PSYCH 210 Lecture 13 Outline of Last Lecture I. Hypothesis Testinga. Formal Steps for z-testb. Nondirectional Hypothesis Testingc. Directional Hypothesis Testingd. What happens when you fail to reject the null?e. Alternative method for finding cutoffsII. Assumptions for a z-testOutline of Current Lecture I. Hypothesis Testing and z testsa. Type I and II errorsb. Powerc. Factors affecting powerCurrent LectureI. Type I, Type II error & Powera. Error built into Hypothesis Testing systemi. System based on probabilityii. MatrixState of the World:H0 TRUE (no effect) H0 FALSE (Tx does have effect)Decision Reject H0TYPE I ERROR CorrectThese notes represent a detailed interpretation of the professor’s lecture. GradeBuddy is best used as a supplement to your own notes, not as a substitute.*POWERFail to Reject H0Correct TYPE II ERRORb. Type I errori. Def: Saying Treatment has an effect when it actually doesn’t (false positive)c. Type II errori. Def: Missing a real treatment effect (false negative)d. Statistical Poweri. Def: Treatment actually works and we also recognize that it does1. Probability we will make the correct decision in rejecting the null2. Ex) Microscope: a higher powered microscope allows one to find cancerous cells more easilye. How can we set up experiments so that error is low and power high?i. Four factors that influence power1. Effect Sizea. How big of a change Tx has on meanb. On graph = distance between two peaks of curves (H0 and H1)i. μ before Tx – μ after Txc. Experimenter does not have much control over thisd. Graphe. Effect of Increasing Effect Size:Type I error (α)Not changing*Power (1-β)IncreasesType II error (β)Decreasesi. *Only time α changes is when experimenter changes it!ii. Type II error and Power are inversely related2. Alpha Levela. Graphb. Effects of Decreasing Alpha Level:Type I ErrorDecreasesPowerDecreasesType II ErrorIncreases!3. Choice of one-tailed vs. two-tailed testa. Graphb. Effects of Changing from two-tailed (nondirectional) to one-tailed (directional):Type I ErrorNo change*PowerLargerType II ErrorSmalleri. *Alpha considered in terms of TOTAL alpha area4. Sample size (n)a. Continue next
View Full Document