Hypothesis Testing
Hypothesis Testing
Hypothesis Testing
Tuhina Manna
Hypothesis Testing
Statistical Hypothesis
–
a conjecture about a
population parameter. The
particular value or range of
values of population
parameter is claimed and we
need to test this claim is true
or not by observing the
sample.
The per capita
There are 30% income of people in
asymptomatic What is a
India are Rs.1,600.
covid infected Hypothesis?
people in West
Bengal
3 Confidence Interval
Step 1: State null and alternate hypotheses
Steps in
Hypothesis Step 2: Select a level of significance
Testing
Step 3: Identify the test statistic
. Right-tailed and left-tailed tests are distinguished by the way the greater than or less than
points in the alternative hypothesis only. They help us with the decision rule.
One-Tailed Tests of Significance
The H1: The mean yearly
alternate commissions earned by full-
hypothesis, time realtors is more than
H1, states a Rs.35,000. (H1: µ>Rs.35,000)
direction
H1: Less than 20
H1: The mean speed of percent of the
trucks traveling on NH- customers pay in
95 is less than 60 km per cash for their
hour. (H1: µ<60) internet charges.
(H1: m<0.20)
Two-Tailed Tests of Significance
No direction is specified in the alternate hypothesis H1.
H1: The mean H1: The mean
amount spent by volume of
customers at diesel
Spencer is not consumed in
equal to India is not
Rs.1500. equal to 25
H1: The mean air quality (µ ≠ Rs1500). billion litres.
index in Kolkata today is
not equal to 50. (µ≠ 25).
(µ ≠ 50).
Step One: State the null and alternate
hypotheses
H 0: m = 0
H 1: m = 0
Three The null
possibilities hypothesis
regarding H 0: m < 0 always contains
means H1: m > 0 equality.
H 0: m > 0
H 1: m < 0 3 hypotheses
about means
When a researcher conducts a study, he or she is generally looking for evidence to support a
claim of some type of difference. In this case, the claim should be stated as the alternative
hypothesis. Because of this, the alternative hypothesis is sometimes called the research
hypothesis.
• 3) A psychologist feels that playing soft music during a test will change the results of the
test. The psychologist is not sure whether the grades will be higher or lower. In the past,
the mean of the scores was 73.
• H0 : H1 :
Example: Decision Errors in a
Legal Trial . What are H0 and H1?
H0 is True H0 is False
Reject H0 Type I Correct
Error Decision
hypothesis.
The z value is based on the
sampling distribution of X,
which is normally
distributed when the sample
Examples: z, t, F, c2 is reasonably large: n>30
(recall Central Limit
Theorem).
Step Four: Formulate the decision rule.
• Critical or Rejection Region – the range of values for the test value that
indicate a significant difference and that the null hypothesis should be
rejected.
• Total area of critical region= P(Type I error) = α
Right-Tailed Test at 5%
Region of
Level of Significance Do not
rejection 𝐻0
P(z>1.645)=0.05 or 𝑧0.05 = 1.645 reject 𝐻0
[Probability =.95] [Probability=.05]
Prob=α
Similarly, for Right-Tailed Test at Prob=1-α
1% Level of Significance
P(z>2.33)=0.01 or 𝑧0.01 = 2.33
0 1.645
1.65
Critical value
ഥ𝟏 −ഥ
𝒙 𝒙𝟐
2. Difference of population mean: Test statistic=𝒛 = ,
𝝈𝟐 𝟐
𝟏 +𝝈𝟐
𝒏𝟏 𝒏𝟐
ഥ𝟏 −ഥ
𝒙 𝒙𝟐
In case population s.d. 𝝈𝟏 , 𝝈𝟐 are not given , 𝒛 =
𝒔𝟐 𝟐
𝟏 + 𝒔𝟐
𝒏𝟏 𝒏𝟐
𝒑−𝑷
3. Single population proportion: Test statistic: 𝒛 = , where p= sample proportion and P= population
𝑷𝑸
𝒏
proportion, Q=1-P.
𝒑𝟏 −𝒑𝟐
4. Difference of population proportion: Test statistic= 𝒛 = 𝑷𝟏 𝑸𝟏 𝑷𝟐 𝑸𝟐
,
𝒏𝟏
+ 𝒏
𝟐
𝒑𝟏 −𝒑𝟐 𝒏𝟏 𝒑𝟏 +𝒏𝟐 𝒑𝟐
In case, population proportions 𝑷𝟏 , 𝑷𝟐 are not given 𝒛 = , where 𝒑 = , 𝒒=𝟏−𝒑
𝟏 𝟏 𝒏𝟏 +𝒏𝟐
𝒑𝒒 +
𝒏𝟏 𝒏𝟐
• Example: It has been reported that the average credit card
debt for college seniors is $3262. The student senate at a
large university feels that their seniors have a debt much less
than this, so it conducts a study of 50 randomly selected
seniors and finds that the average debt is $2995, and the
population standard deviation is $1100.
𝑑ത 0.5
Test statistic Formula for paired sample t-test process: 𝑡 = = 1.08/3 = 0.33
𝑠/ 𝑛−1
This process also the test value < critical value , so null hypothesis is accepted and we conclude the manufacturer’s
claim is valid.
The graph and critical value is same for both the processes but the test statistic value is different.
In a year there are 956 births in a town A of which 52.5% were males,
while in town A and B combined, this proportion in a total of 1406
births was 0.496. Is there any significant difference in the proportion of
male births in the two towns?
Next on Testing of
Hypothesis continuation
ppt:
• Chi Square Distribution:
The probability, assuming that the null
hypothesis is true, of finding a value of
the test statistic at least as extreme as the
p-Value computed value for the test
p-value
Decision Rule
is
If the p-Value is larger If the p-Value is
required than or equal to the smaller than the
for significance level, a, H0 significance level, a,
Practical is not rejected. H0 is rejected.
only Calculated from the
probability distribution Using the p-Value
function or by computer in Hypothesis
Thankyou