Psych 7 Final Study Guide
Psych 7 Final Study Guide 7
Popular in Intro to Experimental Psych
Popular in Department
This 8 page Study Guide was uploaded by Debbie Chen on Friday December 12, 2014. The Study Guide belongs to 7 at University of California Santa Barbara taught by PROTZKO in Fall2014. Since its upload, it has received 99 views.
Reviews for Psych 7 Final Study Guide
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 12/12/14
PSYCHOLOGY 7 FINAL STUDY GUIDE JOHN FOLEY Times amp Dates 0 Tuesday 845 Foey s OH Review Psych 3834 0 Wednesday FINAL 1200 Test 0 Cumulative on what we missed during the midterm amp final 0 Parscore This review guide contains 0 Week 10 Lecture Notes 0 Outline of important concepts from post midterm2 Week 10 Bayesians Office Hours Psych 3834 Tuesday 845 1045 AM John 0 PDH0 0 Null Hypothesis testing 0 How likely would the outcome have happened randomly 0 Cancer problem 0 If you want to know you have a specific type of cancer 0 1 rate 0 You take a test that is 90amp accurate 39 If you have cancer the rest has a 90 chance of coming out positive 7 FP You get a positive result What is the likelihood you have the cancer 125 Why OOOO 0 Bayesian tells you the probability of the hypothesis given the data 0 Baserate is the probability of the hypothesis 0 And then you get a test result with Bayesians Breast Cancer 0 Average cancer rate 00628 0 Probability of Sam having breast cancer 0 00628 0 Positive mammogram 0 False positive How likely are you to have this data if your hypothesis is wrong 6 0 True positive How likely are you to have this data if your hypothesis is right 78 0 Mammogram 0 False positive 60 0 True positive 78 0 PH 0000629 0 PDH 078 0 PDH 06 0 PD 0600113 0 PHD 78 x 0000629 600113 00008 0 PHD PDH x PH PD 0 Probability of cancer Probability of true hypothesis x general probability of true false Sam amp Baserates 0 Sam 01243 0 PHD 0001615 0 Samuel 00013 0 PHD 000017 0 Shows us that base rates really affects things 0 Things that are really rare are really rare no matter what your results for a test are 0 BASE RATES MATTER A LOT Bayesians 0 Gives you the probability of your hypothesis given the data 0 Take Home Message 0 Base rates matter 0 You probably don t have that super rare disease 0 These are degrees of belief 0 What does this mean NHT only gives you PDHo Bayesianism lets you start talking about your hypothesis PHD Even though 80 of shark attacks happen in less than 3 feet of water it doesn39t matter bc most people are in less than 3 feet of water 0 Not impressive 0 Bayesianism in research 0 They don t use the equation 0 What they use is Bayes Factor comparing one hypothesis to another hypothesis given the same data 0 Prior before data what is the probability of your hypothesis being true 39 How I set my priors matter 0 Bayes Factor 39 One over the other 39 H2 could be the null or a competing hypothesis 39 Bayes Factor 30 40 counts as a good enough evidence 39 Is that any less arbitrary than plt005 How do we must treat Major Depressive Disorder 0 We know SSRI help 0 What about exercise 0 RCT 101 patients with mild depression 0 Placebodrugexercise group 0 Exercise amp Serotonin both helped depression 0 But no difference between the groups 0 P 607 tells us we can39t interpret it doesn39t tell us how much 0 But Bayesian statistics you can how much better is exercise than SSRI 0 In general 0 Bayesianism says a good hypothesis is something that takes super extraordinary data and makes it ordinary Pay attention to base rates Bayesianism comes in degrees 01 no such thing as believe and don t believe Belief cannot be 0 or 1 PH 0 then PHD 0 PH 1 then PHD 1 OOOOO 0 Makes the equation unchangeable 0 Problems 0 Declaring PH 39 Uninformed priors increases Bayes Factor 0 Subjective 0 Endless ways to be wrong Lecture 2 Research Ethics The Tuskegee Syphilis Study 0 Progression of untreated syphilis 0 Placebo treatment 0 Didn t give penicillin because they wanted to know progression Forced Sterilization of the mentally unfit 0 Started in America 0 Upheld by the supreme court 0 Technically still legal today Nazi tested for typhus fever 0 Either get a vaccine or placebo then injected with infected blood 0 Nuremberg Code voluntary informed consent valid research that could provide fruitful results 0 Must volunteer and give informed consent 0 Deceptionomit can be okay 39 le bystander effect 39 But must have good reason Willowbrook Hepatitis Study 0 What is the progression for untreated hepatitis 0 Used mentally defective children 0 Tested effect of gamma globulin in treatment 0 Closed doors to new inmates 0 Can accept if new inmates if part of study 0 Cannot force people to be part of your study 0 Prisoners 0 Cognitively Impaired 0 Minors 0 7 child and adult 0 7 adult Tea Room Study 0 Bathrooms for homosexual sex 0 Stood outside and recorded license plates 0 Went to their home and disguised themselves amp went through social health survey 0 Wrote book families of men found out 0 Must disclose privacy and confidentiality Milgrim Study 0 Must have debriefing 0 If you deceive you must debrief Data Faking 0 Andrew Wakefield 0 MMR Autism 0 Fake Data 0 Trying to bring lawsuits against drug companies 0 Jan Hendrik found to fake data 0 No one could replicate it 0 Too perfect data 0 Diederik Stapel 0 Fake data 0 The people he worked with came forward 0 Must rely on scientific integrity 0 Inability to replicate Important Concepts Longitudinal Design 0 Cohort Cohort effects same people 0 No manipulation 0 Advantage fewer people sensitivity individiual differences 0 Disadvantage order effects dropout 0 Predicting ONLY ABOUT PREDICTING NOT CASUALITY 0 Controlling for making the slope for pretest equal changes based on observed scores 0 Multiple regression used when we want to predict the value of a variable based on others 0 Example predicting what we will look like in the future 0 High causality with no actual causality are called spuirious correlations Case studies 0 Phineas Gage 0 Genie 0 Frere Jacques 0 Split brain patients 0 GeneralizabilityLaw of large numbers 0 Solutions replicate more numbers ignore individual differences mechanism 0 Single Case Research Design 0 Multiple Baseline 0 Stable Baseline 0 NaturalisticSystematic Observations Statistical Significance 0 Hypothesis 0 ProbabilityDataNul Hypothesis 0 P value tells you the probability that even if the null hypothesis is true the chance in which you got the result randomly 0 Rejecting null hypothesis ONLY TELLS YOU THERE IS AN EFFECT 0 The higher the P the lower the T 0 P effect sizeN 0 Cohen39s D tells you the overlap between two groups in SD units 0 Type I amp Type II Error 0 P hacking is when you change the statistics to make it significant like using a lot of DV 0 Needs to be replicable 0 Power is the ability to correct reject 0 Type III is when it39s in the wrong direction 0 Parametric aassumptions about populationNonparametric nomnal 0 Interactions are the differences between main effects 0 IV Main effect 0 Levels within main effect 0 Confound alternate explanation 0 Moderator conditional cause 0 NEVER DICHOTOMIZE A CONTINUOUS VARIABLE Bayesians 0 Bayesian probablity of hypothesis given the data 0 Baserate probability of hypothesis before the data 0 True positive how likely the data is if your hypothesis is right 0 False positive how likely you are to have this data if your hypothesis is right 0 Base rates matter a lot 0 Degrees of belief in Bayesianism 0 Bayes Factor compares two hypotheses given the same data Ethics 0 Tuskegee Syphilis Study 0 Willowbrook Hepatitis Study 0 Tea Room Study 0 Milgrim Study
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'