Sie sind auf Seite 1von 5

Name: Nicholas Mack

Block: 3

Group Members names: Jake Sasfai, Maya Gujaral , Salem Balay


Chi Square Analysis Using M&M Statistical Data
Introduction (paragraph form)
A Chi-Square is a test that compares the observed and expected set of data from
an experiment. The primary function of a Chi-Square is to prove or disprove the
null hypothesis of an experiment which, in all scenarios, states: Any difference
between the observed and expected data is due to chance. The task of every ChiSquare analysis is to determine whether the actual experimental data matched the
expected data of the same experiment and, if they are not, the reasoning behind
it. Once both categories are recorded and a Chi-Square value is calculated (see
chart), a student must also calculate the degrees of freedom which is defined as
the number of categories tested 1 or n-1. The reason why a category must be
subtracted is because one category will always be statistically dependent on the
other categories. While these categories have freedom in numerical value, the last
category is never independent because it relies on the other categories to receive
an expected sum. For instance, if three values must add up to a total of 10, the
first two values are degrees of freedom because any number between 1 and 10
will work for these values. The third value, however, is dependent because it relies
on the first two values to get a sum of 10. Once a degrees of freedom is calculated,
a student can then use the value from the Chi-Square and the degrees of freedom
to determine whether the data proves or disproves the null hypothesis.
In this investigation, students used a bag of M&Ms to calculate a Chi-Square based
on the amount of each color in the bag, which included brown, blue, orange, green,
red, and yellow M&Ms. Students were given a chart from the M&M company which
detailed the expected percentage of each color in a standard bag of M&Ms. For this
experiment, the goal was to physically count how many of each color was in a
standard bag of M&Ms and compare those numbers to the data given by the
company. The observed and expected data was collected and used to calculate a
Chi-Square value which proved or disproved the null hypothesis. The null
hypothesis stated that both sets of data were different by chance. This Chi-Square
value was compared to the class data, which comprised of data from several
groups of students doing the same experiment.
Problem (question)
The primary question in this experiment was whether the observed and expected
data were different by chance or by some other factor.
Hypothesis (statement)
If the observed and experimental data differ, then they are different by chance, not
by statistical error.
Materials (list)

The materials used include:


-

1bag of M&Ms

1 paper plate

6 paper cups

Marker

Procedure:
1. Obtain the materials listed above.
2. Open the bag of M&Ms and pour all pieces into a plastic plate.
3. Obtain 6 paper cups and label them based on color. The colors are brown, blue,
orange, green, red, and yellow.
4. Sort all M&Ms based on color and record the total number of each color into
your data chart. This is the observed data.
5. Determine the total number of M&Ms in the bag.
6. Look at the chart given by the M&M company (in HO). Using this reference
chart, multiply the percentage of each color by the total number of M&Ms in the
bag. Record. The percentage is from the expected data.
7. Calculate the difference of the observed data by the expected data for each
color. Record these values in the chart.
8. Square the differences for each color and record these values. Equation: (O-E) 2.
9. Divide the squared differences by the expected value for each color. Equation:
(O-E)2 /E {NOTE: By this point, there should be 3 rows of data in the chart.}
10. Add up the values from this row to get the chi-square value for this set of
data.

Results (data, chart, drawing, table, etc)


GROUP DATA (M&M
Brown
Observed 76
Expected 54.6
Differenc 21.4

Type = Milk chocolate)


Blue
Orange
Green
94
90
46
100.8
84
67.2
-6.8
12
-21.2

Red
75
54.6
20.4

Yellow
33
58.8
-25.8

TOTAL
420
420
X

e
Differenc
e
Squared
(O-E)2 / E
ChiSquare
Value

457.96

46.24

144

449.44

416.16

665.64

8.39
X

.459
X

1.714
X

6.69
X

7.622
X

11.32
X

X
36.193

Mass: 357.47g
Result: Rejection of null hypothesis
CLASS DATA
Brown
Observe
315
Expected 297.18
Differenc 17.82
e
Differenc 317.55
e
Squared
(O-E)2 / E 1.07
ChiX
Square
Value
Mass: 1331.26g
Result: Acceptance of

Blue
532
548.64
-16.64

Orange
474
457.2
16.8

Green
408
365.76
42.24

Red
281
297.18
-16.18

Yellow
276
320.04
-44.04

TOTAL
2,286
2286
X

276.90

282.24

1784.22

261.79

1939.52

.505
X

.617
X

4.878
X

.881
X

6.06
X

X
14.011

null hypothesis Conclusion: Acceptance of null hypothesis

Error Analysis (paragraph form)


There were very few errors in this experiment. However, one potential human error involved the use of
significant figures. In an experiment like this, where an exact calculation is needed to verify the null
hypothesis, significant figures are important. In the chart above, there are values that range from 3
significant figures to 5 significant figures. Some values were also rounded. Should this experiment be
performed again, the student should make sure that all values have the same amount of significant
figures so that the student can get the most accurate and precise Chi-Square value.
Discussion and Conclusion:

Conclusion:
The objective of this investigation was to prove or disprove the null hypothesis that differences in the
observed and experimental data were due to chance alone. To determine the experimental data, as
recalled in the introduction, the group counted the number of M&Ms and sorted them based on color.
They counted and recorded the amount of M&Ms for each of the six colors. From the experiment, after
multiple calculations, it was determined that the Chi-Square value for the group data was 36.193. This
value was very far off from the predicted value of 11.07 (derived from chart given) so, initially; the
group rejected the null hypothesis. To recall, the further away the Chi-Square value is from the predicted
value, the less likely it is that the difference between the observed and experimental data were due to
chance alone. However, from the class data, the Chi-Square value was 14.011. This value was closer to
the predicted value and therefore allowed the class to verify the null hypothesis. The reason why the
class data was used in this conclusion rather than the group data was because the group data contained
data from all the groups rather than one individual group. Because the class data had more accurate
information, its Chi-Square value was used to prove the null hypothesis.
Based on this investigation, the class concluded that the difference between the data given by the Mars
company and the experimental data was due to chance only. Even though the class Chi-Square value
was a bit higher than expected, it was still close to the predicted value given from the chart. Because it
was close to this predicted number, there is little chance that any other factors affected the experiment,
such as human or calculation error. The importance of this investigation is that it teaches how scientists
from all fields use Chi-Squares to verify the certainty of an experimental outcome.
Lab Questions:
1. Based on the class data, should you ACCEPT or REJECT the null hypothesis? Why?:The null
hypothesis was accepted because the Chi-Square value (14.011) for the class data as extremely close
to the value that was predicted (11.07) for 5 degrees of freedom at a .05 probability. Because this
value is closer to the predicted value, we can assume that the difference between the observed and
expected results were do to chance alone.
2. If you rejected the null hypothesis based on the class data, what might be some of the explanations
of your outcome?: If there was a case where I did reject the null hypothesis based on the class data, I
would say the reason why the outcome was not affected by chance alone was because the M&M
company might have added to much or too little M&Ms to some bags that were tested. It is also
possible to say that the class and / or the students miscounted the number of M&Ms.
3. If you accepted the null hypothesis, how do you explain it particularly if you rejected the null
based on your individual groups data? The reason why the null hypothesis was accepted, even
though the individual group data disproved it, was because the class data was simply more accurate.
Unlike the group data where each Chi-Square value varied slightly, the class data was the result of
multiple experiments conducted at the same time. This Chi-Square value produced from the class
data acted as mean value for the entire class. Therefore, it is more accurate because it expresses
every groups data. This Chi-Square value led to the conclusion that the null hypothesis was correct.
4. What is the purpose of collecting data from the entire group? The purpose of collecting data from the
entire group is to ensure the accuracy of our conclusion. If we relied on individual group data, there
might have been a drastic difference in Chi-Square values. Collecting data from all the groups and
getting a value from those results further stabilizes our conclusion. In this case, since the class ChiSquare value was close to the predicted value (11.04), we accepted the null hypothesis, believing

that the difference in the observed and experimental data was due to chance alone.
5. Practice Problem:
GG

Gg

Gg

gg

Phenotype ratio (Green, albino) 3:1

Sources
NONE USED

Das könnte Ihnen auch gefallen