Scribd
Upload
369 views12 pages

Bivariate Analysis PDF

This document discusses bivariate analysis and statistical techniques used for exploring the association between two variables. It defines bivariate analysis and provides examples of research questions it can address. Depending on whether the variables are qualitative or quantitative, different statistical tests are used, such as cross-tabulation and chi-square test for two qualitative variables or correlation for two quantitative variables. Guidelines are provided for selecting variables and interpreting results.

Uploaded by

Hygienix Fsd
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
369 views12 pages

Bivariate Analysis PDF

This document discusses bivariate analysis and statistical techniques used for exploring the association between two variables. It defines bivariate analysis and provides examples of research questions it can address. Depending on whether the variables are qualitative or quantitative, different statistical tests are used, such as cross-tabulation and chi-square test for two qualitative variables or correlation for two quantitative variables. Guidelines are provided for selecting variables and interpreting results.

Uploaded by

Hygienix Fsd
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/ 12

1/13/2020

Bivariate analysis: cross-table


and Chi-square test
Prof. Dr. Rubeena Zakar, MBBS, MPS,
PhD

Outline

• What is a bivariate/bivariable analysis?


• Statistical techniques for bivariate analysis
• Questions relevant to bivariate analysis
• Some hints about selection of variables

1
1/13/2020

Bivariate analysis
• Bivariate analysis is a kind of data analysis that
explores the association between two
variables (X, Y).
or
• Bivariate analysis is the simultaneous analysis
of two variables

Bivariate analysis: Examples


• Does the average score of the examinations differ by sex of
the students?
• Is the association between education and income
significant?
• Are years of experience related with income of the people
working in an industry?
• Are status of drinking alcohol associated with number of
auto accidents?
• Are full-time workers more likely to be unionized than part-
time workers?
• Is there any association between BMI and high blood
• pressure?

2
1/13/2020

• Depend on the composition of the pair of


variables:
o Are both variables qualitative?
o Are both variables quantitative?
o Is one qualitative and another quantitative?
o Are both variables ordinal?
o Is one variable dichotomous and one continuous

Bivariate analysis
• Statistical techniques for bivariate analysis
depend on:
– measurement scales of the variables
and
– distinction between dependent and independent
variables.

3
1/13/2020

Bivariate analysis: Statistical


techniques

Bivariate analysis

4
1/13/2020

Cross table
• Cross table is a two-way table consisting of rows
and columns
• Typically used to determine whether there is a
relation between row variables and column
variables
• As a basic rule, independent variable is on the X-
axis while the dependent variable is located on
the Y-axis
• Normally they have a grand total, row totals,
column totals, cell value

Statistical test for bivariate association


Depend on the composition of the pair of
variables:
– Are both variables qualitative?
– Are both variables quantitative?
– Is one qualitative and another quantitative?
– Are both variables ordinal?
– Is one variable dichotomous and one continuous

5
1/13/2020

Statistical test for bivariate association


• Two qualitative variables
– Use cross table and apply chi-square test
• Two quantitative variables
– Use a correlation matrix and apply Pearson’s
correlation co-efficient (r)
• One qualitative and one quantitative variable
– Use a comparison of means and apply t-test

Hypothesis
• Null hypothesis (H0): The two variables are
independent (not associated)
• Alternative hypothesis (H1): The two variables
are dependent (associated)
• Note: Rejection of null hypothesis means
acceptance of alternative hypothesis

6
1/13/2020

Example of cross table

Chi-qaudrat-tests

7
1/13/2020

Chi-square statistics
• Examine a bivariate relationship using two
qualitative variables
• Perform a cross-table analysis and check the
chi-square statistic
• Interpret the Chi-square statistic
• Are they related?

Statistical significance

8
1/13/2020

Example: Cross table (hypothetical)

Percentage in a table

9
1/13/2020

Percentage in a table

Percentage in a table

10
1/13/2020

Example: Cross table (hypothetical)

Selection of variable
• We should select only those independent
variables which may:
– Have some causal effect on the dependent
variable
– Be correlated with each other

11
1/13/2020

Guidelines
• Select dependent (outcome) variable with possible categories (maximum
3 categories).
• Choose potential independent variables (10-20 variables) from your data
set which may be logically associated with your outcomevariable.
• Recode your variables (if necessary) with meaningful categories (not many
categories).
• Perform cross-table (bivariable) analyses to check the association of
independent variable with dependent variable. If the p-value (based on
Chi-square test) is ≤0.20 for any association, choose this variable for
further analyses.
• If there are many variables which fulfill your inclusion criteria (i.e. p≤0.20),
select around 15 variables which are more important.
• Perform multivariable analysis for your outcome variable (next class). Use
both graphs and tables to present your findings. All graphs and tables
should be self-explanatory with basic information.
• Save the syntax of your analyses

12

You might also like