Unit 5: Nominal & Ordinal Tests
|• Most research is based on samples. |
• Generally, samples represent their populations, and contain sampling errors.
5.2 Sampling Techniques
|• Simple Random Sampling|
• Stratified Random Sampling
• Proportional Stratified Random Sampling
• Convenience Sampling
5.3 One-way Chi-square Test
|• Known as a goodness of fit test, and often used in preference studies.|
• Compares observations and expectations.
• Data must be categorized in one way, and must be nominal or ordinal.
• Use the table of critical values to determine significance.
5.4 Two-way Chi-square Test
|• Used for nominal or ordinal data that is categorized in two ways.|
• Same formula and critical values as used in a one-way chi-square.
• Must first calculate expected values.
5.5 Cramer’s Phi
|Used to indicate the strength of the relationship in a two-way chi-square analysis.|