ST1232

Tutorial 7
1. A person who claims to possess extrasensory perception (ESP) says she can guess more often than not the
outcome of a flip of a balanced coin. Out of 20 flips, she guesses correctly 12 times. Would you conclude
that she truly has ESP? Answer by reporting all five steps of a significance test of the hypothesis that each
of her guesses has probability 0.50 of being correct against the alternative that corresponds to her having
ESP.
1. Condition / Assumption:
Categorical variable (correct or incorrect guess), random , normal [np(1‐P)= 5]
2. H0: p = 0.5 (random guessing)
H1: p > 0.5 (has ESP)
3. Test statistic, Z =( p‐hat – p ) / SE = 0.89
4. P‐value = P(Z > 0.89) = 0.19
5. P‐value = 0.19 > α= 0.05, do not reject H0.
If she has no ESP and did so by random guessing, there is 0.19 probability of getting such a result (12
correct guesses) or more correct guesses, thus there is not enough evidence against H0.
There is not much evidence that the probability of a correct guess is higher than 0.50.

2. For a quantitative variable, you want to test H0: µ = 0 against H1: µ ≠ 0. The 10 observa ons are 3, 7, 3, 3,
0, 8, 1, 12, 5, 8.
a) Show that the sample mean = 5.0, sample standard deviation = 3.71, standard error =1.17, test statistic =
4.26, and df = 9.
n = 10
Sample mean, Y‐bar = ∑Y/n = ∑ (3+7+3+3+0+8+1+12+5+8)/10 = 5.0
Sample standard deviation, s = sqrt [∑(Y – Y‐bar)2/(n–1) ]=3.71
Standard error, SE = s / sqrt(n) = 1.17
Test statistic = (Y‐bar – µ)/( SE) = (5.0 – 0)/(1.17) = 4.26
df = n – 1 = 9.

b) The P‐value is 0.002. Make a decision using a significance level of 0.05. Interpret.
P‐value = 0.002 < 0.05, thus H0 is rejected.
If the H0 is true, there is 0.002 probability of observing this sample (sample mean of 5.0) or values more
extreme, thus there is enough evidence to reject H0.
Decision: Reject H0.

1
 c) If you had instead used H1: µ > 0, what would the P‐value be? Interpret it.
The test statistic = 4.26 (positive test statistic indicates that the test statistic is above the mean) in this
case.
If a right‐tailed test is use, then the P‐value is only half of the P‐value of a two‐sided test,
P‐value = P(Y‐bar > 5.0 |H0 is true) = 0.001
If the H0 is true, there is 0.001 probability of observing this sample (sample mean of 5.0) or values
bigger, thus there is enough evidence to reject H0.
Decision: Reject H0.

d) If you had instead used H1: µ < 0, what would the P‐value be? Interpret it.
The test statistic = 4.26 (positive test statistic indicates that the test statistic is above the mean) in this
case.
If a left‐tailed test is use, then the P‐value is the complement of the P‐value of a right‐tailed test.
P‐value = P(Y‐bar < 5.0 |H0 is true) = 0.999.
If the H0 is true, there is 0.999 probability of observing this sample (sample mean of 5.0) or values smaller,
thus there is not enough evidence to reject H0
Decision: Do not reject H0.

Multiple Choice questions

3. The P‐value for testing H0: µ = 100 against H1: µ ≠ 100 is 0.001. This indicates that
a) There is strong evidence that µ = 100
b) There is strong evidence that µ ≠ 100, since if µ were equal to 100, it would be usual to obtain data such
as those observed.
c) The probability that µ = 100 is 0.001.
d) The probability that µ = 100 is the significance level, usually taken to be 0.05.

4. An environmental action group is concerned that an industrial plant may be exceeding pollution limits.
They took four readings to analyze whether the true mean discharge of wastewater per hour exceeded
the company claim of 1000 gallons. When we make a decision in the one‐sided test using α = 0.05:
a) If the plant is not exceeding the limit, but actually µ = 1000, there is only 5% chance that we will conclude
that they are exceeding the limit.
b) If the plant is exceeding the limit, there is only a 5% chance that we will conclude that they are not
exceeding the limit.
c) The probability that the sample mean equals exactly the observed value would equal 0.05 if H0 were true.
d) If we reject H0, the probability that it is actually true is 0.05.
e) All of the above.

5. For a test of H0: µ = 0 against H1: µ > 0 based on n = 30 observations and using α = 0.05 significance level,
P(Type II error) = 0.36 at µ = 4. Identify the response that is incorrect.
a) At µ = 5, P(Type II error) < 0.36
b) If α = 0.01, then at µ = 4, P(Type II error) > 0.36.
c) If n = 50, then at µ = 4, P(Type II error) > 0.36.
d) The power of the test is 0.64 at µ = 4.

2
 True or False
6. It is always the case that P(Type II error) = 1 – P(Type I error).
False.
P(Type II error) = 1 – Power.

7. If we reject H0: µ = 0 in a study about change in weight on a new diet using α = 0.01, then we also reject
it using α = 0.05.
True.
If H0 is rejected at α = 0.01, it means the P‐value is < 0.01, thus it is less than 0.05.

8. A study about the change in weight on a new diet reports P‐value = 0.043 for testing H0: µ = 0 against H1:
µ ≠ 0. If the authors had instead reported a 95% confidence interval for µ, then the interval would have
contained 0.
False.
A 95% confidence interval is equivalent to a two‐sided significance test at α = 0.05.
Since P‐value is 0.043, H0 of µ = 0 is rejected at α = 0.05, thus 0 is not one of the plausible values, it is not
within the 95% confidence interval.

9. A 95% confidence interval for µ = population mean IQ is (96, 110). So, in the test of H0: µ = 100 against
H1: µ ≠ 100, the P‐value > 0.05.
True.
100 is within the 95% confidence interval, µ = 100 is plausible. Thus H0 of µ =100 will not be rejected (P‐
value > 0.05) for a two‐sided test.

10. For a fixed significance level α, the probability of a Type II error increases when the sample size increases.
False.
When sample size n is increased, Type II errors is decreased.

11. The P‐value is the probability that H0 is true.

False.
P‐value = P(Observing the sample statistics or values more extreme|H0 is true).
P‐value ≠ P(H0 is true).