Hypothesis Testing — Finding the critical region
1. Acoinis tossed 10 times. | believe that the coin is more
hypothesis test with 5% significance.
Ho + abwe ¢ 0 he
pros af
probatcile, of aelking,
head.
Hyp POs
3% Brag emmnce« [rad Keo.
Lik X toe the nemo of hands.
Unde Ho Xa 6(10/0%)
H bw 6 (04s) #8
P(X 7B) = 0.0346 7OOT
P(X 79) = 0- 0107 Koos
tags CF 34 0
Cakeo Megan X74
djec Ho f X74
Naod t o¢h Gos 10 heads by
Suppode lackae Hob 1S boased
bewoards hoods ,
ly to come up heads. Find the critical region for a
Write down your null hypothesis. This will be the
probability of success in the “status quo” case.
Write down you alternative hypothesis. This will
involve choosing either > or < or #
Decide on the significance level and whether the test
is one or two tailed.
Write down the distribution under your null
hypothesis.
If your alternative hypothesis is “p>” then find the
smallest value of c such that P(X > c) is less than the
significance level.
IF your alternative hypothesis is “p<” then find the
biggest value of c such that P(X < c) is less than the
significance level.
Write down two probabilities from the tables — either
side of the significance level.
Make sure you draw a sketch of the bars so you are
sure to choose the right value.
State clearly your critical region in the form X= ¢ or
XS as appropriate.
The critical region is the set of values for X that will
cause you to reject Ho.2. The probability thata certain type of seed germinatgs I 0.4. The seeds undergo anew treatment andi wish
to test to see if the germination rate has neraased. A packet of 10 seeds is tested. Find the critical region for
2 10% significance test, In fact'@ seeds germinate. Is this significant?
Hi p co-4
10%Yo wansficance.
Lab Ke menien of sands trot
oll tent
Under Ho KX 6 (10, 0-4) |
lw @(o.10) = 2
P(X £2) 20-1613 Yor0
P(X gi) 20-0464 or < or #
Decide on the significance level and whether the test
is one or two tailed.
Write down the distribution under your null
hypothesis.
IF your alternative hypothesis is “p>” then find the
smallest value of c such that P(X = c) is less than the
ignificance level.
If your alternative hypothesis is “p<” then find the
biggest value of c such that P(X < c) is less than the
significance level.
Write down two probabilities from the tables ~ either
side of the significance level.
Make sure you draw a sketch of the bars so you are
sure to choose the right value.
State clearly your critical region in the form X= c or
X or ca) S half of significance level
Decide whether the test statistic isin the critical
region or not remembering to consider both ends of
the critical region.
Interpret you results. Ifthe test statistic is in the
critical region then it is an extreme result under Ho
and we state “reject H,". Ifthe test statistic is not in
the critical region we state “Do not reject Ho”
Put this in the context of the question.
With these questions they sometimes ask you to
state the significance of the test. They want you to
add up the probabilities of the two critical regions ~
the answer may not be very close to the significance
tended.
(c 60-0353 + 0 01IS2 = 0-005)