Psychological Statistics - Courses in Psychology - UCLA

Psychology 100A: Psychological StatisticsWinter 2009Meetings:Class: TR 11:00 – 12:15, Chemistry (Young) 76Discussion: (optional) F 12:00-12:50, Franz 1178Instructor:Sean McAuliffeOffice hours: Tue 4-5 & Thu 10:00-11:001346 Franz Hall (middle psychology building behind the vending machines)EMAIL: vpstatman@hotmail.comTeaching Assistant:Simon Xinming, Email: xinming.an@gmail.com, Office hours: Mon 1-2 & Wed 1-2Textbook:No textbook is required for this course. All of the material for the exams will be coveredin class, but some students find a textbook helpful for reinforcing the concepts in the class. Oneproblem is that the textbook will cover additional topics not presented in class, and in some caseswill use different terminology which some students find confusing. The recommended OPTIONALtext is “Statistical Concepts for the Behavioral Sciences (3rd Edition)” by Harold KIESS (availablein the UCLA bookstore).To facilitate learning, comprehensive review materials will be posted on the course webpage for every lecture. In addition, the optional Friday discussion will be used for further reviewevery week. There will be no new material introduced in the Friday reviews, so you are notrequired to attend, but many students find them helpful.Website:www.thebeststatistics.info is a comprehensive website that provides overviews, reviews andstep-by-step problem solvers for many of the types of problems we will go over in class and thatmay appear on the exams. Importantly, the website also provides visualizations to increase learningbecause understanding statistics is often facilitated with the right visual aide. The site costs $30 for3 months. Subscribing to the site is NOT required for the course, but students in the past havefound it very useful. In fact, the website will frequently be used in class to show how to solveproblems as well as visually demonstrate important concepts.Overview: In this course, we will cover the basic data-analysis concepts and procedures usedby research psychologists. The course will briefly review ways of describing data (descriptivestatistics) and methods of evaluating hypotheses and testing psychological theories (inferentialstatistics). Examples will be drawn almost exclusively from the psychological literature.Specific topics include correlation, t-tests, and ANOVA.

Course Requirements and Grading: It is very important that you attend all classes in orderto keep up with the material. Your grade will depend on your performance on the midterms andfinal exams. If you miss some lectures and get lost, you will be in serious trouble becauseconcepts build on other concepts. Statistics is cumulative but the exams emphasize recentmaterials. All exam material will come from lectures.Grading basis:Midterm 1 (Thu January 29) 30 points (30 Fill in the blank questions/problems)Midterm 2 (Tue March 03) 30 points (30 Fill in the blank questions/problems)Final (Fri March 20) 60 points (60 Fill in the blank questions/problems)As you will learn in the course, the distribution of scores can be described in a standardizedway. A Z-score is a standardized score which indicates how far a particular score is away fromthe mean (or average) score in terms of standard deviations (a measure of how wide adistribution is) Exam scores that are above the mean will have a positive Z-score, exam scoresbelow the mean will have a negative Z-score. Your average Z-score on all the exams willdetermine your grade The final exam will count twice as much as the other exams in thisaverage. Below is the correspondence between the average Z-score and the grade you willreceive. Next to the Z score is the approximate % of the class that will get that grade.Average Z-score Grade Approximate % of class receiving score>1.30 A 10>0.85 A- 10>0.55 B+ 10>0.25 B 10>0.00 B- 10>-0.25 C+ 10>-0.85 C 20>-1.30 C- 10>-1.70 D 5

Your grade is up to you. While the course is going on, I will do as much as I can to make sureyou learn as much as you can about this material. If you want to learn more or you are confused,COME TO OFFICE HOURS I may be ugly, but I am not that scary in my office.At the end of every quarter after the final, there are inevitably a handful of students who beg mefor a way to increase their grade because they didn’t seek help when they had the chance. Myanswer to these requests to change grades is always the same – No. No. No. No. No. It would begrossly unfair to all students if I ever allowed anything but the scores determine the grade.Fairness is very important and fairness is the reason why a student will NEVER be able toincrease their grade after the course is done.Missing-Exam: Under certain EXTREME conditions (severe medical illness with completemedical documentation or death of an IMMEDIATE family member), ONE missing midtermexam can be substituted with the lowest Z-score of the remaining exams. In this way, studentswho are consistent good performers need not be hurt by a missing exam.For example:Exam 1 Z=1.0Exam 2 Missing due to EXTREME case of EbolaFinal Exam Z = 1.2In this case, the missing exam would be scored as z=1.0 because 1.0 was the lowest z-score ofthe remaining exams. In this case, the average z score would be (1.0+1.0+1.2+1.2)/4 = 4.4/4 =1.1 which would be an A-.Missing the FINAL exam will result in an F.Only ONE of the 2 midterm exams can be missed.No incompletes, extra credit, etc. are given in 100A.Exam difficulty: The exams are DIFFICULT The scores are curved so this is really not aproblem. In fact, hard exams are good because they allow the students who know more todistinguish themselves from those who don’t know very much An exam that is too easy doesnot allow for a wide enough spread in the distribution. Some of the exam questions will besimilar to questions in the book or examples done in class. Some of the questions will requireyou to understand the concepts well enough to relate them to each other in a potentially novelway.Exam Completion: Your answers on the exam will be written on an answer sheet. It is yourresponsibility to have these answers written down by the end of the exam period. Why are wetelling you this? Because if we don’t, some students will wait until the end of the exam totransfer their answers from their exam to the response sheet, which would effectively give themmore time than the other students. That’s simply not fair and it will not be allowed. Whentime runs out during the exam, students will not be allowed to write anything more on theirresponse sheet.

Why are the exams not multiple choice?: Because multiple choice exams areunfair. They are unfair because they are greatly affected by guessing. The examplebelow shows how random guessing has a dramatic effect on test scores even whenknowledge is equivalent. The dark bars are all students who know 20 out of 30questions and guess on the rest of the questions. The lighter bars are students whoknow 15 out of 20 and guess on the rest. Notice how large the differences are on amultiple choice test between students who have equal knowledge. On a fill-in theanswer test, all students who know 15 questions will get a score of 15 and allstudents who knows 20 answers will get a score of 20. In addition, life is not amultiple choice.65432Knows 15Knows 201015(F)16(D)17(C-)18(C )19(C+)20(B-)21(B+)22(A-)23(A)24(A)Cheating: We hate cheaters and will do everything possible to get you kicked out of school ifwe catch you. No kidding here – cheating is serious business and if we catch you, then you arehistory. Note – we have had a lot of experience at catching cheaters, so just don’t do it. Despitewhat you may have heard, UCLA will kick you out of school for a couple of semesters (ormore) if you caught cheating.

Calculator Type:Fairness is very important in this class. The only way to be fair is to require thateveryone has EXACTLY the same calculator. Because supplies are often limited, I will bringa supply of calculators to class. The calculator will cost $5. Many students will be annoyed thatthey have to buy a $5 calculator when they already have one, but there is no way of making surethat different calculators have the same functionality (we can’t check 150 calculators during theexam). Remember, you are not required to buy a textbook, which costs far more than $5.Any other calculators will not be allowed – NO EXCEPTIONS. In fact, you will lose5 points on exam if you do not have the official calculator – your unofficial calculator will betaken during the exam period and you will be given an official calculator for the exam. 5 pointsis a lot of points. Calculators will be available after class throughout the term.Formula Sheets: No formula sheet will be given out for use during the exams. Studentsoften complain that memorizing a bunch of formulas is stupid because it is not conceptual.Unfortunately, this thinking is wrong. As you will learn in this course – the formula cannot beseparated from the concept and in some cases the formula IS the concept.Schedule:Date Material LectureJan 06 tue Introductory Materials 1Jan 08 thu Central tendency and variability 2Jan 13 tue Normal Distribution 3Sampling Distributions, ConfidenceJan 15 thu Intervals 4Jan 20 tue Introduction to hypothesis testing 5Jan 22 thu More hypothesis testing 6Jan 27 tue Review for Exam 1Jan 29 thu MIDTERM EXAM 1Feb 03 tue two sample independent T-tests 7Feb 05 thu two sample dependent T-tests 8Feb 10 tue power, experimental design 9Feb 12 thu Power/one-way ANOVA (between) 10Feb 17 tue one-way ANOVA (between) 11Feb 19 thu one way ANOVA (within) 12Feb 24 tue two-way ANOVA (between) 13Feb 26 thu Review for Exam 2Mar 03 tue MIDTERM EXAM 2Mar 05 thu Correlation 14Mar 10 tue Regression & Chi-square 15Mar 12 thu Final reviewMar 20 Fri FINAL EXAM 11:30-2:30 in class