What is Your Favorite Color M&M?

Part 1
The company that makes milk chocolate M&Ms claims the following distribution:
12.5% Brown, 12.5% Red, 12.5% Yellow, 12.5% Green, 25% Orange, and 25% Blue. Is this true?
Ho : P,
= 0 125
.
, P2 0 125 P3 0 125
=
.

,
=
.
. %s Ha : At least 2 of the p; are incorrect
1. Take a class sample and record the results:
65704245
Observed values: Brown:_____ 125 Blue:_____
Red:_____ Yellow:_____ Green:_____ Orange:_____ 93
440 = H
Total number of M&Ms:_______
2. As a class, write down hypotheses for a significance test.

H0: The claimed distribution of color for MaMs is true

.

Ha: The claimed distribution of color for MoMs is not true

.

3. Let’s suppose that M&Ms claimed distribution is correct. If they are correct, how many of each
color would we expect to get in our sample.

55 Red:_____
Expected counts: Brown:_____ 55 Green:_____
55 Yellow:_____ 55 Orange:_____ 118
118 Blue:_____
440(0 . 125) =
np Expected
dF = # categories I -

4. Use the table to calculate the test statistic. dF = 6 1 = 5 -

Observed Expected (Observed − Expected)2

Counts Counts (Observed − Expected) (Observed − Expected)2 Expected

3
Brown 18 100 1 82
.

Red 70 15 225 4 09
.

Yellow 42 -
13 169 3 07 .

Green
-

10 100 1 82.

15 225
123
Orange 110 . 05
2
110 17 289
26
Blue
-

Add up all the numbers in the last column. This is our chi-square test statistic: ! ! = ________
(0 z)2
Xi
-

5. What value would we get for ! ! if our sample was very close to what is expected? Explain. E
Close to zero because the , (Observed Expected) values
-

would all be small

!
6. What value would we get for ! if our sample was very far from what is expected? Explain.
.

A large number because the , (Observed Expected) values

-

would be . larger
7. Go to the candy applet at www.stapplet.com/candy.html and enter the observed counts. At the
bottom of the applet, simulate taking 1000 samples from the claimed population of all M&Ms.
Sketch the approximate ! ! density curve. How likely is it to get the ! ! value we got in class today?
at O
&
right
↓ starts

p-vae
e

009
- 0 =
* skewed .
0
.

lower upper of
ima
· X Cdf (15 48 9999 , 5)
. ,
= 0 0085
.

X2 15 48
=
.
 Lesson 12.1 – Introduction to Chi-Square Tests
QuickNotes LT#2 Test statistic + p-value
LT#1
Hypotheses
X test for goodness of fit (60f) x2 =
( Expected np =

Ho: The claimed distribution of

!df
= # of categories 1 -

categorical variable is true .

Ha: The claimed distribution of #

p-value
cdf(X29999 df)
=
X
categorical variable is not true
.
x2 ,

Check Your Understanding

Are births equally likely across the days of the week? A random sample of 150 births give the following
sample distribution: &
claimed distribution
Day Sunday Monday Tuesday Wednesday Thursday Friday Saturday
Count 11 27 23 26 21 29 13

a. State the appropriate hypotheses.

Ho : Births are equally likely across the days of the week

Ha : Births are not equally likely across the days of the week.
b. Calculate the expected count for each day.

Expected 150 ( /) 21 4 for each

np day
= = =
.

↑
Ho true
assume (equally likely
c. Calculate the value of the chi-square test statistic.
2

(1 -
21 4) (27- 21 4) (13 -
21 4)2
X
.

13 6
.

+ +
.

- T =
. on
.

21 4 .
21 4 . 21 4 .

d. How many degrees of freedom for this chi-square test?

df =
7 1 -
=
6
e. Find the p-value. What conclusion can you make?
lower upper If
Xcdf (13 .
6 , 9999
, 6) = 0. 034 Because .
0 034
.
< 0 05 we
.

rejectHo
and we do have that births not
convincing evidence are
equally
likely across days of the week.