Introduction to Statistics in Psychology: PSY 201Greg Francis, PhDDepartment of Psychological SciencesPsychological Sciences Building, Room 3174, (765) 494-6934email: [email protected]://www.psych.purdue.edu/∼gfrancis/Classes/PSY201/index.htmlStudy Guide for Exam 3Exam Date: November 22, 2010 (during the regular class period)The exam will consist of 10-15 short-answer questions that will be similar to theproblems on the homework (but usually smaller). However, some of the questionswill be essay questions where you must describe a concept. You will be expected toperform some calculations, so bring a calculator. If any tables will be necessary, Iwill provide them.I will not provide an equation sheet with the exam. Instead, you should create a“crib sheet” consisting of one piece of paper (both sides can be used). My advice isto (at least) write-out, step-by-step, what you do to carry out a hypothesis test andto create a confidence interval for every statistic we have discussed. The crib sheetmust be hand written (no photocopies). I will be happy to look over your crib sheetbefore the exam to check for any mistakes.Know all of the following information. Generally you need to both know how to dovarious calculations and understand the concepts behind the calculations. The formeryou can verify by doing practice problems. The latter you can verify by creating awritten description of what the calculations are doing. If the written descriptionmakes sense, then you probably understand the concept.In general, you need to be able to do the calculations discussed in lecture andon the homework assignments. For conceptual issues, you need to understand theconcepts discussed in lecture, and the readings will often provide material that willhelp you do that.Lecture 23.1. Understand the need for the Fisher z transform when testing correlations.2. Know the sampling distribution of Fisher z transform scores and the formulafor standard error of zr.3. Know how to run hypothesis tests for correlations using the Fisher z transform.4. Know how to run a hypothesis test for the special case with H0: ρ = 0.Lecture 24.1. Know how to build a confidence interval using Fisher z transform scores.12. Know how to run hypothesis tests for proportions. Know the sampling distri-bution and the formula for standard error.3. Know how to build a confidence interval for a proportion (use the formula forstandard error based on little p and q).Lecture 25.1. Understand the distinction between a one-sample test and a two-sample test.2. Be able to run a two-sample hypothesis test for means, when homogeneity ofvariance is assumed. Know the sampling distribution and the calculation ofstandard error. In particular, know how to pool the sample variances from thetwo samples (using any of the formulas). Know the degrees of freedom.Lecture 26.1. Be able to test for homogeneity of variance. Know the sampling distribution,the test statistic (F -ratio) and how to get each of these terms. Know what theconclusion of the test means about running a hypothesis test for means.2. Be able to run a two-sample hypothesis test for means when homogeneity ofvariance is not assumed. Know the formulas for standard error and degrees offreedom.Lecture 27.1. Understand the distinction between a two-sample case with independent anddependent means. Be able to recognize which case applies to a given situation.2. Be able to run a two-sample hypothesis test for dependent means. Know thetest statistic, the sampling distribution, and the calculation of standard error.Lecture 28.1. Be able to carry out a two-sample hypothesis test involving correlations.2. Be able to carry out a two-sample hypothesis test involving independent samplesfor proportions.3. Know what margin of error refers to.Lecture 29.1. Be able to carry out a two-sample hypothesis test involving dependent samplesfor proportions.2You need to know how to run every type of hypothesis test that we have discussed.1. One-sample correlation (with Fisher z transform): H0: ρ = a [and how tocreate a confidence interval]2. One-sample correlation (special case): H0: ρ = 0 [and how to create a confi-dence interval]3. One-sample proportion: H0: P = a [and how to create a confidence interval]4. Two-sample independent means, homogeneity of variance: H0: µ1= µ25. Two-sample independent means, no homogeneity of variance: H0: µ1= µ26. Test for homogeneity of variance: H0: σ21= σ227. Two-sample dependent means: H0: δ = 08. Two-sample independent correlations: H0: ρ1= ρ29. Two-sample independent proportions: H0: P1= P210. Two-sample dependent proportions: H0: δP= P1− P2=