own data from my own skittle packet, and the
results from my other classmates. I first counted each color of skittles and
sent my data to the professor. The point of this project is to apply things I
have learned in the class. First I made a pie chart showing the total amount
of colors for my whole class.

Organizing and Displaying Categorical Data:

When I first started I thought that each bag of skittles would have the
same amount of skittles in each bag or at least vary by one or two. You

would think that they would distribute the colors evenly in each bag, but
according to the class data this is not true. They do not vary much in
numbers but there are slight differences in each bag of skittles. My personal
bag of skittles were very close to the average of the classes. A lot of the bags
seem to have the same total amount of skittles, but vary more on each of
the colors amount.

Organizing Quantitative Data:


Summary statistics:
Column

Number of
skittles

15

Mean

Variance

Std. dev.

Std. err.

60.6 5.5428571 2.3543273 0.6078847

Median Range Min Max Q1 Q3


61

55

63

59

63

Candies in my bag: 61

The histogram shows that most of the skittles lie in between 58-64 and that
not very many were less than 58. This graph does a good job showing that
each bag has a slightly different amount.

Reflection:
Categorical data is statistical data consisting of categorical variables or
group data. Quantitative data is certain quantity measurements using units.
Pie graphs should be used for categorical data because it easily shows the
differences in each category. Pareto charts or stem leaf charts should be
used for quantitative data because you need to be accurate with your data
and shouldnt be misleading.
Confidence Interval Estimates
The general purpose and meaning of a confidence interval is to guess
the true value of a population parameter. There is a range of values as a
parameter and a specified probability that a value is in the parameter.
1. Construct a 99 percent C.I. estimate for the true proportion of yellow
candies. We are 99 percent confident that our C.L. is in between .176
and .246 for the true proportion of yellow candies.

.
2. Construct a 95 percent C.I. estimate for the true mean # of candies per
bag. We are 95 percent confident that the mean number of candies will
fall in the parameter of 59.03 and 61.57.
3. Construct a 98 percent C.I. estimate for the standard deviation of the #
of candies per bag. We are 98 percent confident that the standard
deviation of the number of candies per bag will fall in the parameter of
1.589 and 3.969.

Hypothesis Tests:
The purpose and meaning to these tests are to find out if the
hypothesis seems to have significance.
1. Use a 0.05 Significance level to test the claim that 20 percent of all
skittles candies are red. Because the value .249 falls in the middle
and not in the critical area that means that there is not significance
to test the claim that 20 percent of skittle candies are red.
2. Use a 0.01 significance level to test the claim that the mean
number of candies in a bag of skittles is 55. Because 8.964 falls in
the critical area, this means that there is significance to test the
claim that the mean number of candies in a bag of skittles is 55. So
you will reject the noll.

Reflection:
The conditions for doing a confidence interval is that the sample
is a simple random sample, conditions for the binomial distribution
are satisfied, and there are at least 5 successes and at least 5
failures. For this project and constructing the confidence interval,
requirements were met to proceed. The hypothesis testing is a
standard procedure for testing a claim about a property of a
population. First identify the null and then the alternative
hypothesis, which were both found in our data.
A problem with our data might be that someone couldve not
been counting correctly about how many they have in their bag.
Also the amount of samples for the class is not very large so the
data would be more meaningful if there were more peoples data
involved. By doing this statistical research I have learned that there
is a lot more than you think that you can figure out about data.

You might also like