PSYCH 210 Lecture 12 Outline of Last Lecture I. Hypothesis Testinga. Use of probability to answer experimental questionsb. Steps of the processII. Hypothesesa. Nullb. Alternativei. Non-directionalii. DirectionalOutline of Current 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-testCurrent LectureI. Hypothesis Testinga. Formal Steps for z-testi. State hypothesesii. Determine cutoff areasiii. Collect data; find Miv. Make Decision about H0v. State ConclusionII. Nondirectional Hypothesis Testinga. State Hypothesesb. Determine cutoffsThese 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.α=0.05Half of p (0.05) in each tailLocate z-score for p 0.025 on tablePositive and negative z-scoresz = +/- 1.96c. Collect data: find MM = μ + zσσM=σ/√n = 5/√16=1.25M=65+1.96(1.25)62.55, 67.45d. Make decision about H0: Reject H0e. Conclusion: The advertising blitz resulted in a significant difference between mean # of FB posts about candidate from sample campaign to original population.μIII. Using a one-tailed (directional) alternative hypothesisμ = 65, σ = 5, n=16a. HypothesesH1= >65μH0 = ≤μ 65b. Determine cutoffsα=0.05Graph: tail all on one sidez=1.65c. Collect Data: Find MσM= 1.25M = μ + zσM=65+1.65(1.25) = 66M=67.06 d. Make decision about H067.06>66Reject H0e. ConclusionThe ad blitz resulted in a significant INCREASE in mean # of FB posts @ candidate from sample camp to original population .μIV. What happens when you fail to reject the null?a. Ex) Conclusion: i. The advertising blitz DID NOT result in a significant increase in mean # of FB posts @ candidate from sample camp to original population .μV. Alternative method for determining cutoffsa. Leave z-scores as is (do not translate into M)b. Step 3: convert M to z (z=[m- ]μ /σM)c. Compare z-scores to generate conclusionVI. Assumptions for a z-testa. Random Samplingb. Interval or ratio datac. Independent observationsi. Data given by one person not influenced by data of anotherd. Normal Distributione. Variance and SD unchanged by Tx (act as if applying constant to every
View Full Document
Unlocking...