Hypothesis Testing Tutorial
Hypothesis Testing Tutorial
Hypothesis Testing Tutorial
1. A random variable X is known to have a normal distribution with variance 36. The mean
of the distribution of X is denoted by . A random sample of 50 observations of X has
mean 20.2. Test, at the 1% significance level, the null hypothesis = 22 against the
alternative hypothesis 22.
2. A supermarket manager investigated the length of time that customers spent shopping in
the store. The time, x minutes, spent by each of a random sample of 150 customers was
measured, and it was found that x 2871 and x 2 60 029 . Test, at the 5% level
of significance, the hypothesis that the mean time spent shopping by customers is 20
minutes, against the alternative that it is less than this.
(N91/P2/10)
x 4.344, ( x x )2 0.8022
Carry out a 2-tail test, a the 5% level of significance, to test whether the mean of X is
4.58. State your null and alternative hypothesis clearly.
(FM N2003/P2/6)
4. An automatic dispensing machine is set to dispense 200 cm3 of liquid into cups. A
random sample of six cups was taken, and the cups were found to contain the following
volumes of liquid in cm3 .
(i) State suitable null and alternative hypotheses for this test.
(ii) State an assumption about the distribution of the volume of liquid dispensed which
you need to make in order to use a t-test.
(iii) Carry out an appropriate t-test, using a 5% significance level, to test whether the
machine is dispensing too much liquid.
Practice Questions
1. A coin is chosen at random from a population of recently produced coins. The discrete
random variable X is the age, in years, of the coin. The population mean of X is denoted
by μ, the population standard deviation is denoted by . A random sample of 120
independent observations of X was taken and the results can be summarised as follows.
Age (x) 0 1 2 3 4 5
Frequency ( f ) 14 26 21 23 17 19
2
(i) Calculate the exact values of the unbiased estimates of and
(ii) Using a 10% significance level, test the null hypothesis 2.4 against the
alternative hypothesis 2.4 .
(iii) Explain whether it is necessary to use the Central Limit Theorem in your test.
2. The mass, x kg, of the contents of each packet in a random sample of 80 cereal packets is
measured, and the results are summarised by
x 79.53, x2 100.4621
Test, at the 4% significance level, whether the population mean mass of the contents is
less than 1.10 kg.
Explain, in the context of the question, the meaning of ‘at the 4% significance level’.
In another test, using the same data and also at the 4% significance level, the hypotheses
are as follows.
Given that the null hypothesis is rejected in favour of the alternative hypothesis, find the
set of possible values of 0 .
(N2003/P2/29/mod)
(i) It is given that the population variance is 15. Carry out a 2-tail test of the null
hypothesis = 9.5 at the 5% significance level.
(ii) It is given instead that x2 6172.31 . In a 1-tail test of the null hypothesis
= 11.5, the alternative hypothesis is accepted. State the alternative hypothesis,
and find an inequality satisfied by the significance level of the test.
(N2004/P2/30)
4. A plant breeder claims that a new variety of fruit bush he has produced gives a higher
yield of fruit than the variety it will replace. A random sample of 10 bushes of the new
variety is grown and the yields of the bushes recorded. The old variety has an average
yield of 5.2kg/bush. The yields of the sample of new bushes are summarised by
It may be assumed that the yield from each bush is an independent observation from a
normal distribution. Test at 5% level of significance, the breeder’s claim.
Another plant breeder also claims that his new variety of fruit bush gives a higher yield
of fruit than the old variety. A small sample of 15 bushes of the new variety is taken
from this breeder. The standard deviation of the yield from this sample is 0.2. When a
test is conducted at 5% level of significance, it was found that there is insufficient
evidence to justify this breeder’s claim. Give an estimate of the maximum possible total
yield from this sample.
5. The number of minutes that the 0815 bus arrives late at my local bus stop has a normal
distribution; the mean number of minutes that the bus is late has been 4.3. A new
company takes over the service, claiming that punctuality will be improved. After the
new company takes over, a random sample of 10 days is taken and the number of
minutes that the bus is late is recorded. The sample mean is t minutes and the sample
variance is k 2 minutes2. A test is to be carried out at the 10% level of significance to
determine whether the mean number of minutes late has been reduced.
(i) State appropriate hypotheses for the test, defining any symbols that you use.
(ii) Given that k 2 3.2, find the set of values of t for which the result of the test
would be that the null hypothesis is not rejected.
(N2014/II/9/mod)
In a test at the 5% significance level it is found that there is significant evidence that the
population mean talk-time is less than 5 hours. Using only this information, and giving a
reason in each case, state whether each of the following statements is (i) necessarily true,
(ii) necessarily false, or (iii) neither necessarily true nor necessarily false.
(a) There is significant evidence at the 10% significance level that the population
mean talk-time is less than 5 hours.
(b) There is significant evidence at the 5% significance level that the population
mean talk-time is not 5 hours.
(N2001/P2/10 modified)
1. A motorist records the time taken, T, minutes, to drive a particular stretch of road on
each of 64 occasions. Her results are summarized by
t 876.8, t2 12657.28 .
(ii) Test, at the 5% significance level, whether the mean time for the motorist to drive
the stretch of road is greater than 13.1 minutes. Explain whether it is necessary to
use the Central Limit Theorem in your test.
(2011 TPJC CT)
2. At an early stage in analysing the marks scored by the large number of candidates in an
examination paper, the Examining Board takes a random sample of 250 candidates and
finds that the marks, x, of these candidates give x 11 872 and x 2 646 193 .
Using the figures obtained in this sample, the null hypothesis = 49.5 is tested against
the alternative hypothesis < 49.5 at the α % significance level.
Determine the set of values of α for which the null hypothesis is rejected in favour of the
alternative hypothesis.
It is subsequently found that the population mean and standard deviation for the paper
are 45.292 and 18.761 respectively. Find the probability of a random sample of size 250
giving the sample mean at least as high as the one found in the sample above.
(N88/P2/10)
3. Ben, a fruit seller, only sells mangoes from an orchard. In the orchard, the weight of the
mangoes are normally distributed with a mean weight of 200g and standard deviation
39g.
One day, Andy, a business partner of Ben, brought in a large batch of mangoes from the
orchard. 10 mangoes were randomly selected from this batch and the weights (in g) of
the mangoes were as follows:
If a test at 2.5% significance level indicates there is a decrease in the mean weight of the
mangoes the orchard, find the greatest value of x , correct to 1 decimal place.
(2007 HCI Prelim mod)
4. The average number of particles emitted from a cathode ray tube is 20 in a millisecond.
A scientist makes improvement to 10 such tubes so that the mean and standard
deviation of particles emitted are 21 and 2.5 per millisecond respectively. Determine
whether there is really an improvement at the 5% significance level. State any
assumption(s) made.
Challenging Question
1. (a) A large department store wants to find out how much its customers spend on
Gundam model kits. From a random selection of 100 transactions, the results are
summarized by
x 3500 , x2 220 400 ,
If the researcher had carried out a z-test instead, determine which of the following
2 statements is correct, giving clear reasons to support your claim.
(I) The researcher would have concluded that there is significant evidence at
the 5% significance level that the population mean breaking strain is
different from 0 kN.
(II) It is not possible for the researcher to conclude, by using only the
information given, whether there is significant evidence at the 5%
significance level that the population mean breaking strain is different
from 0 kN .
Numerical Answers
Tutorial Questions
5(ii) t 3.48
6(a) Necessarily true 6(b) Neither necessarily true nor necessarily false
2. 4.02 ; 0.0321
3. 208.2
4. p-value = 0.130 (Do not reject H 0 )
Challenging Question
1(a) p-value = 0.0559 (Do not reject H 0 )