So to begin our team project for statistics, each of us bought a bag of Skittle candies

and reported to the Professor how many we had of each color. We then used this
data throughout the project, computing different statistics.
First, our small group took guesses on what we thought the proportion would
be for each color, then we computed the true frequency/proportion as well as
graphed our results into both a histogram and a pie chart:

1.
a. Janelle- Guessed that the highest proportion would be Yellow and Green and
the lowest would be Red. Because in the individual packages she buys, there
always seem to be a lot of Yellow/Green and only a few Red.
Rosa- Guessed that Purple would be the highest proportion because her
package had the most of that color.
Bryce- Guessed that Purple would be the highest proportion because he had
mostly purple in his package.
Nidha- Guessed that Purple would be the highest proportion, and Yellow and
Red would be the least.
b.
Skittle
Color
Red
Orange
Yellow
Green
Purple

Freque
ncy
1238
1148
1238
1133
1135

Relative
Frequency
0.21
0.195
0.21
0.192
0.193

Proport
ion
21%
19.50%
21%
19.20%
19.30%

2.

3. Yes, this is a random sample. The definition of a simple random sample

means: A sample of size n from a population size N is obtained through simple
random sampling if every possible sample of size n has an equal chance of
occurring. In our Skittle project, each student was asked to by a 2.17 ounce bag
of Skittles. Each student bought there bag at random, as in we probably bought
them at different stores, gas stations, or online websites. The sample is the 2.17
bags of skittles that each student bought. There is no bias in this experiment
because we each bought our own random bag.

The population is every bag of 2.17 ounces that is produced from the Wrigley
Company in Illinois. We as students do not even scratch the surface of how
many Skittles are out there produced by the Wrigley Company. But from our
small class sample, we can make observations about which colors are mostly
produced in each bag.

I then wrote my own reflection of how I felt about the results versus what
my guesses were:

1. The graphs definitely werent what I expected. I was so surprised how close in
count every color was, I thought there would be more of a difference, but I was
correct in assuming that Yellow was the most. Although, I thought for sure that
Red would be one of the lowest counts, because its my favorite one and I never
feel like theres enough. But, as not only my bag but also the class count
showed, it is the second highest. I guess I never realized and just always want
more Red! I dont feel like there are any outliers at all. The distribution is just a
little different from my bag of Skittles. The Red was the highest count in my bag
with Yellow in second, whereas those frequencies were switched in the class
count. This was also the case with Purple and Orange. However in both counts,
Green was the lowest count.
We then started looking at the statistics of the number of candies per bag
(not focusing on the color) and reported the results as well as made a
histogram and a boxplot:
Mean number of candies per bag: 60.1
Standard Deviation of the number of candies per bag: 5.6

5 Number Summary of the number of candies per bag: 37, 58, 60, 62, 82

Next we worked on confidence intervals, how confident we are about the

population of all skittles candies, based on our sample, in regards to
certain colors and also the number of candies per bag
A confidence interval is a range of values which is believed, with certain degrees
of confidence, that the true value of the population lies within its bounds. The
purpose of it is to confidently estimate a parameter of the population (such as
mean, standard deviation or proportion) without having to gather all the extra

data (as youre obtaining this confidence interval from a sample rather than
having to gather all the data for an entire population).

1. Total yellow candies: 1189

Total candies: 5650
P=0.2104
Z =2.576

Lower Bound:

PZ 0.005

P (1P )
0.2104 ( 0.7896 )
0.21042.576
=0.196
n
5650

Upper Bound:

P+Z 0.005

P ( 1P )
0.2104 ( 0.7896 )
0.2104+2.576
=0.224
n
5650

Margin of error:

0.2240.196
=0.014
2

A 99% confidence interval for the proportion of yellow skittles is (0.196, 0.224).

2. Total candies: 5650

Total bags: 94
Mean: 60.106
Standard Deviation: 5.556
T 0.025=1.96

Lower Bound:

s
5.556
X T 0.025
60.1061.96
=59.961
75.166
n

Upper Bound:

s
5.556
X +T 0.025
60.106+1.96
=60.251
75.166
n

60.25159.961
=0.145
2

Margin of error:

One can be 95% confident that the mean candies per bag is between the upper and
lower bounds.

3. Total candies: 5650

Standard Deviation: 5.556
X 2R=0.01=135.807
X 2L=0.99=70.065

( n1 ) s 2
( n1 ) s2
< <
X 2R
X 2L

5649 30.869
5649 30.869
< <
135.87
70.065

35.833< < 49.888

Margin of error:

49.8835.833
=7.028
2

We can say, 98% confidence, the population standard deviation is between (35.833,
49.888).

Reflection:

I think what I learned the most was that statistics arent always what you expect. When I
first thought about skittles, I thought there are never enough of my favorite one, which is
Red, so that has to be the lowest count in the bag, and Orange and Yellow being the
most. Yet when we computed the statistics for 94 bags, we found that the quantities of
the different colors were not far apart at all.
Second, I am currently majoring in Psychology and Criminal Justice, so
these skills will definitely help me in a lot of future classes I will be taking. I will now be
able to compute my own statistics regarding things such as drug abuse, mental illness,
basically anything that has to do with my majors, and Ill know how to understand them
and apply my results to any project I may work on.
So all in all, I think what I took most from this is that you cant take things for face
value as they arent always what they seem- and that is a part of why we have statistics.
You cannot base results on your own judgement of how you perceive things, you need
to actually study and observe, record your results and compute the true statistic. This
class has also taught me to be more observant and cautious of statistics that are given
to the public as well. I now have the ability to see if something was accurately observed
and recorded or if I believe the statistic or take it with a grain of salt.