This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Chapter 4 Research Hypothesis- H 1 ; predicts a difference between populations; can be directional or non-directional ; μ 1 ≠μ 2 Null Hypothesis- H ; predicts no difference between populations; μ 1 =μ 2 Population Distribution- the amount of people in a certain area Type 1 Error- reject the null hypothesis, but the null hypothesis is true We say μ 1 ≠μ 2 , when actually μ 1 =μ 2 The chance of making a Type 1 error is the same as the level of significance of the test- .05 significance level 5% chance of making an error Alpha Error (α)- p<.05, α<.05 Type 2 Error- we accept the null hypothesis, when the null hypothesis should be rejected We say μ 1 =μ 2 , when actually μ 1 ≠μ 2 Beta Error (β)]- The more lenient the significance level, the smaller the chance for making a beta error Relationship Between Type 1 and Type 2 Error- • If you protect against Type 1 error, chance of Type 2 error increases • If you protect against Type 2 error, chance of Type 1 error increases Protecting against one, increases the chances of the other occuring Directional Hypothesis- when a clear direction is stated (“increase”, “higher”, “lower”); 1 tailed test; μ 1 >μ 2 or μ 1 <μ 2 Non- Directional Hypothesis- mixed theory, no direction prediction; 2 tailed; μ 1 ≠μ 2 Rationale for Hypothesis Testing- we want there to be differences between populations Conventional Levels of Significance- 5% (p<.05) & 1% (p<.01) Chapter 5 Population Distribution vs. Distribution of Means * Individual Scores * Means of groups of individual scores * Used when comparing...
View Full Document
- Fall '07
- Statistics, Null hypothesis, Statistical hypothesis testing, Statistical significance, Statistical power, Effect size