0% found this document useful (0 votes)
28 views9 pages

Chi Square

ANALYSIS OF ENUMERATION

Uploaded by

Prince Hakim
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
28 views9 pages

Chi Square

ANALYSIS OF ENUMERATION

Uploaded by

Prince Hakim
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 9

ANALYSIS OF

ENUMERATION (COUNTED)
DATA
Learning objectives
By the end of this session,
participants will be able to:
 Recognize data that is

appropriate for contingency


tables
 Construct contingency tables for

counted data
 Apply the chi-square test in

analysis of contingency tables


Key concepts
 Contingency tables
 Chi-square
 Validity of test
 Collapsing
 Exact test
 Correlated counts
Introduction

 Data occurs in form of counts


 Cross-classification with

respect to two or more


variables - categorical
 Presentation in form of

CONTINGENCY Table
Data example
Study has 500 patients,
 100 given treatment 1 (70

cured, 30 not cured)


 200 given treatment 2 (160

cured, 40 not cured)


 200 given treatment 3 (168

cured, 32 not cured)


Test of hypothesis
 Null hypothesis: No association
between the two variables
 Alt hypothesis: There is association
 Reject if value of test stat > table
value
 Df = (r-1)(c-1)
 Obtain p-value
Chi-square test statistic
 Expected value
 What would be the value in each cell if
null hypothesis true?
 (Row total X Column total)/grand total
 For each cell calculate
 (O – E)²/E
 Sum up the values to get χ² statistic
 Df = (r-1)(c-1)
Are results valid?
 Problem with validity if
 Any expected value < 1
 More than 20% of E values < 5
 What to do if this happens?
 Collapse!
 Till when?
 2 by 2 table
 If problem persists see a doctor!
(USE FISHER’S EXACT TEST)
Case of correlated data
 Assumption of independence of
observations – not satisfied here
 Example
 Before & After situation
 Case control situation
 In continuous data (comparison of
means)
 Use PAIRED t-test
 In enumeration data
 Use McNemar’s χ² statistic

You might also like