Introduction to Statistics in Psychology Greg Francis Fall 2010 MWF 2 30 3 20 pm Peirce Hall Room 277 http www psych purdue edu gfrancis Classes PSY201 index html Introduction to Statistics in Psychology PSY 201 Professor Greg Francis Lecture 01 How to avoid being a schmuck MAKING JUDGMENTS A PROBLEM we have to make judgments all the time people are not very good at answering these kinds of questions we make systematic errors take PSY 200 PSY 285 or PSY 318 Do nicotine patches help people stop smoking Is Pepsi better than Coke How is alcohol consumption related to depression Is this a good buy for a stereo Is diabetes related to arsenic people in the know can take advantage of these tendencies politicians retailers drug companies activists 2 3 SIGNIFICANCE AVOIDING COLDS A GOOD STUDY people are easily influenced by the context in which they make decisions Many people recommend the herb echinacea to reduce the severity of colds and or to increase your immunity to getting a cold there appears to be a link between low levels are arsenic and type 2 diabetes this is a problem because context is easily and subtly manipulated to avoid being a schmuck you need to learn how to make decisions properly What should happen if echinacea does work STATISTICS What should happen if echinacea does not work it is not always easy but it is worth it Why is it popular 4 5 788 adult medical tests found a fourfold increase in the risk of diabetes for those exposed to low levels of arsenic in their urine compared to people with even lower levels of arsenic previous studies have found high levels of arsenic in some drinking water a follow up study with 4000 people is planned 6 A GOOD STUDY A BAD STUDY A BAD STUDY But even a good study has limits The Daily Express subheading The World s greatest newspaper ran a story claiming statistical evidence that suicides were higher around mobile phone masts According to the quoted researcher masts are placed an average of 800 meters from each home across the country while for 22 highly publicized suicides the masts were an average of only 356 meters away arsenic comes from lots of different sources drinking water coal burning seafood it s not known how arsenic could contribute to diabetes But what does this actually mean it could be that people with diabetes tend to retain arsenic in their urine How was average distance computed the press CNN tries to scare people about the issue Mobile phone masts are not distributed evenly across the countryside so is an average even a reasonable measure of anything no indication of the actual risk numbers Maybe more suicides are in urban areas which also happen to have more mobile phone masts If you look long enough you can find something in common among any number of people 7 8 9 COURSE GOALS COURSE GOALS WHY IS IT HARD What are the differences between good and bad studies What are the limitations of good studies What is needed to get good experimental data How is statistics used in psychological sciences 1 Descriptive statistics several reasons How to describe data Using graphs How to summarize data 1 Little differences in presentation can make a big difference in understanding Hypothesis testing Comparing descriptive statistics 3 It involves mathematics 2 Inferential statistics 2 It is hard to get good measurements 4 It goes against our intuitions anecdotal evidence 5 If you don t ask the right type of question it is worthless 6 Sometimes the answer is I don t know 10 11 12 COURSE OUTLINE TEXT TEXT statistical terms describing data identifying relationships EXAM I 10 Online Statistics An Interactive Multimedia Course of Study provides a pretty good introduction to many topics we want to discuss designed for the behavioral sciences sometimes weak in presenting exactly how to carry out the computations probability hypothesis testing estimation EXAM II 10 hypothesis testing statistical power EXAM III 10 ANOVA FINAL 15 cumulative free You will earn course credit by finishing reading assignments by the due date given in the syllabus 10 of your class grade You must login to earn the course credit and stay logged in for the specified duration required to read the material rarely more than 10 minutes our discussion follows a different order so we jump around the chapters a bit If you log off or point the web browser to a different page you have to start over Beware scheduling of the final exam 13 14 15 LECTURE NOTES HOMEWORK HOMEWORK reduced format of 6 slides to a page homework problems are available for download from the class web site grading will focus on whether you are doing the problems correctly not just on whether you get the right answer make sure that you show that you know what you are doing available before lecture available on the world wide web in Portable Document Format readable by Adobe Acrobat due on the day specified in the syllabus if turned in by the next class it is worth 1 2 credit any other time is worth no credit http www psych purdue edu gfrancis Classes PSY201 index html total homework grade is worth 20 of your class grade necessary for some topics we will discuss 16 17 18 STATLAB STATLAB COMPUTERS a new part of the class is a set of on line experiments that measure some aspect of your behavior You will earn course credit by finishing a lab assignment by the due date given in the syllabus 10 of your class grade some software packages handle lots of statistics at the end of each experiment you are guided through a statistical analysis of the results You must login to do the labs You need to know what you are doing to complete the statistical analysis When you finish the lab answer all the statistical question correctly you get credit for that lab I don t recommend you use them to complete the homework problems because you need to show intermediate steps a spreadsheet program like MS Excel is pretty useful for computing terms quickly but be sure you know how to compute things by hand Excel is especially useful for the StatLab assignments free you will not be able to use the packages on the exams 19 20 21 CLASS PARTICIPATION GRADING IMPROVING YOUR GRADE every now and then I require assistance during class straight scale no rounding Attend class and do the homework correctly I will go through the class list and pick names at random I will note if you are present or not and that will count toward your class participation grade Your total class participation grade is worth 10 of your final class grade I view