Bryan Redd
Homework # 21
.
Problem 1.
An inspector for the Atlantic City Gaming Commission suspects that a particular
blackjack dealer may be cheating (in favor of the casino) when he deals at expensive tables. To
test her belief, she observed 500 hands at the $100limit table and the $3,000limit table. For
each hand, she recorded whether the dealer won (code = 2) or lost (code = 1). Can the inspector
conclude at the alpha level of 0.10 that the dealer is cheating at the more expensive table? That
is, can we infer at the alpha level of 0.10 that the dealer is winning more hands at a more
expensive table? The
output, and not the data itself,
is stored in the attached file
(homework_21_data.xls, use "problem 1 output" sheet). Note: this is a new problem and it is my
admonition to solve it using decision tree and paying particular attention as to whether the
variable of interest is nominal or interval. If, Ftest is applicable, then for the Ftest use alpha of
0.1 also.
a.
Who is the researcher?
An Inspector from Atlantic City Gaming Commission
b.
What is/are the population(s)?
Expensive Black Jack Tables
c.
What is the variable of interest?
Wins and Losses
d.
Of what type (i.e., either nominal or interval) is the variable of interest?
Nominal
e.
Are the samples related or are they independent? Why?
Related because she was comparing the outcomes between. two blackjack tables
f. What is the population parameter (i.e., either proportion or the mean) of interest?
Proportion
g. State clearly in symbols both H
0
and H
a
.
(Please make sure and look at the data before you state the hypotheses, and remember
that if this is a related samples problem, the Excel subtracts the variable2 range from the
variable 1 range. In other words, the column of data that you would specify for variable 2
range will be subtracted from the column you would specify for the variable 1 range).
Ho: P2 ≤ P1
Ha: P2 > P1
h. Is H
a
twotailed, lowertailed, or uppertailed?
Uppertailed
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Documenti.
Is it required to run an Ftest? Explain! If the answer is yes, go ahead and run the Ftest
and interpret the results!
No, because there are not two constant hypothesis.
j. What is ttest or tstat or ztest or zstat?
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '05
 Jensen
 Statistics, Statistical hypothesis testing, researcher

Click to edit the document details