Scribd
Upload
15 views

Hypothesis Testing and Chi - Square

Uploaded by

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

Hypothesis Testing and Chi - Square

Uploaded by

Ahmed mahyoob Hamoud
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 35

Testing Hypothesis

Chapter 8
Introduction
Hypothesis testing allows the physician to be take decisions regarding comparisons
of population parameters with the theoretical value or another population.

In the testing of the hypothesis, the comparison will always be referred to


population parameters but the decision of accepting or rejecting the hypothesis
will be performed with the values observed and computed in random sample.
Definition of hypothesis testing is: the statistical procedures and
techniques that have as main objective the establishment of valid
conclusions for a population by studying a random sample of it,
measuring the degree of uncertainty in terms of probability.
Typeꓲ and type ∏ Errors, level of significance,
critical regions of hypothesis testing
Null hypothesis may include two different types of errors:

- error type ꓲ (alpha), that’s, 𝐻0 is true and it’s rejected.

- error type ∏ (beta), that’s, 𝐻0 is false and it’s accepted.

These types of errors are always present in hypothesis test, and when
it’s tried to diminish one the other increases. Only both diminished
when the sample size is increased.
The value with which error type ꓲ is fixed is called level of
significance and it’s denoted by Greek letter alpha (α). The
level of significance is related to the certainty or confidence.

The level of significance is could be defined as the maximum


probability of risk of committing an error type ꓲ.
For taking a decision in hypothesis test (accepting or rejecting null
hypothesis). The procedure for testing hypothesis takes this on
account when establishing the rejective region of test. This region is
named the critical region of the hypothesis test.
Testing of Mean

The hypothesis test consist in comparing the sample value (in


this case X) with a supposed or theoretical value of population
mean (µ0 ). To compare two values, the standard error of the
mean (SE x) is taken in to account, using it for the calculation
of Z. if 𝜎 (standard deviation of the population parameters) is
known, then Z calculated by:
Testing of percentage

• In medicine, most of reports and researches are dealing with


qualitative variables such as diagnosis, most frequent complications,
patients evaluations, symptoms .. Ect. The data are presented in
percentages. To take a decisions about the results it is necessary to
make a hypothesis test of percentages.
Chi- square test
Chapter 9
Introduction
In most of scientific papers and researches work, it is in deed to find
out if a variable has influence over other variables, if they have
relationship or they are independent.
The chi- square test gives that possibility.
Chi- squared test is non- parametric test, which means that the results
obtained are not in relation with population parameters. But in this
case regarding relationship among the variables.
Contingency tables, observed frequencies,
expected frequencies.
• Observed frequencies: it’s what has been observed or found
during the research or study performed.
• Expected frequencies: it is defined as the theoretical
frequencies that hypothetically should appear in the cells of
the contingency table if the variable don’t have relationship.
It is the frequencies that should be in each cell or box off the
contingency table if variables were independent, that is, if
the variables have no relationships between them.
The expected frequency is denoted by Eij, where i is number of rows
and j is number of columns, similarly as observed frequency Oij.
The expected frequency is calculated by:
(𝑛𝑖 )(𝑛𝑗)
Eij=
𝑁
Where: 𝑛𝑖 = marginal total of rows
𝑛𝑗 = marginal total of columns
N= grand total or sample size.
Wish you all the best..
Dr. Thekra Mahmoud

You might also like