Professional Documents
Culture Documents
Chapter 13 - Lecture 1 Two-Way Contingency Tables: Andreas Artemiou
Chapter 13 - Lecture 1 Two-Way Contingency Tables: Andreas Artemiou
Introduction
Test for independence of two categorical variables
Test for homogeneity of proportions
Exercises
Chapter 13 - Lecture 1
Two-way contingency Tables
Andreas Artemiou
Introduction
Review
Exercises
Review
Notation
Hypothesis test
I Null Hypothesis: H0 : Variable 1 and Variable 2 are
independent
I Alternative Hypothesis: HA : not H0
k
X (Oi Ei )2
I Test statistic: 2 =
Ei
i=1
I Rejection Region: 2 2,(I 1)(J1)
I The conditions so that the test is valid, is to ensure that in
every cell Ei 5.
I If this is not true then we can merge cells together.
I To calculate the test statistic you need to construct two
Tables. The Table of observed counts and the Table of
Expected counts
Andreas Artemiou Chapter 13 - Lecture 1 Two-way contingency Tables
Outline
Introduction
Test for independence of two categorical variables Hypothesis testing
Test for homogeneity of proportions
Exercises
Hypothesis test
I We ask 1000 students in PSU what color they like the most
among the three: black, yellow, green. The students are
divided into 5 populations. Freshmen, Sophomores, Juniors,
Seniors, Graduates. Freshem answered 150, 80, 20.
Sophomores answered 130, 60, 60. Juniors answered 100, 70,
30. Seniors answered 100, 40, 60 and Graduates answered,
30, 40, 30. Are the 5 populations Homogeneous in terms of
color preference at significance level 0.01?
Exercises