Parametric vs' Nonparametric Tests

1
Parametric vs. Nonparametric Tests

• Parametric
• Testing hypotheses about parameters, e.g., value
  of m, m1 m2, etc.
• Assumptions about population distributions, e.g.,
  normal, t, F
• Assumptions about values of parameters, e.g., s1
  s2

2
Parametric vs. Nonparametric Tests

• Nonparametric tests
• No distribution assumptions
• Useful where DV not quantitative

3
One Variable Chi Square

• Goodness of fit test
• Data in the form of frequency of observations in
  two or more categories
• Example Faculty survey on 1A football

4
Chi Square

• Survey Results
• Category fo
• Yes 27
• NO 44
• No Opinion 29
• What are the expected frequencies?

5
Chi Square Test

• http//www.graphpad.com/calculators/chisquared1.cf
  m

6
Chi Square Test for Independence

• Does faculty preference on 1A football depend on
  sex of the respondent?

7
Calculation of Chi Square

• http//home.ubalt.edu/ntsbarsh/Business-stat/other
  applets/Catego.htm

8
Expected Frequencies

• Derived from empirical probabilities
• P(Male) 107/200 .535
• P(Yes) 83/200 .415
• P(Male and Yes) P(Male) x P(Yes) .222
• Expected frequency for Male Yes
• P(Male) x P(Yes) x N 44.405

9
Expected Frequencies
10
Chi Square Requirements

• Independent Observations
• Expected frequency for a category or cell should
  be at least 5