Title: Colored Words Experiment
11
2Introduction Thoughts Behind The Experiment
- We wanted to chose an experiment that was easy
to complete, organized, and some what
entertaining. - Our original hypothesis, was
that people who ate more Candy on a Regular Basis
and had a stronger preference for Sweets rather
than Chocolate would have more success and be
able to correctly identify more skittles in one
minute than those who do not.
2
3The Task
- 40 Subjects
- Objective
- How many Skittles can the Subject identify in one
minute based solely on their Taste Buds - Categorical Factors
- Gender
- Male or Female
- Candy Consumption
- Frequent, Sometimes, Never
- Candy Preference
- Chocolate or Sweet
- Hair Color
3
4Overall Quantitative Data
Mean 2.07962 Standard Dev. 1.45791 Min 0 Q1
1 Median 2 Q3 3 Max 6
The Data is right skewed and unimodal. The center
is at the median of 2 skittles with an IQR of 2
skittles as well. The range was (0, 6) skittles.
All data points are within the Upper Lower
Fence (UF6) (LF-2), thus there are no outliers.
4
5- 1 sample T Interval on the true average of
skittles - Conditions
- Statement
- Formula
- Interval (a, b)
- Conclusion
6Quantitative Data by Gender
Male Mean- 2.07692 Standard Dev.- 1.45791 Min-
0 Q1- 1 Median- 2 Q3- 3 Max- 4
Female Mean- 2.10526 Standard Dev.- 1.6632 Min-
0 Q1- 1 Median- 2 Q3- 3 Max- 6
The plots of our data broken down by gender shows
that the both females and males had the same
median at 2 skittles. The range for females was
larger, which was (0,6) skittles, compared to the
males which was (0,4) skittles. However they both
had an IQR of 2 skittles. Finally, the males
data appears to be more symmetrical, while the
females data is clearly right skewed. Both are
unimodal when we look at histograms of the data.
6
7- 2 sample t test on male vs. female average
skittles - Conditions
- Statement
- Hypotheses Ho µF µM Ha µF ? µM
- Test statistic
- P-value
- Conclusion
- If reject, complete a confidence interval
8Quantitative Data by Candy Consumption
SUMMARY STATS NEEDED FOR EACH VALUE OF THE
VARIABLE
8
9Quantitative Data by Candy Consumption
(frequently, sometimes, or never)
We chose to compare Candy Consumption to the data
results to see if there was an association
between people who ate Candy Frequently,
Sometimes, or Never with the number of Skittles
they correctly identified.
- Our categorical variable that worked best with
the data was if the subject ate candy on a
regular basis (Frequently). The Frequently
variable was more consistent overall than the
Sometimes and Never variables - Frequently had a median at 3.5 skittles, which
was the highest. Next was Sometimes with a
median of 2 skittles, And never was the lowest
with a median of ½ skittles. - The Sometimes variable had the largest range from
(0,6) making it the most inconsistent, where as,
the Never variable had a range from (0,1) and the
Frequently variable had a range of (2, 5). The
IQR of Frequently was - The shape of frequently was ..
9
10- X2 GOF test on type of Candy Consumption
- NEVER SOMETIMES FREQ
- OBS 10 28 30
- EXP (uniform)
- Conditions
- Statement
- Hypotheses
- Test Statistic
- P-Value
- Conclusion
- NOTE
- If your categorical variable only has 2 values,
instead of doing a Chi-Square GOF test, complete
a 1 prop Z test to see if the of one of the
values is 50
11Simple Two-Way Table of Categorical Data
11
12Marginal Distribution for Gender
Male 52.5 Female 47.5
12
13Marginal Distribution for Candy Preference
Chocolate 50 Sweet 50
13
14 C S
F 57.89 42.11
M 42.86 57.14
List the in each category Of the females C
57.89 S 42.11 Of the males C 42.86 S
57.14
15- X2 Test for Association
- Observed Table Expected Table
- Conditions
- Statement
- Hypotheses
- Test Statistic
- P-value
- Conclusion
(made in Excel copied in)
16 17Sources of Error and Bias
17
18Conclusion
- Make a conclusion about each of the
tests/intervals you did - Make a conclusion about each graph you made
18
19EYEWEAR
Glasses/Contacts None total
GENDER M 12 18 30
F 18 23 41
total 30 41 71