This

**preview**shows half of the first page. to view the full**3 pages of the document.**Chapter 13

îPresent data in a cross-tabulation table (it compares two nominal ordinal variables at

ounceî essential for testing a hypothesis about the relationship between these two

variables

îThe # in a cell represents the frequency of joint occurrences of the categories of the 2

variables

îJoint occurrence is the combination of categories for a single individual

îColum% : is a cellâs frequency as a % or the marginal total/ Row %: is a cellâs frequency

as the row marginal total

îThe existence of a relationship between two nominal ordinal variables is established

with a hypothesis test called the âchi-square testâ

îIn a cross tab table we put the independent variable in the columns and the dependent

variable in the rows

îThe hypothesis is stated in a way that allows us to know what sampling outcomes to

expect when hypothesise is true. With a chi-square test the hypo is stated as one of no

relationship Ho:x2 =0 or Ha:x2>0

îAssuming no relationship we can use the marginal frequency to predict the expected f of

each cell

îWe compare the expected f to the observed f of each cell, and if about equal we allow

hypothesis of no relationship/ but if a large difference between expected and observed

then we expect a relationship between the variables

îThe chi-square test tells is of the difference are so great that they are not simply the

result of sampling error

îExpected cell frequency: tell us how many cases should fall in a cell if each cell is

proportional to the marginal frequency

îE cell= (column marginal total for a cell) (row marginal total for cell)/ grand total

îThe effect of the test is the difference between what is observed in the actual sample

data and what is hypothesise for the Ho

îText statisticî x2= sum (o-e)2/ e 9it measure the like hood of differences between

observed and expected cell frequencies

www.notesolution.com

###### You're Reading a Preview

Unlock to view full version