Department

GeographyCourse Code

GGR270H1Professor

Damian DupuyStudy Guide

FinalThis

**preview**shows half of the first page. to view the full**2 pages of the document.**GGR 270 – Lecture 12 – December 1, 2010

Non-Parametric

Nominal/Ordinal data or categorical data

Normality of population distribution is unknown

Use non-parametric tests:

oGoodness of fit

oContingency table or cross tabulated data

Chi square or x2

Most common

Relationships between cross-tabulated variables

Using observed frequency counts of a variable

oLooking for significant difference between actual and expected frequencies

Null hypothesis assumes no significant difference between actual and expected

frequency counts

Uses x2 distribution that is by default one-tailed

Assumptions

oNominal/ordinal

oAt least 2 mutually exclusive categories

oWhen only 2, expected frequency in each must not be less than 5

oWhen more than 2 categories, no one category should have an expected

frequency of less than 1, and not more than 1 category in 5 should have an

expected frequency of less than 5

oWe may need to amalgamate categories to achieve aussumptions

www.notesolution.com

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

Unlock to view full version