SBC 3305
SBC 3305
A drug manufacturer advertise that a product dissolves completely in gastric juice within 45 sec. To
determine whether its dissolving time is longer than advertised, the dissolving time is investigated and
the following data was obtained; 42.7,43.4,44.6,45.1,45.6,45.9, 46.8,47.6. Perform a hypothesis test to
determine whether the mean dissolving time is longer than advertised. Use a significance level of α =
0.05.
Solution The One-Tailed t Test for the Hypotheses H0 : µ ≤ 45 sec and HA : µ > 45 sec
Dissolving times (in sec) of a drug in gastric juice: 42.7, 43.4, 44.6, 45.1, 45.6, 45.9, 46.8, 47.6.
H0 : µ ≤ 45 sec
HA : µ > 45 sec
α = 0.05
n=8
X̄ = 45.21 sec
SS = 18.8288 sec2
s2 = 2.6898 sec2
ν=7
t0.05(1),7 = 1.895
If t ≥ t0.05(1),7 , reject H0 .
Conclusion: do not reject H0 .
(n − 1)S 2
χ2 =
σ2
follows a chi-square distribution with n−1 degrees of freedom. Therefore, if we’re interested in testing
the null hypothesis:
H0 : σ 2 = σ02
against any of the alternative hypotheses:
(n − 1)S 2
χ2 =
σ02
and follow the standard hypothesis testing procedures.
Example: For the data in the previous example, a manufacturer might be interested in whether the
variability in the dissolving times of the drug is greater than a certain value-say, 1.5 sec. Thus, Ho :
σ 2 ≤ 1.5sec2 and HA : σ 2 > 1.5sec2 .
(n − 1)S 2
χ2 =
σ02
where σ02 = 1.5 sec2 is the hypothesized variance under the null hypothesis.
Substituting the values, we get:
ν =n−1=8−1=7
We compare the test statistic to the critical value of χ2 at α = 0.05 and ν = 7 degrees of freedom.
From chi-square distribution tables:
χ20.05,7 = 14.067
Since 12.5524 < 14.067, we fail to reject H0 .
Conclusion: At the 0.05 significance level, there is not enough evidence to conclude that the vari-
ability in the dissolving times is greater than 1.5 sec2 .
3
Example 2: A manufacturer of hard safety hats for construction workers is concerned about the mean
and the variation of the forces its helmets transmits to wearers when subjected to an external force. The
manufacturer has designed the helmets so that the mean force transmitted by the helmets to the workers
is 800 pounds (or less) with a standard deviation to be less than 40 pounds. Tests were run on a random
sample of n = 40 helmets, and the sample mean and sample standard deviation were found to be 825
pounds and 48.5 pounds, respectively.
Do the data provide sufficient evidence, at the level, to conclude that the population standard devi-
ation exceeds 40 pounds?
H0 : σ 2 = 402 = 1600
against the alternative hypothesis:
HA : σ 2 > 1600
Therefore, the value of the test statistic is:
(40 − 1)48.52
χ2 = = 57.336
402
Is the test statistic too large for the null hypothesis to be true? Well, the critical value approach
would have us finding the threshold value such that the probability of rejecting the null hypothesis if
it were true, that is, of committing a Type I error, is small... 0.05, in this case. Using a chi-square
probability table, we see that the cutoff value is 54.572.
That is, we reject the null hypothesis in favor of the alternative hypothesis if the test statistic is greater
than 54.572.
From the above results, the test statistic (57.336) is greater than 54.572. Therefore, we conclude that
there is sufficient evidence, at the 0.05 level, to conclude that the population standard deviation exceeds
40.
Practice Question: Do the data provide sufficient evidence, at the α = 0.05 level, to conclude that
the population standard deviation differs from 40 pounds?
Two-Sample Hypotheses
Among the most commonly employed biostatistical procedures is the comparison of two samples to infer
whether differences exist between the two populations sampled. As an assignment, you will research
and read about hypotheses comparing two population means and variances (or standard
deviations) and do the assignment that follows. You will discover another very important sampling
distribution, the F distribution—named for its discoverer, R. A. Fisher and further use of Student’s t
distribution.
• Null hypothesis: The null hypothesis is that the two groups have equal means.
• Alternative hypothesis: The alternative hypothesis can be one-sided or two-sided. A one-sided
hypothesis states that one group’s mean is higher or lower than the other’s. A two-sided hypothesis
states that the two means are different.
• Test statistic: The test statistic is a standardized difference between the two sample means.
4
• Paired or unpaired data: The data can be paired or unpaired. Paired data means that there is
a one-to-one correspondence between the values in the two samples.
• Variances: The variances of the two samples can be assumed to be equal or unequal.
• Statistical software or tables: The test statistic can be used to determine if the null hypothesis
should be rejected. This can be done using statistical software or a table in a statistics textbook.
The appropriate statistical test to use depends on a number of factors, including the assumptions
made about the data and the hypothesis being tested.
After reading about Two-Sample Hypothesis, attempt the following questions (As part of
CAT 2 -Take away). Show all steps of your calculations and explain your reasoning.
Question 1 In an experiment, adult male rabbits were divided at random into two groups, one group
of six and one group of seven. The members of the first group were given one kind of drug (called “B”),
and the members of the second group were given another kind of drug (called “G”). Blood is to be taken
from each rabbit and the time it takes the blood to clot is to be recorded.
Do the data provide sufficient evidence, at the α = 0.05 level, to conclude that the mean of the
population of blood-clotting times of all adult male rabbits who might have been administered drug B
(let’s call that mean µ1 ) is the same as the population mean for blood-clotting times of all adult male
rabbits who might have been given drug G (call it µ2 )?
Question 2 Using the following data, test the null hypothesis that male and female turtles have the
same mean serum cholesterol concentrations.
Serum Cholesterol (mg/100 ml) Male Female
220.1 223.4
218.6 221.5
229.6 230.2
228.8 224.3
222.0 223.8
224.1 230.8
226.5
Question 3 A random sample of 10 doses of a liquid medication dispensed by Dispenser A had a mean
volume of 203 ml and a standard deviation of 3 ml. A random sample of 15 doses dispensed by Dispenser
B had a mean volume of 206 ml and a standard deviation of 5 ml. Assume that the amount dispensed
by both dispensers follows a normal distribution. Test, at the 0.05 significance level, the hypothesis that
there is no difference in the variability of the volumes dispensed by the two dispensers.
Question 4 Korn Krispies are a type of breakfast cereal, and they are packed in boxes with a nominal
net mass of 296 grams. Owing to overwhelming demand, the manufacturers have installed a new and
faster machine to fill the boxes with cereal. However, to meet government regulations, amongst other
things, the variability in the packed masses of these boxes should not increase over present levels. The
table below gives the masses of a random sample of 10 boxes of cereal from the original packing machine,
and the masses of a random sample of 12 boxes of cereal from the new machine.
5
Assuming that these independent samples came from underlying normal populations, use a 1% level
of significance to determine whether an increase in variance has occurred.
(b) Test for equality of the variances of the two kinds of air pollutants.