New User Special Price Expires in

Let's log you in.

Sign in with Facebook


Don't have a StudySoup account? Create one here!


Create a StudySoup account

Be part of our community, it's free to join!

Sign up with Facebook


Create your account
By creating an account you agree to StudySoup's terms and conditions and privacy policy

Already have a StudySoup account? Login here

Exam 1, 2, and 3 Study notes and Lectures

by: Jimbo Notetaker

Exam 1, 2, and 3 Study notes and Lectures

Marketplace > University at Buffalo > Psychlogy > > Exam 1 2 and 3 Study notes and Lectures
Jimbo Notetaker
GPA 3.89
Scientific Inquiry

Almost Ready


These notes were just uploaded, and will be ready to view shortly.

Purchase these notes here, or revisit this page.

Either way, we'll remind you when they're ready :)

Preview These Notes for FREE

Get a free preview of these Notes, just enter your email below.

Unlock Preview
Unlock Preview

Preview these materials now for free

Why put in your email? Get access to more of this material and other relevant free materials for your school

View Preview

About this Document

Exams 1,2,3 material including possibly notes, video descriptions, textbook content and outline, Including pictures of diagrams for reference.
Scientific Inquiry
Scientific Inquiry University at Buffalo
75 ?




Popular in Scientific Inquiry

Popular in Psychlogy

This 76 page Bundle was uploaded by Jimbo Notetaker on Monday May 25, 2015. The Bundle belongs to at University at Buffalo taught by Wendy in Spring 2015. Since its upload, it has received 174 views. For similar materials see Scientific Inquiry in Psychlogy at University at Buffalo.


Reviews for Exam 1, 2, and 3 Study notes and Lectures


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: 05/25/15
SI January 28 2015 o What is science 0 A way of knowing how do you gain info about the world 0 Many ways of knowing ex is eating gluten free diet goodbetter for you 0 Example will eating a vegan diet be healthier and make you live longer 0 Ways of knowing acceptance of beliefs based on the idea that quotwe have 0 always known it to be this way and that makes it truequot a traditions quotcan t teach a dog new tricksquot quotopposites attractquot quoteverything happens for a reasonquot Tenacious is powerful marketers know this and use this as an advantage Hurtina car Huuuuuuuuge buffalo 0 So easy a But what about GF diets What is ingrained in culture Problems Not the best if information is actually wrong 0 Still powerful to the point where people might lie to back it acceptance of a belief because an authority gure tells us it is so Authorities are power gures and we rely on authority a lot GF diet 0 Dr feegood says everyone should be gluten free m The authority may be wrong also It makes our lives easier to take their word for it but they 0 could still be wrong 0 Acceptance of a belief because it conforms to the rules of logic i all men cannot cook 0 Roger is a man 0 Therefore roger cannot cook It is logical but it is ridiculous Most of the people in the cooking profession are indeed men 0 It39s ridiculous bc it is not accurate Ex All GF diets are healthy 0 eat GF therefore I am healthy That is a logical way of thinking but it is wrong White rice is GF But to eat white rice all the time is not healthy 0 acceptance of a belief based on a feeling that it is true It is not something we can observe it39s just a feeling Ex quotShe likes me I can tellquot 0 But his intuition was wrong nobody likes him l o Acceptance of a belief based on our senses that there s a shared understanding Ulcers motivations Ulcers were always thought to be caused by stress and stomach acid But it turns out that this is false Ulcers are caused by a pylorate Bacteria M o It makes sense that eliminating it works Problem with both INTUTION and COMMON SENSE feeling shared and understanding could also be wrong misinterpreted o acceptance of a belief based on our own expedences Ex endless sunshine is boring or Don t brake and turn in the snow I know this because it is something that happened to be I expenencedit M Improvement over other work 0 Based on the direct experiences better than other bc Global warming video shown in class A man in snowy weather was saying how global warming was not real because it is snowing right now A woman in Australia says it is true because weather only increases and its really hot right now Those are their person experiences at the same exact time during the day in different parts of the world Problems all this could also be wrong 0 Our experiences are limited We cannot experience everything possible or we might interpret things differently from one another 0 Isolated you are just one data point We can39t conclude things from only one data point we need a collective data points to come to a conclusion Potentially bias actorobserver effect o acceptance of a belief based on the results of an observation and experimentation We care about what the evidence and data say not other reasoning Accuracy is tested There are no problems science is the only one that can be tested M oGF diets tend to be more calorie intake Raises risk in health olt39s expensive Scienti c Method39s 4 main steps Observing a phenomenon Formulating tentative explanations Doing more observations amp experimentations Re ne amp retest these explanations n science you cannot fail in love with an idea you have to test it o No other way of knowing has such powerful check on validity of conclusion gtUJNl 39 SI January 28 2015 I Why does 11 use a scientific approach To explain questions like 0 Why are we the way we are 0 Why do we think and behave the way we do I Using science puts is in a place of accuracy I Goals in 1I 0 To describe behavior thoughts feelings behavior I Ex Working together I Why do some people like working together and others don t 0 Class described listed things as to why they prefer dodon t I Dependable division of labor trust I People work less when in a group vs when alone 0 To predict behavior I Ex Wingman s worker will stock less merchandise when in a team than alone 0 To determine cause of behavior 39Ex I When we work in groups responsibility is diffused btwn everybody in the group which means there is less on the individual and tend to be less responsible 0 They slack off 0 To explain behavior I Ex occurs when people work less hard in a group than alone I One cause of this is diffusion of resp 0 May be related to genders in group 0 Types of sciences 0 There is vs no one science is better than the other I Hard ex Chemistry physics I Soft 1P 0 Ex in chemistry I Chemical A Chemical B Chemical C I Always the same I Easily quantifiable mL L ounces 0 Ex III I Psychological phenomena are messy 0 Individual differences gender sex age 0 Difficulty of measurement or manipulation O INTELLENCE nobody agrees on a clean cut measurement on intelligence 0 WELLBEING can t measure it Ex self esteem O PRIVATE THOUGHTS we might not want to share or unaware we can t measure that we might lie 0 CAN T MANIPULATE we simply cannot manipulate things like depression We can t control how much depression someone can have to experiment on 0 Science is science 0 Challenges of 11 science 0 Fundamental difficulty I Some questions are difficult to answer via scientific means 0 People are not always honest 0 Animals don t communicate 0 How do we 1P researchers trv to overcome I INGENUITY coming up with creative ways of solving a problem 0 Ex can animals see colour 0 If not direct I have to find an indirect way to find out How 0 On mice we use a skinner box I Press lever to get food 0 Put a light in the box w different colours I If it can distinguish blue apart from green it will show us they distinguish if they behave differently and try w different colours as well 0 Remember people performs science 0 So it benefits from human capabilities I Ingenuitv creative thinking of new ways to ask I Intelligence compare and contrast I Good sense different from intelligence 0 Don t make anything more complicated than it needs to be be simple 0 PRINCIPLES OF DRUNKARD S SEARCH 0 The story of a drunk man looking for his keys at night under a lamp post When asked where he last saw them he pointed to some place completely different from where he was but he was in the place because there was light 0 Don t look in the convenient place look at the likely place 0 More technology doesn t mean easier answers Don t use more technology than needed I Enthusiasm you might have an ingenuine experiment and end up w no answers we need scientists to stay motivated and enthusiastic about keep going I And falls prey to human pitfalls 0 Imperfection we are not perfect you want to do well 0 Gallo The French and HIV they were too competitive and it was bad They wanted individual glory 0 Competition if you want glory you will have clouded judgment I Fraud 0 Karen Ruggiero Targets of Prejudice Research 0 Quinton s story on fraud on the lady who stole slept on her couch and stole her experimentresearch lied about data 0 Karen from Harvard took professional data and published it People questioned Karen39s data and was eventually caught for fraud 0 Andrew Wakefield Vaccine and Autism 0 Video He changed data he claimed kids had autism when they were actually perfectly fine and for the kids who reported of developing autism were actually already sick and had it from before they were vaccinated 39Motive might have been money glory 0 Diedrick Stapel 2 30 publications all fraud SI February 2 2015 Theories and Hypothesis Theory an organized set of ideas that seeks to explain why two or more events are related Hypothesis a testable prediction that follows from a theory allows testing of theories gives rssearch a direction and makes theories practical De ning Variables Variable Event situation behavior or characteristics w more than 1 value 0 Theoretical conceptual definition The meaning of a variable in abstract or conceptual terms ex In words 0 Operational definition The meaning of a variable in terms of the methods used to measure it 0 Independent Variable IV Variable the experimenter manipulates to determine whether it has effect on another variable the dependent 0 Ex We are manipulating contact in a prejudice example Dependent Variable DV Variable the experimenter measures to determine whether another variable the IV has had an affect EACH HAS A THEORETICAL amp OPERATIONS DEFENITION Concept in Practice 0 Theory By comparing w others we learn about ourselves 1954 Looking at other people to see how and where we stand 0 comparing ourselves with someone better than us 0 Can make us feel worse about ourselves comparing ourselves to someone worse than us 0 Can make us feel better about ourselves SociaI Concept Theorv Hypothesis Morse amp Gergen 1970 o When waiting to be interviewed for a job Those who wait w another applicant who has desirable characteristics will make the waiter feel worse about themselves Vice versa SeIfEvaIuation the DV 0 Theoretical de nition how one assesses one s own quantities abilities and attitude 0 Operational de nition participant39s responses on a survey assessing beliefs about their perceived abiIities Comparison Referent the IV 0 who the applicant waited with TWO LEVELS ONE upward comparison Mr clean Mr Clean comes in w a suit briefcase and starts reading a stats book TWO downward comparison Mr dirty Mr dirty comes in w a smelly sweatshirt ripped pants and asks if anyone has a pencil 0 Who would you rather wait with DV completed before and after interview 0 What do findinos mean 0 Was the hypothesis supported YES 0 What were implications have for theory IT SUPPORTS THE THEORY What about 0 A researcher says caffeine causes a greater romantic attraction by increasing overall arousal The researcher predicts that college study partners who are paired based on potential for romantic attraction will be more attracted o m Two levels those who get caffeine M those who do not 0 m one level Attraction The researcher replicates the study but measures arousal amp attraction o M Two eves SAME AS ABOVE o mTwo eves attraction TWO arousal What about ex 0 A researcher says vigorous exercise effects on mood bc exercise releases endorphins Researcher predicts endorphins will be higher and mood will be good 0 M One eve Vigorous exercise 0 mTwo eves ONE mood TWO level of endorphins The researcher predicts that vigorous exercise will affect the number of oxygenated blood 0 M One eve ONE vigorous exercise 0 m One eve ONE oxygenated blood 0 What is the problem with this part of the experiment We have to stay relevant to what we are trying to prove in the main theory Yes you can experiment this hypothesis but what is oxygenated blood telling us about improved moods and endorphins Nothing therefore it is not necessary Sl February 4 2015 Generating hypothesis 0 We generate hypothesis from theories test them look at results see if data is supported 0 Theory Exerting selfcontrol depletes resources which makes additional self control more dif cult Theory of Ego Depletion Buameister et al 1998 Muraven et al 1998 o What they are arguing is much like a muscle when you deplete it it makes additional selfcontrol muscle Hypothesis People asked to force themselves to eat radishes instead of chocolate will be more likely to quit working on a dif cult puzzle than people not asking to exert selfcontrol over eating o Is it getting to the idea of the theory Is it exerting selfcontrol 0 Does this hypothesis stem from this theory YES 0 How do we manipulate it 0 Theoretical De nition of IV exerting selfcontrol refers to denying immediate grati cation It doesn39t have to be tied to the theory It is just de ning what selfcontrol means 0 Operational de ne of IV Selfcontrol instructions selfcontrol eat radishes vs no control eat whatever you want 0 How do we measure it 0 Theoretical De nition of DV Exerting additional selfcontrol refers to additional denial of immediate grati cation 0 Operations De nition of DV Persistence on aversive task dif cult puzzle measure in minsec We did an in class assignment Five Criteria of Good Theorv Predictive accuracv It should reliably predict behavior 0 Internal coherence It has ideas that logically follow from one another Parsimonveconomv It39s no more complicated than it needs to be 0 A good theory is a simply theory Fertility it should suggest new ideas for further study Veri ability its testable through empirical methods 0 If somebody tells you they have ESP mind reader its not veri able 0 Not veri able not a good theory Idea of proof 0 We never quotprovequot a hypothesis or theory 0 Hypothesis and theories are IDEAS 0 Why How many observations do you need to prove an idea INFINATE More observations quotthe nextquot could be the one that debunks it 0 We test hypothesis derives from theories 0 If the results consistent with hypothesis Hypothesis and theory are supported 0 Ex cheating boyfriend You observe him 1000 times and he doesn39t cheat But maybe it39s the 1001 times that he does SI February 6 2015 o A good example of the in class exercise 0 O O O O 0 Theory Familiarity breeds liking Hypothesis kids repeatedly exposed to sushi at home will learn to like sushi more than those not exposed to sushi at home Theoretical de nition if IV Familiarity is repeated exposure to an exotic food Operational de nition of IV Exposure kids who are exposed to sushi vs not Theoretical de nition of DV Liking means having a positive attitude towards something Operational De nition of DV Kids39 desire to eat different variety of sushL Example 2 O O O O O 0 Theory Familiarity breeds liking Hypothesis People will like their new haircut better after a week than on the initial day of the haircut Theoretical de nition of IV Familiarity is repeated exposure to the unfamiliar haircut Operational De nition of IV One group regularly Theoretical De nition of DV IN PHOTO Operational De nition of DV IN PHOTO Common types of relationships between variables 0 0 Positive linear relationship INCREASES in one variable are accompanied by INCREASES in the other variable The variables are moving together as one goes up the other goes up As one goes down the other goes down m Height weight 0 As people get taller the weight increases same as as shorter the less you weigh Age of car mileage units spent on textbooks What does it look like Negative Linear Relationship Increases in one variable are accompanied by DECREASE in the other variable EL miles from home to campus trips to campus pe Units taken last semester hours worked at outside times been broken up with trust in relationship partnlt Curvilinear Relationship INCREASES in one variable are accom by both INCREASES amp DECREASES in the other variable r wee job ars panied Ex Marital satisfaction Stress doing a task Seerv et al 2010 No Relationship Changes increases or decreases in one variable are not associated with changes in the other variable Ex siblings books read last year Age of roommate age of car units completed heights 0 What do correlations tell us 0 Variables change together They are associated o Predictability Knowing that there is a correlation we can predict Can predict probable value of one variable by knowing value of other variable Predictor variable X Variable used to predict Criterion variable Y Variable whose value is being predicted not associated with changes in the other variable Ex siblings books read last year Age of roommate age of car units completed heights 0 Ex use caffeine consumption predictor variable predict rate of speech criterion variable Y o No Relationship Changes increases or decreases in one variable are X to o What correlations DON T tell us 0 Does not tell us anything about CAUSATION 0 Correlation Causation Can39t ever say that X caused Y if you39re only looking at correlation Why not ca rtoon d iagra m 0 Man quotI used to think correlation implied causation Then I took a stats class Now I don39tquot o Girl quotSounds like the class helpedquot 0 Man quotWell maybequot SI February 11 2015 Experimental Research Has two de ning characteristics 2 Control over extraneous Variables o Extraneous Variable any variable not manipulated in experimental but that still may affect outcome 0 Ex Age personality traits mental states time of treatment complexity of treatment 0 Ex Study on sugary snacks vs healthy snacks on mood In the ice scream condition looks amazing to the eye In the healthy snack chopped up vegetables on a paper plate You are not manipulating only sweet vs healthy but also presentation how it looks Try to control every extraneous variables as much as possible 0 How do we control extraneous variables Make sure treatments are exactly alike except for level of N o Run all conditions in the same rooms 0 Use placebos 0 Keep the experimental attire presentation style the same Even with animals No perfume 0 Script andor automate instructions 0 Keep cage conditions treatment diet same Don t favor any animals because you like it more or cuter Make sure remaining extraneous variables operate in entirely random fashion 0 Randomize experimenters treatment daystimes 0 Random as nment of participants to condition This means that every participant has equal chance of being assigned to any condition lfl have a high anxiety person they have an equal chance in being in any condition placebocondition We do this because it makes it highly likely that extraneous variables will be distributed equally across groups ON EVERY PARTICIPANTS VARIABLE no matter what it is Random Assignment 0 We randomly assign subjects to levels of the IV 0 What does it do 0 It makes the groups EQUAL on everything EXCEPT THE IV 0 No order Truly random makes everyone equal 0 Study using Random Assignment 0 IV Antidepressant drug no drug vs drug 0 DV Depression 0 Starting out two groups be EQUAL in depression Ex Group A is going to be approximately a the same level of depression as group B 0 After experiment If there is a difference after treatment we know that it39s due to the drug Some wavs to randomlv assidn Flip a coin Roll dice Pick condition out of a hat 0 Once you pick somebody from a hat you have to put it back so it can still be random Use a table of random numbers Use a random number generator on a computer 0 m Randomorg You do not have equal numbers Random assignment and vou Two groups randomly assigned based on random generator 0 Groups 0 Group 1 n78 0 Group 2 n2107 Age 0 Group 1 M2033 0 Group 2 M2107 Height 0 Group 1 M6588 0 Group 2 M6531 RANDOM ASSIGNMENT IS ESSENTIAL FOR EXPERIMENTAL RESEARCH 0 Without it cannot Call study an quotexperimentquot Make causal inferences o Is it an experiment What is being manipulated o If the answer is nothing being manipulated then it39s a non experiment Is there control over extraneous Variable Random assignment Same Resea rch Effect of intervention program on low birth weight infants Rauh et al 1988 o 78 mothers amp their low birth weight infants 0 Randomly assigned mom amp baby to experimental or control group by ip of coin 0 Exp Group moms received training in special needs for low birth weight infants control group moms did not 0 Measured the mom39s selfcon dence amp the baby s development What did they nd 0 Compared to control groups mom in exp Group were more self confident and babies had greater cognitive development SI February 13 2015 0 Assessing the Rauh et al Study 0 Where the two conditions for an experiment met Manipulation of IV INTERVENTION PROGRAM Controlling extraneous variable 0 Hold them constant Whatever we can39t constant it we randomize by random assignment THEY FLIPPED A COIN FOR THE CONTROL GROUP 0 Can we conclude that the intervention program caused the difference in outcomes We can only say CAUSE when it39s in an experiment YES 0 Criticism Ethical AT THE TIME YES Is it the nurse visitingthe information or both WE DON T KNOW Class assignment 0 Difference in reported weekly alcohol intake among college students vs non college attending 0 Manipulation SI February 16 2015 0 Issues with the in class exercise 0 Some struddled with gave incorrect reasons for why it WOULD NOT be a difference that could be tested experimentally No control group No way to measure the DV Confused IV with the DV quotunethical to manipulate alcohol intakequot 0 Gave incorrect reasons for why it WOULD Possible to put P39s into two groups there are two groups Confused the IV with the DV you can control alcohol intake 0 Remember to focus on the two criteria for an experiment 0 Is it possible to explore these ideas EXPERIMENTALLY o A researcher is interested in relationship satisfaction among couples who either complete a challenging physical task together or individually m relationship satisfaction IV completing challenging task together or individually can you manipulateYes IS an experiment 0 A researcher wants to know whether people with Type A blood are more or less likely to donate blood than those with Type 0 blood DV Likely to donate IV Blood type can I manipulate No IS NOT an experiment 0 A researcher is interested in whether people who are left handed have greater activation in their occipital lobes when presented with visual stimuli in their right vs left visual eld DV activation occipital lobe IV right vs left vision eld manipulate Yes IS an experiment 0 Pros amp Cons of Different Approaches o PRO You can look at relationships of interest that cannot be investigated otherwise you only can able to look at thing are you can experiment CON Inability to infer causenonexperience you can39t cause only relationship PRO Ability to identify and describe causal relationships CONS Cannot be used if you cannot manipulate your variable of interest experimental is limited to what you can and can39t study 0 Also Tight control over extraneous variables limits generalizability you may be able to infer cause bc controlled extraneous variables but you can39t infer that39s how it will be in the real world Bc the real world nothing is controlled 0 All Experiments Are NOT Created Equal 0 Question every study Quaity depends on design Internal Validity The certainty with which results can be attributed to manipulation of an IV rather than some other variable 0 We want to know if internal validity to be high Uncontrolled extraneous variables threaten internal validity Internal Validity o a Is new teaching method better than traditional method Study 0 4th grade students randomly assigned to new vs traditional method 0 Mr V age 23 does the new method Mr F age 61 does traditional method 0 Both use same text material and tests 0 Teaching effectiveness measured by test scores Finding Students taught with new method outscore those with traditional method Can we conclude that the new method is better than traditional Manipulated teaching methods 0 They used the same text material and tests Randomly assigned 4th graders Still NO 0 Why Mr V amp Mr F Likeability IdenU cann Enthusiasm Expectations might evoke self ful lling prophecy Other factors 0 Grading styles 0 Atmosphere of classrooms o Workload What happened in this study SI February 18 2015 Internal Validity o What happened in this study Confounding 0 When effects of two or more variables cannot separate o Is study high or low internal validity 0 Low internal validity How is internal validity 0 Highest internal validity controlled lab setting 0 Field setting improvement We want to avoid confounding whenever possible o It ruined internal validityexperiment External Validitv Degree to which results can be generalize beyond sample 0 More relatable amp looks like the real world 0 Can happen vs does happen Common complaints 0 We are in ll to be able to generalize results its important 0 But mostly we ask what CAN happen not what DOES happen Internal and external validity are not opposites There is a tradeoff between the two 0 The more I control extraneous variables the more internal validity which is great 0 But the real world is not like that its more chaotic Steps to increase one often decreases the other Compromise 0 We care a lot about generalization but we never sacri ce internal validity Ex A dr was studying what couples do but her setting looked just like a living room She has a lot of external validity but she doesn39t compromise internal validity she compromises external makes it looks like a regular home Measurement Concepts When we measure something DV39s these measurements uctuates changes Error that tends to push measurements up and down around an exact value Ex If you are 6 ft tall and I measure you sometimes I will get 6 1 sometimes I will get 5 9 Thev kev thing about it If I take the average and look at how close it will be to exact value it will be really close Random error is more benion that the next error Error that tends to push measurements in the same direction If I take the average of all the measurements it will not be close to the exact measurements It will tend to be skewed to one direction oAnd so the average will also be skewed to be an underestimate or over estimate of the exact value Ex A grocer is weighing grapes and you buy grapes every day for a whole year oRANNDOM ERROR Scale weighs heavy on humid days light on dry days 0 The scale will uctuate between increasing and decreasing You will eventually be paying the same thing oSYSTEMATlC ERROR Grocer puts thumb on scale 0 You are buying more because he is constantly increasing the weight oWant to keep both types of error LOW How assess accuracy of measures Two main methods o Stability or consistency of a measure It is measuring what it39s measuring in the same way every time We want measurements to be as reliable as possible Ex If you are 6 ft we want the ruler to measure 6 ft every time o If we have a rubber ruler it39s going to be uctuating we want a reliable ruler reliable measures On multiple measurements oUnreliable measures D different readings rubber ruler oReliable measures D same highly similar readings metal ruler Ex Rosenberg Selfesteem scale 0 They ask the same questions different ways to get accurate ratings from the person taking the test Wavs to asses reliabilitv Measure same Ps at two points in time 0 Correlation between two scores Measure The P39s at one point in time ocorrelate one random half of in items with other half ocorrelate each item with every other item Then calculate mean of all correlations Assessed by multiple observations of same behavior 0 Correlate judgment of raters How much do you need oGeneral rule of thumb does not apply in all situations Adequate reliability Is 70 or higher o It measures what it s supposed to be measuring doing what it39s supposed to do Essential that measures are valid indices or phenomenon of interest Extent to which an instrument measure what it claims to measure 0 Its an umbrella term which all other types of validity fall 0 Unlike reliability which has numeric guild lines There are many different facets in validity o No numerical guild lines SI February 20 2015 Validity Essential that measure are valid indices of phenomenon of interest Construct validity Extent to which an instrument measure what it claims to measure Umbrella under which all other types of validity fall 0 Types of Validity Convergent validity Measure is related to other measures it theoretically should be related tospeci c Ex measure of depression should be 0 Positively related to measure of negative mood o Negatively related to measure of selfesteem Discriminant validity measure does not relate to other measures it theoretically should not relate to 0 Ex measure of depression should not be related to measure of spatial ability 0 Ho relate to construct validity Reliability amp Validity Can we have something that is reliable but not valid YES 0 Measuring ears with a metal ruler longer ears means happier life Reliable Yes because metal ruer doesn39t uctuate Valid No ear size has nothing to do with life happiness Can we have something that is valid but not reliable NO o A watch is always too fast too slow ad stops Reliable No Valid no we can39t have something that is not reliable to be valid Reviewing scienti c Method Video Understaninding expectations SI February 25 2015 Nonexperimental research Methods ways of observing aka quotfield workquot Observing and describing naturally occurring phenomena with little experimenter interventions 0 You are watching not experimenting 0 Has two functions Gaining descriptive knowledge gure out what39s going in here what do I see Gaining insight about patterns and relationships we might notice that a couple pf variables or behaviors go together in this setting 0 Tends to be global in scope The focus is broad interested in everything that I come across in this setting o Michael Fay39s megatransect across equatorial Africa Interested in preserving natural resources Conservationist walked across African jungle from sept 1999 through Dec 2000 c He walked across the Republic of Congo Cameroon Gabon 0 Walked through thick forests swamps rivers He came across few human settlements Goal record what39s there to spark conservation efforts Encountered and recorded Plant life Animal life behavior excrement Ebola Poaching where people kill wild animals for pro t People villages not many but some He endured malaria hepatitis other infections o Rosenhan39s 1973 study quotOn Being Sane in lnsane Placesquot What39s the power of being labeled how do we diagnose illnesses how does it feel to be admitted through the patients perspective lnterested in experiences of patients in mental hospitals 8 sane people gained admission to mental hospitals claiming they heard voices oThe only thing they lied about was that they heard voices and their names Once admitted patients behaved normally did not take medication Stay at the hospital ranged from 7 to 52 days Discharged with schizophrenia in remission after M19 days oThey were in different hospitals Documented treatment by doctors quotpatientsquot feelings of powerlessness depersonalization 0 They have no control no agency they are not a person anymore Careful observation of one or more speci c behaviors in a particular setting may include experimenter intervention 0 Tends to be a narrow scope focusing on certain behavior or relationships often tests a hypothesis or two 0 Robber39s Cave Studies Sherif et al 1961 lnterested in the relationship between competition over scarce resources and prejudice It was done at a camp focused on two groups of 11 yearold boys at summer camp Boys at a camp in different cabins Boys did everything with their own group for the rst week Rattlers vs eagles Second week they competed with other group for desired prizes pocket knives medals After two weeks groups showed all components of prejudice stereotypes and discrimination lncuded intervention attempt at prejudice reduction 0 Working towards a superordinate goal goals that could not be met by one cabin alone 0 Messed up water supply 0 Tickets for movies 0 Stranded a truck in the mud o Stopping Signaling at Stop Signs Lebbon et al 2007 In 2002 680000 accidents per year at stop signs 3500 were fatal Focused on drivers at Tintersection near large college campus Trained observers are watching for Legal stops tires stop rotating Turn signal use Conclusion Oncoming traf c is signi cantly related to driver stopping amp turn signal use 0 Legal stops 100 with traf c 46 without o Turn signal use 63 w traf c 44 without This is a systematic because they are only focused on two very speci c behavior Not EVERYTHING that happens 0 Kev differences in examples Naturalistic observations megatransect quotOn Being lnsanequot Global Scope 0 Examples 0 Foster homes 0 North Korea 0 Environment of Mars 0 Cult dynamics 0 Amish 0 Witness protection Systematic Observations Robber39s Cave39 Stop Signs Narrow Scope 0 Examples OOOOOO Mexican borders who gets through who do not Homeless people sleeping habits in winter People watching who checks other people out Who goes inside restaurant vs drive through Who washes their hands in the bathroom Watching people drinking at bars SI February 27 2015 Nonexperimental research Methods Wavs of Observing 0 Nonexperimental research designs 39 Case studies I Archival research 39 Strengths and weaknesses Nonexperimental Research Methods Ways of Observing 0 Case Studies Analyses of the experiences of a particular person or group I Extraordinary experiences that are difficult or impossible to recreate 39 Used most by 0 Clinical psychologists 0 Behavioral neuroscientists 0 But any type of psychologist can use them 0 Case study example Phineas Gage 0 Railroad foreman summer 1848 o 43 inches tamping iron prefrontal cortex he survived 0 Iron and skull are on display at Harvard 0 What happened to him Personality changed Emotionally volatile emotional mess 0 Why is it interesting There is something about the brain that changed Daguerrotype Mystery 0 They are a couple who liked to collect pictures and post them on icker 0 They had a picture of Phinease Gage Nonexperimental research Methods Ways of Observing Archive research Using previously collected information to answer research questions What might we look at Common types 0 Census data 0 Public Records 0 Ex Grosvenor Room Buffalo Center Library 0 Where people originally settled people tend to still settle in these areas now 0 Communication records 0 Survey Database 0 Information collected over time Ex Price of cotton and Lynching in the American South Hoyland amp Sears 1940 0 They wanted to study conditions that would tackle this Looked at 7 states looked at the price of cotton and number of lynching This was after the civil war and slaves were freed o postcivil war there was competition between whites and blacks for cotton 0 prices were recorded 0 They were also interested in prejudice o Lynching recorded 0 Predicted bad economic conditions D frustration o Frustration racially motivated violence 0 Time spam 18821930 0 Cotton prices good cotton year high prices Bad cotton year low prices 0 4761 recorded lynching most likely an under estimate more unrecorded lynching Prediction subported their prediction that prejudice and cotton prices correlate o Displaced aggression scapegoating Real world data revealed pattern Ex Physical Mental health Among Civil War Veterans Pizarro et al 2006 0 They were interested in how conditions of the world had effect on mental and physical health 15027 Union Veterans 0 Records were kept by American government so records were good Military pension records Confederate were not kept by US government records were bad 0 Predicted Severity of war experience D poor physical mental health 0 Findings 0 Greater company killed you survived but everyone around you died increase postwar cardiac GI disease increase postwar nervous disease 0 Younger age during enlistment lt 18 years increase postwar majority risk of early death increase postwar nervous disease Content Analysis Systematic analysis of contents of documents 0 Investigation of email addresses Quinton et al 2007 0 1855 college student email addresses 0 Coded for 4 attributes Reveals sexgender Indicate youth Mentions appearance Mentions af uence nancial status 0 Coding scheme goes into detail what each attributes they were looking for o Coders blind to addressee39s sexgender Predictions o 1 women will reveal sexgender indicate youth mention appearance gt men 0 2 men will mention af uence gt women 0 Findings Support for 1 not 2 0 Maybe bc men are young and don39t have money they didn39t advertise that they have money Nonexperimental research design 0 Consider the weaknesses We can39t infer causation only related 0 But also remember the strengths We learned a lot about competition amp prejudice about the brain post war heal Sl March 2 2015 Survey Research Is Used To 0 Evaluate Speci c attitudes or behaviors o Are you in favor of a tuition increase at UB o What house hold product do you use Predict behavior 0 Are you going to enroll in the U35 201 5 summer session 0 How likely are you to buy a Tayota Steps in conducting Survev research 0 Step one Designing your questionnaire so many issues could be a whole course 0 Must clearly de ne topic to gear uestions towards that topic Ex Did you have sex with your last boyfriend of girlfriend This is not a speci c question lnstead asses different levels of sexual activity Be more speci c How many times 0 Goal Enough to throughouly asses topic not so many as to become tedious or confusing Select the appr0priate questionnaire format Open ended Allow participants to provide response in own words Close ended Participants selects response from a limited number of given options 0 Example I am in favor of gay marriage Askino pe0ple about themselves on questionnaires or in interviews 0 Very widely used Warranty cards Phone calls asking your opinion on social political or community issues lnternet surveys Important distinctions 0 Survey research vs naturalistic or systematic observation Observation methods 0 You do not administer any measures you observe behavior Survey research 0 You directly question participants Survey research vs Experimental research 0 No variable manipulated l survey research 0 Its Nonexperimental lf questionnaires then survey research 0 Not necessarily Sl March 4 2015 0 Many DVs in experiments consist of questionnaires Steps in Conducting Survev Research 0 Step 1 Designing your questionnaire 0 Write effective items not too many words simple questions easy to answer not dif cult Pilot testing and revision Testing questions on a sample of people to get feedback and revise Keep language clear and simple terminate or end END 7th8th reading level Avoid vague questions Vague quotHow do you feel about the moviequot Precise quotHow much did you enjoy the moviequot Avoid Biased questions Biased quotDo you think pornography is disgustingquot Unbiased quotWhat is your attitude toward pornographyquot Avoid unnecessary negatives Bad quotShould the president not nominate a conservativequot Better quotShould the president nominate a conservativequot Avoid doublebarreed items When asking two things in one queonns DB quotWhen thinking about the incident I got upset and sought helpquot 0 Some people get upset but don39t seek help Or vice versa 0 Better quotWhen thinking about the incident I got upsetquot AND quotWhen thinking about the incident I sought helpquot Include reversescored items to avoid response set answering When you answer the same questions with the same response 0 Measure of depression quotI feel depressedquot BUT quotI feel happyquot reversedscored Format questionnaire in an organized and visually appealing manner 0 Why Format things really well when it looks like researcher puts thought into it you are going to take it more serious Sl March 6 2015 Test Anxiety Less effective Items De ne Nervousness prior to and during evaluation exams GOOD Do you think about tests before taking them BAD I study the day before a test BAD Do you eat food before any of your tests BAD I do not enjoy taking exams BAD I usually nd exams to be challenging BAD Do you typically become feeling overwhelmed with a sense of dread when you are taking exams that may have a direct effect in your grades BAD Very effective ltems While taking an exam I often feel anxious about howl will perform GOOD During exams sometimes I feel so nervous that l have trouble remembering things GOOD The night before a big exam I often have trouble sleeping GOOD My heart often races when I think about taking an exam GOOD I usually feel calm in testtaking situations GOOD Life satisfaction 0 0000000 0 o O O O 0 Less effective De ne an individual39s feeling of overall contentment with the quality of hisher life GOOD Im happy with my body BAD I wish I had a better car BAD feel good about myself and my life BAD My job is very stressful BAD I am happy with my exam grades BAD Are you nancially stable BAD I feel energized BAD l have a lot of friends BAD Very effective I am satis ed with my life right now GOOD I feel happy with the path I have taken the way my life has turned out GOOD My life is currently exactly as I would like it to be GOOD I would like many things in my life to be different than they are now GOOD Steps in conditioning Survey Research 0 Step Two Acquiring a sample of participants Populations whatever you set the population to be who you are targeting 0 Sample you want a sample of the people that you are interested in Samples should represent the population of interest 0 Representative sample sample that closely matches characteristic of population Biased sample sample that does not closely match characteristics of population 0 Ex bag with 300 balls 100 white 100 yellow 100 orange select sample of 30 Real life example of biased sampling 0 1936 presidential poll done by Literary Digest 0 What went wrong They surveyed the wealthy and the wealthy tend to lean more republican Variety of sampling techniques are discussed An important distinction Random assionment When P39s are assigned to conditions in an arbitrary and unbiased fashion 0 Not expensive 0 Random selection when every individual in the population has an equal opportunity of being selected into the sample 0 It39s expensive 0 Most often experiments do not do this 0 Random Digit Dialing 0 Large scale surveys tend to use this 0 Step three administering your questionnaire By mail Advantaoe convenience Disadvantaoe o May be important differences between nonresponses and responders lbiased sample By telephone 0 Advantages reach a wide variety of people 0 You can collect data quickly with multiple callers Disadvantaoe Behavior of interviewer may affect responses 0 People are hesitant to answer telephone survey39s In groups Advantage Collect large amount of data in brief time o Assured a high response rate Disadvantages Ps may feel pressured to participate unethical we don39t39 want people to feel pressured by conforming to the group 0 May not be as truthful as when alone Face to face interview Advantage Rapport with interviewer more serious truthful responses OR less truthful MINI QUIZ Andwers 1C 23 3B 1 2 3 Disadvantage behavior of the interviewer may affect responses 0 Expensive and time consuming Over the internet Advantage inexpensive 0 Easy to collect data from large pool of respondents Disadvantage Sample may not be representative of population of interest 0 Dif cult to prevent multiple responding If you wanted to control extraneous variables in your experiment you should be sure to aRandomly select participants from the population LDo a controlled interview cRandomly assign participants to conditions LHand out questionnaires in a laboratory If you decide to do a study on Americans overall life satisfaction and you want to makes sure every American has the same change of being in your study sample Therefor you should aUse random assignment LUse random selection cUse the internet LUse a mailed survey To have an experiment you must aRandomly select participants LRandomly assign participants c Both a and b SI March 9 2015 Experimental Design 0 Goals 0 Two Independent Groups Design Basics Examples Experimental Research Designs 0 Research Design Outline plan or strategy used to investigate a question 0 Main goal of experimental research designs way to the participants in the other variable One independent variable two levels Determine cause To do this you need to 0 Manipulate an IV 0 Control extraneous variables Good vs Bad designs 0 Valued based on how clear the causal inferences are that can be made from the experimental design ie the conclusions that can be drawn about cause 0 Bad designs are much easier to do than Good designs Good Experimental Research Designs 0 Two Independent Groups The participants in one variable are not related in any Dependent variable always remains the same across groups Treatment IV Response DV Posttest Group One No training I Employee Output Group Two Efficiency Training Employee Output Program I Often experimental group control group control groups are a good way to control extraneous variables Other variations 0 Ideal dosage for new drug 50mg vs 100mg In example find difference in output between two groups 0 Group 2 efficiency training program measured greater output than group 1 no training 0 Caused by training program 0 YES As long as we control extraneous variables Ex Drooling amp Lachman 1971 Properties of Memory 0 Read and remember passage 0 Recall as many exact words as you can 0 Three Sturdy sisters peaks and valleys calm vastness 0 IV Title of Story Absent vs Present 0 Title Christopher Columbus Journeys to America DV of words remembered 77 in passage Findings 0 No title 1325 words 0 Title 1575 words 0 Title made a significant difference Conclusion 0 Titles I Serve as a cue help us make sense of things I Cause title causes better recall Implications 0 Framework matters ie study strategies Ex Alter amp Oppenheimer 2009 Cognitive Fluency and Self Disclosure Fluency Positive experience that results from ease of information processing ie when we understand what s going on we have a positive experience Situations with greater ease of information processing I greater uency Hypothesis Fluency I greater selfdisclosure 0 Method All participants in study completed Social Desirability Scale SDS which measures social desirability bias I How much people claim 0 Virtuous but impossible No matter who I am talking to I am always a good listener I How much people deny 0 Common human frailties I like to gossip at times I Lower scores more selfdisclosure more honesty about personal aws IV Fluency Via Questionnaire Font Clarity O ClearFluency 12point font 100 dark 0 DifficulttoreadDis uent 10 point font 50 gray DV Scores on SDS Findings 0 of questions answered in socially desirable NONself disclosing manner was much higher in the dif uent condition relative to the uent condition Conclusions 0 Fluency I Increases causes greater selfdisclosure I Why 0 2nd study showed Dis uency primes thoughts of risk 0 Implications 0 Mental health professionals physicians rely on honesty from patients so they must ask questions that are simple and clear SI March 112015 Good Experimental Design Two Independent GrOUps Desan Advantages missed Disadvantage 0 Investigation of only one IV at a time Ex What about training AND workspace can39t do that 0 Limited information about effective of IV and DV it is limited by the 2 levels Ex Full vs some ef ciency training can39t do that either 0 Ex was posttest only design 0 Variation this includes 0 Pretest testing measure DV before and after the IV Bene ts 0 Ensuring group equivalence random assignment however in small groups random assignment doesn39t work very well outliers are heavy on data vs n large groups 0 Checking reasons for mortality people drop out of studies 0 This is a problem because in a heavy smoking study you might have the heaviest smoker drop out and if we don39t do pretests we don39t know if the people who drop out are heavy or lighter Drawbacks Sensitization you are giving them assessments and you are giving them what you are testing and it39s not so blind Timeconsuming Multiple independent GrOUps can be just posttest or pretestposttest Treatments Response posttest 0 Group 1 no training 0 Group 2 Effeciancy training program 0 Group 3 0 Difference in output between groups caused by training program 0 Hold extraneous variables constant Manipulate IV control DV random assignment 0 So YES 0 Ex Lee Fredrick amp Ariely 2006 Beer preference Research Question How does knowledge affect gustatory expenence Speci cally negative knowledge offensive ingredient Focus on timing of disclosure of negative info 0 Method Never vs Before vs After consumption 0 Used Bar patrons sampled two beer samples 0 0 Regular beer Bud Wiser quotMITquot brew Regular beer w balsamic vinegar offense ingredient added 0 IV presentation of info about offensive ingredient 3 levels 0 DV Beer preference regular vs quotMITquot brew Findings blind above 50 before 30 After Above 50 liked the balsamic vinegar Conclusion Expectations affect realtime experience itself 0 Disclosure of negative info after experience did not reduce people39s liking Implications If you want to like food eat rst ask later Parent39s strategy for picky kids 0 Crab cake quotSea Hamburgerquot YUM 0 Ex Labert et al 1020 Gratitude and relationship value Research questions how does experiencing gratitude affect one s view of a relationship Hvbothesis Gratitude increases communal strength 0 Communal strength the degree to which you feel responsible for partner39s welfare Methods 0 IV expression of gratitude of friend 4 levels 0 O O O 0 Expression of gratitude Email note tell you appreciate something heshe does Thouohts of gratitude thinking about think you appreciate Positive interaction focus on positive memories Neutral think about daily activities Wrote reports about each activity biweekly 0 DV The DV was the amount of communal strength between the two friendships 0 They found communal strength was higher when expression of gratitude was present Conclusion expressing gratitude increases the expresser39s feelings of communal strength in the relationship Gratitude works like relationship glue 0 May D reciprocate effect Implications 0 Want a closer relationship Express gratitude Advantages to this desion o Allows complex testing of effects of an IV 0 Easy data analysis Run a oneway ANOVA o chart of training to mean of candies made per hour Look for a main effect on training Must do a post hoc test to determine where differences are Disadvantages to this desion 0 We can only investigate one IV at a time o Requires more P5 and time than 2GD Which design should you choose Pick the best design that answers YOUR questions It depends on what you are researching Fl March 23 2015 came in late Experimental design recap 0 Two independent groups design 0 One IV two levels 0 Data analysis t test looking for main effect of IV This is the difference between 2 means 0 Multiple independent group design 0 One IV three or more DV 0 Data analysis one way ANOVA looking for main effects of IV Post hoc test to check difference 0 However this in ates typel error 0 Posttest only vs pretestposttest o Pretest bene t check group equivalency Reasons for mortality 0 Pretest drawbacks sensitization time consuming generalization Issue More good experimental research designs 0 Repeated measures withinsubjects designs 0 Each P5 is exposed to ALL levels of the IV rather than being randomly assigned to one level Everything is in all levels Ex 10 P5 we are interested in looking at cognitive ability to remember a list of words Repeated Measures Withinsubiect design 0 Why repeated measures They are doing the DV what we measure is measured repeatedly Why Withinsubjects We are looking at within the P5 if there is change 0 Changes in responses ae measures within each P across the Tx IV Tx treatment Contrasted w Independent groups or betweensubjects design 0 No random assignment can we still be considered an experiment 0 Yes because the role of randomization is to make everything equal across treatments In a repeated measures design groups are exactly equal prior to treatment as they are the same people 0 So if differences in memory for 3 vs 5 letter words Caused by word length 0 Held extraneous variables constant Yes 0 YES YOU CAN 0 Ex Bellock et al 2002 Attention and Sensorimotor Skill 0 Research questions Can paying attention work against you 0 Hypothesis focusing attention on execution how you do a task while performing a well learned skilwill result in poorer performance Pooper than what What are we comparing this too Focusing attention AWAY from the task Why 0 When you have a skill that you learned very well that skill becomes automated you don39t need to think about that skill any more Paying attention goes against automated nature of well learned skill 0 Method Sample Experienced college Golfers Task Golf putting on indoor green 0 Hit X o M Direction of attention 2 levels Skillfocused Attend to speci c aspect of performance 0 Exact moment club head stopped followlength Dueltask Attend to an outside stimulus during performance 0 All Ps did both levels 0 DV Distance from X 0 Finding In the condition that golfers were asked to focus on puck Conclusion paying attention to welllearned task caused poorer performance 0 Implications Choking under pressure 0 Advantages All P related factors are identical across Tx 0 Age personality sex is constant across all Ps So Any differences in DV across levels of Tx cannot be due to P5 factors individual differences 0 More powerful than betweensubjects design Less error variance than between subjects design 0 Fewer Ps than betweensubjects design Disadvantages o OrderCarryover affects When previous Tx alters behavior observed in subsequent treatment PracticeLearning if P task in rst Tx performance likely to be better on Subsequent Tx Ex get better at learning list words Fatigue if performance in earlier sz drains energy performance in later sz may deteriorate Ex tiring after measures of grip strength Habituation Repeated exposure to a stimulus can lead to reduced responsiveness to that stimulus 0Ex Reaction to loud noise Contrast Exposure to the rst treatment affects response to second treatment 0 Ex rating attractiveness of faces SI March 15 2015 need hep Latin square 0 Modi ed counterbalancing with limited set of order 0 Rules 0 Each condition must appear at each ordinal level 0 Each condition precedes and follow each other condition one time 0 Number of order number of condition Other disadvantage of repeated measure design Counterbalancing does not solve everything 0 Loss of naivety 0 Problems with cover story Take test twice 0 Procedures causing lasting or permanent change SI March 27 2015 Even more good experimental research designs Factorial designs design with more than one IV 0 They are called factorial because they have multiple factors le o Advantage Allows investigation of complex relationships between variables Study more than one thing at a time o Disadvantage Complex relationships are more difficult to interpret than simple ones Things are harder to de ne Not impossible just harder o What are the effects of the amount of studying 5hrs vs 10 hrs and having eaten breakfast no breakfast vs breakfast on test performance DV Test performance of IVs TWO 0 1 Amount of studying 2 breakfast How many levels Each has two levels P o Terminology levels of lV1 x levels of lV2 o In our test performance example 2 x 2 pronounced 39two by two 2 amount studying 5 hrs vs 10 hrs X 2 breakfast none vs breakfast 0 ln factorial designs multiply levels of all le to determine of experimental conditions ln 2 x 2 design how many conditions are there FOUR o In this example 4 is the minimum Eaten breakfast No breakfast Breakfast Amount of hours gtlt gtlt 5 hrs study 10 hrs study So Ex Was a 2 X 2 what about 0 2 x 3 How many le TWO How many levels of each One has TWO the other has THREE How many condition 23 SIX What does it look like V2 LVL 1 LVL 2 LVL3 3 X 3 o How many IVs TWO 0 How many levels each IV Each has THREE levels 0 How many conditions 33 NINE V2 Lvl 1 lvl 2 lvl 3 LVL 1 LVL 2 LVL 3 IT DOES NOT MATTER WHERE YOU PUT IV 1 AND IV 2 only depending on which level has what 0 2 X 2 X 2 o How many IVs THREE 0 How many levels each IV Each has TWO levels 0 How many conditions 222 EIGHT v2 Leve1 V3 Level 2 LVL 1 LVL 2 LVL 1 LVL 2 2 X 3 X 4 X 2 o How many IVs FOUR o How many levels of each IV TWO THREE FOUR TWO 0 How many conditions 2342 48 0 We are not going to draw it but we draw it the same as the previous one 0 Results In design w two IVs like our ex there are three different effects o The rst effect The main effect of IV1 Amount of studying on DV which is test performance 0 The second effect Main effect of IV2 Eaten breakfast on DV which is test performance The number of a factorial design or any design possible is equal the the number of IVs main effect is to of IVs 0 Interaction between IV1 and IV2 on DV The whole reason we do a factorial design is to look at the combination of hrs studying and breakfast vs no breakfast 0 Main Effects how many are possible in 0 Ex a study looking at the effects of color red vs blue on mood One INDEPENDENT VARIABLE ONE MAIN EFFECT ONE 0 Ex A study looking at the effects of color red vs blue vs greed and snack cookies vs carrots in mood Color 2 levels Snack 2 levels TWO 0 Ex A study looking at the effects of feedback or noise quiet vs noisy and heat 70 vs 80 vs 90 degrees on aggression THREE o When the effect of one IV depends on the level of another IV 0 Ex Effects of having eaten breakfast on test performance depends on whether the person studied for 5 hrs or 10 hrs 0 Key things to remember about interactions lnteractions are between IVs not levels of IV 0 If there are 2 IVs only 1 interaction is possible 0 It s a rule If you have more than 2 IVs than more than 1 interaction is possible o How many interactions are possible in the following 2X2 ONE 2X3 ONE 4X4 ONE 2X2X2 FOUR example of 3 students and how they can interact 2 then 2 then 2 or all 3 FOUR 0 Q Before spring break W majors and business majors study the same of hrs per week but after BS W students study more hrs per week than business students What the effect of SB on studying It depends It depends on major Which major studies more It depends Depends on if it s before or after spring break 0 Q Rookie referee calls are more signi cant more inaccurate during the NCAA tournament than during the NCAA tournament but veteran Who is more accurate rookie or veteran ref It depends on conference Which conference has more accurate calls Depends on ref SI March 30 2015 Understanding interactions Factorial Designs Interaction When the effect of one IV depends on the level of another IV 0 Ex breakfast and test performance 0 Presenting information I I F a39 Number represents Within cell tests performance kfast Amount of Studvinlt 5 hrs study 50 50 10 hrs study 50 I Marginal cell 50 Assume equal cell sizes in this class but 0 Must say condition must be met for same cell sizes Example one Q I I I 5 hrs 10 hrs I No breakfast I Breakfast Look at the marginal means to investigate main effects Look at within cell means to investigate interactionsls 0 Graph within cell means and it will never leave you astray Eaten Breakfast No breakfast Breakfast Amount of Studyir 5 hrs StUdy 10 10 hrs study 90 10 50 To know what is considered signi cant it would be given in pic its in orange 0 I have a signi cant amount of study Those who study 10 hours do signi cant better than those who study 5 hours 10 50 0 There is a signi cance in in the amount of breakfast Those who has breakfast do signi cantly better than those who do not 0 Graph the within 0 Label everything clearly o Is something going on in one level of the IV than the other level of the IV Yes So we have an interaction 0 Once you have the graph and compare you have to describe the interaction by walking your way through it We have an interaction between amount of study and breakfast Such that at 5 hrs of study there is no difference in test performance between people who did not eat breakfast and those who did However among those who study 10 hrs and those who ate breakfast did signi cantly better than who did not Example one 5 hrs 10 hrs I No breakfast I Breakfast 0 30 70 60 C530 Eaten Breakfast o No breakfast Breakfast Amount of Studyir 5 hrs StUdy 10 10 10 hrs study 90 50 m 90 o We can tell that any difference was because of the amount studying not if eaten breakfast or not Example one Q bo V0 o 1 5 hrs 10 hrs I No breakfast I Breakfast 39 between amount of study and breakfast Such that at 5 hrs of study there is no difference in test performance between people who did not eat breakfast and those who did However among those who study 10 hrs and those who did not eat breakfast did signi cantly better than who did Eaten Breakfast o No breakfast Breakfast Amount of Studvir 5 hrs StUdy 90 50 10 hrs study 90 10 E 50 Both graphs look the same We are decreasing by 40 in each both no breakfast did better so there is no interaction 0 5 hrs increases by 80 and 10 hrs decreases by 80 0 Cross over or disordinal interactions Example one b9 5 hrs 10 hrs I No breakfast I Breakfast Eaten Breakfast o No breakfast Breakfast Amount of Studyir 5 hrs StUdy 50 4O 10 hrs stud y 90 80 50 7O If you get a signi cant interaction at one level differ in magnitude We have an interaction between amount of study and breakfast Such that at 5 hrs of study there is a difference in test performance between people who did not eat breakfast and those who did However among those who study 10 hrs and those who did better but not signi cantly 0 Looking at pattern of data is important Sl March 1 2015 Campaign Slur Tactic No slur Slur PoliticalAf iction Democrat 52 141 Republican 25 11 215 38 0 Mean differences gt30 points are signi cant Sprinkles No sprinkles Sprinkles Ice Cream Flavor Vanilla 15 14 Chocolate 36 28 0 Example three Sprinkles No sprinkles Choc Sprin Rainbor lce fife an Flavor zinillal t 22 oco a e I 20 12 31333 E 165 25 38 o l have a main affect between no sprinkles and choc sprink l have a main affect between 0 Suck that ice cream with rainbow sprinkles is eaten sig more than ice cream with choc or no sprinkles But there is no difference between in amount of ice cream eaten between chocolate sprinkles and no sprinkles 0 There is a sig interaction between sprinkles and avor on number of scoops eaten Such that when there are no sprinkles people eat slight more ice cream than chocolate sprinkles However Cho Sig more than van Rainbow is the same choc and vanilla Graphong different types of variables 0 Catagorical le They fall into discrete category one or the other no in between O 0 Ex political af iction ice cream avor Bar graphs appropriate Continuous iv aka predictable variablequot No distinct categories it s a range 0 Ex selfesteem measures on a scale from 16 Line graphs are apocopate Describing graph in pic For the people who studied 5 hrs the higher self esteem the higher higher their test performance When they study for 10 hrs there is no relationship SlAp l32015 0 Interaction the other way around 0 Which of these re ects an interaction 0 A As humans grow older their height declines In this example what is the DV Height Age on the x axis took a picture There is no interaction just shows a relation 0 B Women s selfesteem is lower after graduation from college than before graduation but men s selfesteem does not change from before to after graduation What is the DV selfesteem on the y axis There is an interaction took a picture 0 C UB students respond more positively to print ads than radio ads but Buff State students respond more positively to radio ads than print ads What is the DV positivity of response on the y axis School or college two levels Type of ads two levels There is an interaction 0 D As anxiety level increases performance on an easy task increases but performance on a hard task decreases DV Performance on the y axis Anxiety is continuous It is an interaction because they cross 0 E People like lattes better than plain coffee and lattes better than tea but there s no difference in how better than how much people like plain coffee and tea What is the DV Liking on the y axis On the test if there is only 1 independent variable automatically there is no interaction 0 F When drinking lattes people like them more in the morning than at night but when drinking plain coffee or tea there s no difference in liking based on time of day morning vs night DV liking y axis Type of drinks is categorical There is an interaction Liking in lattes is signi cant in the morning than at night And for plain or tea does not matter time of day they are liked the same 0 Mini quiz answers 1b 2a 3a 4d One look at factorial designs and main affects 0000 Two interactions if there are 2 lV s one interaction more than 2 people more than one interaction Three repeated measures mean you are using the same participants If I wanted to do it between 266 20120 What research uses content analysis archive research SI lecture 1 o QuasiExperimental Designs 0 quotkindofquot quotsortofquot quotalmostquot experiments 0 there39s only partial control over IVs E P5 are Designs where put into one or more conditions by some means other than random assignment 0 Because of this it makes it not an experiment 0 Why use a Quasi 0 Cannot always do true experiment E cannot assign gender 0 Does this mean we shouldn t study gender differences E Cannot assign natural disasters E Cannot ethically assign 0 HIV smoking pregnancy abortion diet exposure to toxic chemicals prejudicial attitudes o QuasiExperimental Designs vary in quality 0 Some allow for stronger conclusions than others If you can39t draw strong conclusions it39s been a waste Even if its quotthis had a negativepositiveno affectquot Determination of cause a NO 0 But it is possible to eliminate many alternative explanations 0 Need to know about the bad designs so you can Avoid using them Design better quasiexperiments Bad quasiexperimental Designs In til it39s rare in the medical field it is Psychologists are more trained in experimental designs 0 One group posttest Treatment Response posttest Pain Pain Management H program Ratmg 5 quotgoodquot due to treatment vvnyr You have nothing to compare it to This design is of virtually no value 0 Onegroup pretest posttest Pretest Treatment Response posttest Pain Pain Pain Rating gt Management Rating program again If response gets better t NO 0 Why Because people are already showing up with high pain it could be a lot of other variables that made them feel better We don39t know what caused the change 0 Only a small improvement over OneGroup Posttest Rival hypotheses threats to internal validity in onegroup pretest posttest designpart of bad Quasi experiment Can screw experiment 0 developmental or experiential changes in P5 between pre and posttest Ex Growing older hungrier tired or skilled 0 Changes due to an event that curs between pre and posttest Ex 911 lay offs in a company o Changes due to having taken pretest Ex Realizing performance was low or rating were high greater awareness of behavior can affect the score of the posttest o Changes due to differences in measuring instruments Ex Changes in calibration raters raters slip over time scoring procedures things are scored by people like professors grading an exam o Changes due to the loss of P5 from pretest to posttest Ex When Ps drop out after pretest or they die o Changes due to unreliability of observed scores Ex When Ps score very low or very high on a pretest then more toward average on the posttest Ex Rookie of the Year SI jinx Probem Difference between observed scoreperformance and true score oTrue scores more stable than observed scores oRTM especialya problem when Ps selected on the basis of extreme score legressio to the mm 7 Concusion Is it effective No it39s a bad design SIAp l102015 Better QuasiExperimental Designs Nonequivalent control group PretestPosttest design o Investigates differences between groups not randomly assigned to conditions 0 a Design is strengthened by including pretest AND a comparison group 0 PC Ex Kyrchenko et al 2006 Investigated effectiveness of HIVAIDS education intervention program P5 1516 year olds at two different high schools in Ukraine 0 One HS had the program other did not control Program Six sessions 1 per week focused on 0 HIV transmission 0 Dangers of drug use 0 Preventative measures Partner condom negotiation PretestPosttesL o HIVAIDS knowledge 0 Attitude towards those w HIVAIDS o Selfef cacy for maintaining safer behaviors Posttest assessed at two different time periods 0 23 days postintervention o 3 months postintervention missed a pic Kyrychenko et al 2006 Selfef cacy Tl Baseline Pumas 1 Profiled Did not get picture of ATTITUDE graph looked similar to knowledge 0 Intervention program was associated with signi cant improvements in all 3 domains knowledge attitude and selfef cacy Time series desion o Interrupted Time Series Researchers makes several observations 0 over time prior to and immediately after a treatment or event 0 01 02 O3 O4 EVENT 05 O6 O7 08 0 It39s the same design we are just making it stronger by having multiple observations Ex Berkowitz amp Macaulay 1971 o Examined crime rates before and after Kennedy assignation 0 Method FBI records in major US cities including Buffalo 0 Violent Crime homicide rape aggravated assault robbery if 2 a 3quot 5 393 Z quot391 O I LL J quotJ q 1 I Pattern shows an association between JFK assignation and increases in violent crime 0 Nonviolent crimes did not show the same pattern petty theft Supports idea of violence contagion 0 Control seriesVariations on ITS that involves including some kind of control group Ex Observe graf ti in two high schools O 01 02 O3 04 NO GRAFFITI PROGRAM 05 O6 O7 08 O 01 02 O3 O4 ANTIGRAFFITI PROGRAM 05 O6 O7 08 Advantage over ITS Comparison group Better QuasiExperimental Designs continued 0 Person x by treatment Measure at least one IV and manipulate at least one other IV 0 Aka IV x PV factorial o How done Measured variable PV 0Assessed priorduring experiment Random assignment to Tx condition IV 0 Ex Major Quinton amp Schmader 2003 Goal Investigate factors that increase recognition of prejudice Method 0 P5 women 0Creativity test male evaluator all Ps get quotDquot Design 2 x 3 Factorial 0 Group identity PV Low vs High 0 Cues to prejudice IV No clues vs Ambiguous vs Blatant Cues DV Attributions to Sex Discrimination P identification gun w WWW mm Arrhiguous Blatant Cues No to Kev ndings 0 Group identirty associated with increased ATD when prejudice is ambiguous 0 Individual differences play role in weak situations not strong ones Drawbacks of quasi experiments They are nonexperimental they are correlational Developmental research Design Evaluate changes in affect cognition and or behavior related to changes in a person39s chronological age Special case of Quasiexperimental design where age is the PV Also have applications outside dev psy Ex looking at time as the PV How does passage of time relate to adjustment to life event ex tragic life event How does intelligence change with age Hypothesis inteligence increase steadily during childhood adoescence levels off in adulthood and decrease l old age How would you study this question Crosssectional Designs involve selecting different Ps from each of a number of age groups Groups are create based on the age of P5 at the time of the study Ex15 years 30 years4560yrs75yrs Longitudinal design foow same group over time invoved following a single group of PS over same time period Ex 2011 age 6 2020age 15 2035age 30 Advantages and Disadvantages of each Developmental Design 0 Cross sectiona design advantages can collect useful developmental data in short period of time Disadvantages Age difference may not be due to chronological age but to generational differences Cohort generational effects 0 Difference in experience bw people of different age groups 0 Alternative explanation for difference Ex older people fewer educational opportunities Longitudinal Designs Advantages provider best data for studying age related changes cohort same throughout Disadvantages ong time needed to collected data expensive participants mortality Disadvantages of longitudinal Design Testing Effects Tony grew up in east end of England dreamt of becoming a race jockeyhorse rider at 15 he left school to work at a race horse events at 35 him and his Wife are cab driver Tony also does some acting in small parts they have 3 children Tony parent had passed away he works very hard and feels that the economy is taking advantages of him he wants to leave England At 42 his marriage was in serious trouble and at age 49 he is enjoying his life and marriage Neil suburban boy dreamt of going to Oxford but didn t get in dropped out of university after 1 year ended up homeless By age 28 Scotland still homeless began to exercised his religious value Still homeless at 35 but by 42 he became a politician later in life he joined a church by 49 he had left London he had a great friend name Brue but they distant apart Wish he would be met someone special religion gave him peace of mind Suzy Very well to do background she cynical about married she met a man Roget get by 28 been married 27 years WRupert they had quotups and downs As a child she always wanted children by 28 she had a baby then another by 35 in her 50 she states that she feel comfortable in her own skin the upset rm has been really hard on her Miss class 15 SIAp l172015 Ethics in research A system of ethics is a set of principles for behaving in a way that is morally correct 0 To behave ethically is to do what is quotrightquot 0 It39s a potential open question so we set guild lines however it39s a hard line to draw First code of ethics in LIJ in US published 0 1953 The code has been revised several times 0 The most recent revision was in 2002 Why do we need a code of ethics Tuskegee Syphilis Study US Public Health Service wanted to study effect of untreated syphilis Enrolled 600 African American men from rural Alabama 0 They worked very hard for very little earn 399 of the has syphilis 201 did now controls Study ran 1932 to 1972 40 years Men tod procedures ex Spinal taps were quotspecial free treatmentquot 0 However the men Ps did not know that they had syphilis Penicillin was available in 1943 o It becomes the standard cure for syphilis by 1951 The men where never told that they had syphilis or that treatment was available 0 They were told that they had quotBad Bloodquot Extraordinary measures were taken to deny treatment for length of study 0 Researches did They ordered doctors not to treat them Broke state and deferral law by preventing men from getting treatment when drafted during WWII Death rate of those with syphilis twice that of controls In July 26 1971 the research got exposed 0 When it was exposed 128 men had died of syphilis or its complications 40 wives have been infected 19 children contracted disease at birth Lawsuit against US govt settled for 10 million in 1992 0 Men living w syphilis 37500 0 Heirs of deceased men w syphilis 15000 0 Living men who served as control 16000 0 Heirs if deceased men who served as controls 5000 US Govt ordered to 0 Provided cont 0 PICTURE 0 Public acknowledgement 0 Presidential Apology in 19797 0 Announced grant program at Tuskegee U Effects of Tuskegee According to US Dept of Health and Humans services 0 29 of those waiting for organ donations are African American AA o In 2011 only 14 of organ donors where AA 0 Greater chances of compatibility win ethnic group Distrust in medical system is related to donationdonation intention ex Krux et all 2007 Russell et al 2012 Guatemala STD Study 0 Revealed in 2010 0 Samples infected by researchers Lawsuits currently pending Willowbrook Institution Study 0 To study progression and treatment of viral hepatitis 0 Study ran 1956 to 1070 at Willowbrook Institution for Mentally Retarded Children Staten Island NY 0 Why 0 Study was initiated because hepatitis rampant at institution These kids do not use the bathroom properly and they thought that it would be easier to study them 0 1 to 10 new admissions purposely infected with hepatitis Early fecal extract Later Puri ed Form 0 Parents forced to consent 0 Justi cation Children would likely contact disease anyway Sl lecture 2 Watson amp Raynor39s 1920 Little Albert Studies To study whether fear could be conditioned Participants 11 month old quotAlbert Bquot Littler Albert has no natural fear of white rat Steel bar hammer 0 Become terri ed of rat Showed that fear generalized justi cation About Albert would learn fears anyway What ever happened to Little Albert 0 He wasn t deconditioned is he still afraid of animals quotTearoom Tradequot Study HumPhI39Y 197 Aim To study those who engage in impersonal sex acts quotTearoom Tradequotquot refers to such acts 0 Men public restrooms 0 Sometimes no words exchanged Humphry39s role Wrote down the men39s license plate 5 Humphreys contacted the men Conducted interviews at men39s homes Descriptions made men easily identi able Unveiled previously hidden underground behavior 0 50 of these men identi ed as heterosexual ha wives and families 0 By what means did we get this information Milgram39s 1963 Classic Obedience Studies MTG investigate people39s blind obedience to authorities Participants quotTeacherquot Confederate quotLearnerquot Teacher ordered by experimenter authority gure to shock leaner for incorrect answers Ps told to continue even when they asked to withdraw 0 Being told to stay in the experiment was the point Clear psychological distress observed but study continued Were all Ps told the truth about the study Zimbardo39s 1972 Stanford Prison Simulation Aim To simulate prison experience record behaviors Guards 0 Uniform Billy clubs power Prisoners o Locked in barren cells forced to wear humiliation out ts Guards became cruel prisoners rebelled andor broke down Experiment was supposed to go on for 2 weeks had to be called off in 6 days SI lecture 3 What did we learn from Lil studies Often found out counterintuitive and surprising ndings 0 Fear responses can be learned 0 A whole subculture that was remained unstudied for the longest time the bathroom sexual acts 0 Obedient and obeying authority is done by majority Roles guard or prisoner shaped them quickly and intensely Scienti c merit importance 0 Underlined understandings Ends justify means 0 Does the harm done worth all the new info 0 It s a balance 0 We have a need for ll knowledge 0 We have to consider Ps rights mostly humans but animals as well Ethical consideration 0 1979 The Belmont Repost was published 0 Set guidelines to protect rights and welfare of P5 in behavioral and biomedical research 0 Previous guidelines to 1979 were existing but not followed Serious violations occurred anyway 0 You can t anticipate everything 0 You don39t know whats going to happen you really don39t know every harm in doing this That39s why you need to be very careful from the beginning APA Codes of Ethics 5 Basic Principles Principle 1 Respect for persons and their autonomy 0 Researchers have responsibility to ensure that 0 Ps know what they will get into 0 If you are going to hook them up to equipment when they are going to talk to other Ps they must know all this ahead of time P5 are free to choose whether to participate or not Informed consent Tell Ps about the study and get their written agreement to participate 0 Basically a contract saying this is ok 0 Many situations where informed consent is unnecessary or impossible Ex doing something hat is in the public domain so its ne Like civil war soldiers they are dead 0 Using public records is ne no consent Ex Full informed consent is impossible when you are using deception You tell them about the research but you don39t tell them the part that is deception that parts that are fake Principle 2 Bene cence amp Nonmale cence Bene cence quotDoing goodquot Nonmale cence quotNot doing harmquot 0 Must consider Costsrisks and bene ts of doing any study 0 What do we compare it to Everyday life Costsrisk of NOT DOING a particular study o If we don39t study a study could it harm us more 0 Should we not study somewhat controversial and could do hard O 0000 Principle 3 lustice New medications and treatments Sexual activity Obedience Helping behavior Predictors of suicide 0 The burdens and bene ts of research should be distributed fairly across those involved Injustice of Tuskegee What if the Tuskegee Study tested effectiveness of new drug eg penicillin to cure syphilis Methodologicalstrategy Better Strategy 0 Yes comparing to another treatment 0 Decision to includeexclude Ps must be justi ed on scienti c grounds Principle 4 Fidelity amp Responsibility 0 Researcher must honor relationship of trust with Ps R must follow informed consent 0 You said 3 tasks you do 3 tasks You don39t throw in extra because there is time o R keeps Ps responses con dential If deception involved Ps must be fully debriefed During the debrie ng O O O R tes P true purpose of study R answers all the P5 questions R explains why deception necessary was used 0 After debrie ng O O Principle 5 integrity P should leave with sense of dignity intact P should not feel that hisher time has been wasted o Pursuit of accurate knowledge is an ethical objective 0 Poor quality research Waste of resources 0 Description of ndings Fraudulent research Are researchers held to these ethical principles 0 YES 0 How Any institution that receives public funding is observed by an IRB Institutional Review Board IRB approval 0 Scientists and nonscientists R submit research proposal 0 Member review proposal judge ethical appropriateness 0 They can veto or shut down a research at any time Ethics in the real world 0 What about quotRealityquot TV 0 Fear factor wipe out bridal plasty survivor 0 They could all cause potential harm but at what cost Entertainment ISSED CLASS FRIDAY APRIL 24 Sl lecture 4 Stats refresher Scales of Measurement 0 Nominal Levels of variable are different categories a Sex Malesfemale m Handedness leftright m Answer Yesn0 m Behavior presentabsent a Race language political party college major 0 Ordinal Levels of variable arranged in quantitative order m Ranking of ice cream avors Pr0f 0395 1 Spum0ni 2 Coffee Chip 3 Pistachio 4 Hazelnut 5 Chocolate chip 0 Interval levels of variable arrange in quantitative order AND intervals between levels are equal in size a How much do you like Spum0ni ice cream c not at all 1 2 3 4 5 very much Here the interval or the difference between 1 and 2 is the same as between 2 and 3 0 Ratio Levels of variable arranged in quantitative order intervals between levels are equal in size AND there is an absolute zero point Why an absolute zero important Because there is an absence in the variable a How many ounces of spum0ni ice cream did you eat 0 002 102 202 302 402 502 Why are scales of measurement important o It determines the statistics and statistical tests you can use 0 Stats has guild lines on how you can use it and know when you can use them and on what level data Descriptive stats 0 Measure of Central Tendency Tell us about the location of central or typical values 0 Mean M average score Can only be calculated on interval or rati0 Outliers can only affect the mean 0 The mean is sensitive to 0utiers 0 Median Mdn Sc0re that divides group of scores in the half or middle Ordinal interval or ratio 0 Mode Most frequently occurring value All 4 can be used n0minal 0rdina interval rati0 PIC DIAGRAM Measure of Variabilitv Tell us the amount of spread in a set of scores o Variance squot2 aka quotmean squarequot Average deviation in a set of scores from the mean in squared deviation units lt39s calculated by computing the mean of the squared deviations of the scores from their mean 0 We have to square it so we don39t cancel each other out o This is not user friendly 0 Standard deviation 5 Average deviation in dependent variable units Calculated by taking the square roof of the variance PIC DIAGRAM Correlational Analysis 0 Correlation coef cient 0 Range 10 t 10 0 Tells you Direction and strength of linear association between two variables Ex Correlation between Exam1 Exam 2 scores 70 Extra Credit 2 score 02 Exam 2 score 35 Partial correlation Allows investigation of relationship between two controlling for a third variable 0 Ex Look at the relationship between extraversion and selfesteem controlling social connection extraversion PIC DIAGRAM lts partial because 48 went down to 21 not all the way down to 0 PIC DIAGRAM same test run again 0 We like it because it allows us to look at the missed 0 Regression Equations 0 This has many uses 0 Can be used to predict Y when X is known Ex If there is a relationship between height at age 8 and height at age 18 can predict at the 18 from height at 8 0 Multiple regression aka multiple correlation It allows you to look at a combination of factors as predictors and each independently of others 0 m Predict height at 18 from LOOK AT ALL THESE FACTORS Height 8 amnt of protein in diet 8 fathers height mother height siblings height 0 Ex Blascovish et al 2004 0 Question What psychological states predict athletic better vs worse OOO performance 0 Sample College baseball and softball players both amlea nd female 0 Method asked to imagine clutch situation 0 Conference championship game bottom of the 9t 7th inning Two outs tie game runner on 3rCI You are up to bat O 0000 Give speech about that and Why you make a good friend control topic Withtin 55 design speech order counterbalanced DVs Psychological measures of challengethreat Positive state eyes dilate blood vessels dilate allowing more blood to ow through the body Threat state measures Performance subsequent season runs created Analysis multiple regression predicting performance Predictors PIC DIAGRAM Findings PIC DIAGRAM Serena Williams is a clutch game player she will nd a way to win The other dude is a British player not as good known as a choking playerclutch win or lose in last minute choking student will he gave up SI lecture 5 lnferential Stats Basic Concept lnferential Statistics Designed to determine whether results from sample data are generalizable to a population 0 Key question Can you infer that results from vour sample Would be what you would get with other samples from sample population Logic of lnferential Statistics ETWO independent groups Experiment 0 5 College students 0 M No caffeine vs 80 mg caffeine 0 Test performance 0 Before IV administered Test performance means from each group viewed as 0 Sample means Assumed to represent means of the underlying population college students At this point group should be equivalent 0 However difference between any to samples means rarely zero 0 Random error error due to chance not the IV 0 After IV administered Each sample mean is assumed to represent the mean of its underlying population N0 caffeine population college students Caffeine population college students on 80 mg caffeine Any difference between sample means re ects 1 True difference in test performance between populations 2 Random error Use inferential statistics to do the following 0 Make judgements about differences between populations based on our sample data Ex college students vs college students on 80 mg caffeine 0 Determine probability that difference in samples means is due to random error What else would difference be due to The work ofthe IV Hypothesis testina 0 State null hypothesis 0 Hsub zero u of no caffeine group uof 80 mg of caffeine group Says IV has no effect any difference was due to random error 0 State the alternative research hypothesis 0 Hsub 1 u of no caffeine group u of 8 mg caffeine group Says IV has affect 0 We reject the null hypothesis when probability that results are due to random error is LOW lt05 aka less than 5 0 quotSignificantquot result has low probability of occurring if population means are equal Errors in hypothesis testing 0 Key point decision to rejectretain null hypothesis is based on probabilities 0 Don39t know true population means 0 So in using inferential stats might make error 0 PIC DIAGRAM Types of error 0 Type 1 error rate is equal to chosen signi cance level 0 Lower alpha level lower type 1 error rate 0 Type 2 error rate affected by 0 Alpha level As one type error goes up the other one goes down and vice versa It39s a delicate balance 0 Sample size Larger sample size less type 2 error 0 Effect size matters Stronger the effect the less type 2 error SI lecture 6 Lecture outlines Inferential statistics decisionmaking how does science deal with error Type of error Risk of Gullibility fallingvs Blindnessmissing something Type 2 is I don39t see Often scientists the US justice system fear gullibility 9Type1 errorgt blindness Type 2 Error Ex Fear sometimes realized Anthony Capozzi has mental disorder and 2 decade Type 1 error Alternative hypothesis is he didn39t do it In class exercise 1 Null hypothesis there is nothing there 2 Alternative hypothesis there is something there 3 Type 1 error saying you are pregnant but you are not pregnant 4 Type 2 error opposite of Type 1 error 5 Two types of correct decisions correct about the null and alternative Scenario A You are social worker must decide whether to pull children from home where child abuse is suspected Scenario B you are a lab tech responsible for running Important of replication Single study does not carry much weight Science is selfcorrecting Research goals Immediate Goal Assess relationship between IV and DV lnterna validity Extent we can meet this goal what we need Longranged Goal Generalize results to and across different persons settings times Externality validity Extent we can meet this goal it help with random sample Issues The rarity of random sampling Most often used in population surveys some correlational studies Rare in experimental research Butwebbased studies Problem Sample may contain characteristics that threaten external validity Population Validity Ability to generalize from study sample to larger population of interest Target population larger population of interest Ex All college students Experimentaly accessible population population available to research ExUB undergraduates Population Validity Two inferential steps Sample gtstep1 experimentally accessible population gt Target population Step1 Generalizing from sample to EA Pop Step 2 generaize from EA pop to T pop Convenience samples Organisms most often used in psychological research Sprague Dawey albino rat College student Key Question Are there good reasons why effect or relationship would not be found wother types of subjects Ecological Validity Ability to generalize result of a study across settings ex Generalizing from lab to real world Problem Arti ciality consider Festinger and Carismith 1959 Paid 1 vs 20 to lie about how fun it was to turn knobs Have you ever done something just to impress a friend and then convinced yourself that you like it Mundaneeveryday Realism Similarity of an experimental setting to reallife Does it look like the real world EXPERIMENTL Realism Psychological impact of experimental situation does it feel the real world Most important type of realism How Real are Experiments Lab setting may overestimate underestimate magnitude of effects Exab to real world might change Effect of audience or serious consequences on attributions to discrimination ATD ay overestimate ATD in lab Research on relationship dynamics among couples May underestimate problem in lab Temple Validity Extent to which results can be generalized across time Problem social political structural trends may affect generalizability Ex Political attitudes and activism Gender roles Life goals Expression of certain attitudeseg prejudicial poitica correctness and the bogus pipeline Changeable characteristics Generalizing Result Matter of degree not absolute


Buy Material

Are you sure you want to buy this material for

75 Karma

Buy Material

BOOM! Enjoy Your Free Notes!

We've added these Notes to your profile, click here to view them now.


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'

Why people love StudySoup

Bentley McCaw University of Florida

"I was shooting for a perfect 4.0 GPA this semester. Having StudySoup as a study aid was critical to helping me achieve my goal...and I nailed it!"

Anthony Lee UC Santa Barbara

"I bought an awesome study guide, which helped me get an A in my Math 34B class this quarter!"

Steve Martinelli UC Los Angeles

"There's no way I would have passed my Organic Chemistry class this semester without the notes and study guides I got from StudySoup."

Parker Thompson 500 Startups

"It's a great way for students to improve their educational experience and it seemed like a product that everybody wants, so all the people participating are winning."

Become an Elite Notetaker and start selling your notes online!

Refund Policy


All subscriptions to StudySoup are paid in full at the time of subscribing. To change your credit card information or to cancel your subscription, go to "Edit Settings". All credit card information will be available there. If you should decide to cancel your subscription, it will continue to be valid until the next payment period, as all payments for the current period were made in advance. For special circumstances, please email


StudySoup has more than 1 million course-specific study resources to help students study smarter. If you’re having trouble finding what you’re looking for, our customer support team can help you find what you need! Feel free to contact them here:

Recurring Subscriptions: If you have canceled your recurring subscription on the day of renewal and have not downloaded any documents, you may request a refund by submitting an email to

Satisfaction Guarantee: If you’re not satisfied with your subscription, you can contact us for further help. Contact must be made within 3 business days of your subscription purchase and your refund request will be subject for review.

Please Note: Refunds can never be provided more than 30 days after the initial purchase date regardless of your activity on the site.