Scribd
Upload
96 views

Hypothesis Testing Quick Reference - 1

Hypothesis testing is used to test claims about a population by considering a sample. It involves stating a null hypothesis that assumes the claim is false and an alternative hypothesis that assumes it is true. There are different types of hypotheses tests depending on the relationship between the null and alternative, such as one-tailed tests when the alternative is greater than or less than the null. Errors can occur when the test result does not match the truth, including type 1 errors when a true null is rejected and type 2 errors when a false null is not rejected. The significance level and p-value are used to determine the likelihood of type 1 errors.

Uploaded by

Swathi Mithai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
96 views

Hypothesis Testing Quick Reference - 1

Hypothesis testing is used to test claims about a population by considering a sample. It involves stating a null hypothesis that assumes the claim is false and an alternative hypothesis that assumes it is true. There are different types of hypotheses tests depending on the relationship between the null and alternative, such as one-tailed tests when the alternative is greater than or less than the null. Errors can occur when the test result does not match the truth, including type 1 errors when a true null is rejected and type 2 errors when a false null is not rejected. The significance level and p-value are used to determine the likelihood of type 1 errors.

Uploaded by

Swathi Mithai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as XLSX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Hypothesis in English mean 'assumption'.

Hypothesis testing is a statistical technique used to test if an assumption or claim


considering the given data.

Hypothesis testing is formulated such that assumption or claim that has to be proved is mentioned in the Alternate Hypothes
always hold a condition which we want to reject.

In the world of analytics hypothesis testing is widely used to check if the sample being considered is a representation of the
performed on the mean and variance of the data. Additionally there are hypothesis testing functions available to perform n
check correlations etc.

One Tail Right Sided Hypothesis Test One Tail Left Sided Hypothesis Test
Alternate hypothesis will have a '>' sign Alternate hypothesis will have '<' sign

Example: Pizza official's competitor claim that the Example: Pizza officials claim that the mean
mean delivery time is more than 30 mins delivery time is less than 30 mins

Hence the null and alternate hypothesis is: Hence the null and alternate hypothesis is:
Ho: µ = 30 mins Ho: µ = 30 mins
Ha: µ > 30 mins Ha: µ < 30 mins
Representation: Representation:

While we do the hypothesis testing there is possibility that there will be some errors. The output of the test may not be true
Below table represents the possible type of errors.

Ho True (Should not reject)


Reject Ho Type I Error (α)
Fail to Reject Ho Correct decision

α : level of significance
Confidence level = 1 - α

The Type I Error returned by hypothesis test is the p-value of the test.
If we want to be 95% confident on the result of the hypothesis test i.e. confidence level = 0.95, we keep threshold of 'α' a
compare the p-value with 0.05.
If p-value < 0.05, null hypothesis is rejected for a confidence level of 0.95.
used to test if an assumption or claim made is indeed true

mentioned in the Alternate Hypothesis. Null hypothesis will


to reject.

considered is a representation of the population. Tests are


sting functions available to perform normality test, test to

Two Tail Hypothesis test


Alternate hypothesis will have '<>' sign

Example: A customer complains that the mean


delivery time of pizza is not equal to 30 mins

Hence the null and alternate hypothesis is:


Ho: µ = 30 mins
Ha: µ != 30 mins
Representation:

he output of the test may not be true for few data points.
e of errors.

Ho False (Should reject)


Correct decision
Type II Error (β)

Power of the test = 1- β

e p-value of the test.


vel = 0.95, we keep threshold of 'α' as 0.05 (1-0.95) and
.
nfidence level of 0.95.

You might also like