In: Statistics and Probability
Two marathon training procedures are tried for comparison
purposes. Their efficacy is to be determined in a marathon race.
Assume race times are given in minutes. Use α=0.056α=0.056 for all
calculations.
Procedure_A Procedure_B
146 144
147 147
149 150
142 147
144 141
147 147
143 147
148 148
147 148
146 151
147 145
144 141
144 153
147 150
140 141
151
144
156
150
153
155
140
138
146
141
151
(a) Do the populations appear to be normal? Hint: use a
probability plot to decide
A. Yes, procedure B appears to be normal but
procedure A does not appear to be normal
B. Yes, Both procedure A and procedure B appear to
be normal
C. Yes, procedure A appears to be normal but
procedure B does not appear to be normal
D. No, Both procedure A and procedure B appear to
be normal
E. No, neither procedure A appears to be normal
nor procedure B appears to be normal
F. Yes, Both procedure A and procedure B appear to
be non-normal
(b) Choose the most appropriate statistical hypotheses to see which
procedure results in longer marathon times (be very careful in how
you formulate your hypothesis, this one is tricky! Use the sample
data to make any assumptions necessary.)
A. H0:μ1=μ2HA:μ1>μ2H0:μ1=μ2HA:μ1>μ2
B. H0:μ1=μ2HA:μ1<μ2H0:μ1=μ2HA:μ1<μ2
C. H0:μD=0,HA:μD<0H0:μD=0,HA:μD<0
D. H0:μ1=μ2HA:μ1≠μ2H0:μ1=μ2HA:μ1≠μ2
E. H0:μD=0,HA:μD≠0H0:μD=0,HA:μD≠0
F. H0:μD=0,HA:μD>0H0:μD=0,HA:μD>0
(c) Using technology available to you, test to see if the variances
are equal or not?
A. There appears to be more variation in the 2nd
procedure then in the 1st.
B. They appear to be equal.
C. There appears to be an unequal amount of
variation in the 2nd procedure then in the 1st.
D. There appears to be more variation in the 1st
procedure then in the 2nd.
(d) Report the p-value of the test you ran in (c) use at least
three decimals in your answer (this is from Levene's test).
P-value =
(e) Test the statistical hypotheses in (b) by carrying out the
appropriate statistical test. Find the value of the test statistic
for this test, use at least two decimals in your answer (from your
test of middle).
Test Statistic =
(f) Determine the PP-value for this test, to at least three decimal
places (from your test of middle).
(a)
From above probability plots, we see that both data sets are obtained from two independent normal populations.
Hence Option B. Yes, Both procedure A and procedure B appear to be normal.
(b)
Option: D. H0:μ1=μ2 HA:μ1≠μ2
(c)
Test and CI for Two Variances: Procedure_A, Procedure_B
Method
Null hypothesis Sigma(Procedure_A) / Sigma(Procedure_B) =
1
Alternative hypothesis Sigma(Procedure_A) / Sigma(Procedure_B) not
= 1
Significance level Alpha = 0.05
Statistics
Variable N StDev Variance
Procedure_A 26 4.526 20.486
Procedure_B 15 3.697 13.667
Ratio of standard deviations = 1.224
Ratio of variances = 1.499
Tests
Test
Method DF1 DF2 Statistic P-Value
F Test (normal) 25 14 1.50 0.433
Levene's Test (any continuous) 1 39 0.70 0.409
Since P-value>0.05 so
Option: B. They appear to be equal.
(d) P-value=0.409
(e)
Two-Sample T-Test and CI: Procedure_A, Procedure_B
Two-sample T for Procedure_A vs Procedure_B
N Mean StDev SE Mean
Procedure_A 26 146.38 4.53 0.89
Procedure_B 15 146.67 3.70 0.95
Difference = mu (Procedure_A) - mu (Procedure_B)
Estimate for difference: -0.28
95% CI for difference: (-3.07, 2.50)
T-Test of difference = 0 (vs not =): T-Value = -0.20 P-Value =
0.839 DF = 39
Both use Pooled StDev = 4.2471
Test Statistic = -0.20
(f) P-value for this test=0.839