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


by: Nellie Runte


Nellie Runte
GPA 3.8


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

Class Notes
25 ?




Popular in Course

Popular in Psychlogy

This 16 page Class Notes was uploaded by Nellie Runte on Thursday October 22, 2015. The Class Notes belongs to PSY 7 at University of California Santa Barbara taught by Staff in Fall. Since its upload, it has received 11 views. For similar materials see /class/227111/psy-7-university-of-california-santa-barbara in Psychlogy at University of California Santa Barbara.




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: 10/22/15
Unit 3 Monday February 24 2014 Chapter 10 o increase number of levels of IV 2 levels can39t provide much info about exact form of relationship between IVampDV o If curvilinear relationship is predicted least 3 levels must be used Factorial Designs designs with more than one Vorfactor All levels of each IV are combined with all levels of other IV simplest 2X22 IV each having two levels 0 ex 0 0 IV misleading or unbiased 2nd IV knowledge of the crime 3esL naive questionermisleading Questions naive questionerunbiased Qs 3X39 conditions main effect effect each variable has by itself one for each IV overall relationship between IV and DV o averaging across all participants in each group o Mean across for main effect and down for main effect interaction no interaction between 2 IV effects of one IV depends on particular level of the other variable graph means for all conditions Factorial design experimental manipulated and non experimentalmeasured or non manipulated IV X PV IVXPV one manipulated IV least 2 levels and one participate variable with 2 levels PV age gender ethnic group subject variables or attribute ex extroverts vs introverts vs silent and distraction when studying moderator variable or operation influences relationship between two other variables ex misleadingmore errors than unbiased Qs analysis of varianceasses statistical significance of main effects and interaction in factorial design a Simple main effects examines main effect of IV averages across levels of other IV results are analyzed as if we had separate experiments at each level of the other IV a Two ways assigning participants to conditions 1 Independent groups design differ participants assigned to each conditions of study 2 Repeated Measures Design sam individual to all conditions 3 mixed factorial design combination of two o ex of mixed extraversion independent distractionrepeated o 2X36 conditions ex easy hard and anxiety levellow mod high LECTURE Tues Feb 25th Monday February 24 2014 Factors and levels Factor is another term for IV Experiment randomly assign subjects to drink coffee or drink decage One Factor whether or not drink coffee o Levels refer to different conditions within a factor 2 levels in coffee example regular vs decaf experiment with a 3rd group drink caffeinated soda would have 3 levels but still one factor o One Factor Design with gt 2 Levels Basic Reasons for having 2 levels 1 Need more than 2 to deter curvilinear or complex relationships between variables o 2 May want multiple control groupsbe interested in multiple conditions Bandura Study of Imitative Aggression Children35yrs assigned to 1 to 3 Conditions Observe aggressive adult beat up Bobo Doll Observe nonaggressive adult not beat up Bobo Doll No exposure to any adult model Matched Pairs Design children were enrolled in Stanford s nursery school rated by teachers for degree of aggressiveness matched for aggressiveness before being RA to conditions Hypothesis 1 kids exposed to aggressive models would perform more aggression than those in other 2 groups 2 exposure to nonaggressive model would inhibit aggression prediction no expose kidsgt aggression than nonaggressive condition Dependent variable measured in different room in the absence of adult models Prior to test for imitation however all subjects were subjected to mild aggression arousal to insure that they were under some degree of instigation to aggression Experimenter lets child play with toys for 2 mins then tells them these are the very best toys she didn39t let just anyone play with them and she decided to reserve toys for other kidsquot Modeling apparently doesn39t work without this added instigation observers counted of aggressive acts in specific categories behind oneway mirror over 20 min intervals FACTORIAL design involve gt 1 independent variable or factor All levels of each IV are combined with all levels of other lV Ex balanced placebo design study Factor 1 alcohol expectations Level 2 expect Vodka Level 2 expect Tonic Factor 2 alcohol Intake Level 1Drank vodka Level 2 drank tonic A 2 rows X 2 columns design Multiplicative notation 2 Monday February 24 2014 a 3X 4 designs1 factors with 3 and 4 eves12 total Main Effects Effects of each IV considered separately eg does actually drinking vodka produce higher levels of shock delivered to learner than drinking tonic main effect of alcohol intake main effects compare effects of levels within a factor averaged across all levels of the other factors Example Factorial Design RA to 2X2 factorial design DV is performance on math problems correctly finished out of 100 in specific time interval Factor 1 testing condition alone or in front of an audience Factor 2 difficulty of problems easy or hard posttest only design with RA across 4 cells Main Effects Easy Hard Aver MAIN EFFECTs of contain audience gt alone Alone 60 4O Audience 90 3O 60 75 35 Main effect easy gt hard Interaction Effects when the effect of one IV depends on the level of another IV Cannot fully understand influence of one V without reference to another V Cant fully understand effect of audience vs alone without reference to difficulty factor if easy problems better to have audience if hard problems betterto be alone Main Effects in American Idol Love song Rock Song Male 100 200 150 FgtM females do better than males Female 300 200 250 200 200 No main effect No difference in song choice Is there an interaction Graphing Result of Factorial Design For main effects compute average for each condition and plot as bar graphs which is larger For interaction effect plot all the data as line graphs nonparallel lines indicate an interaction 0 plot Female point at 300 and point at 200 and connect Male point at 100 and point at 200 and connect Balanced Placebo Design alcohol no alcohol alcohol 51 35 43 no alc 50 34 42 505 345 interaction effect lines are parallel when graphed no interaction Monday February 24 2014 only one main effect Simple Main Effect Effects examine the effects of one factor at one level of another factor Eg do subjects solve more math problems in audience vs alone when problems are easy Qs are isolated to easy problems one level of the difficulty factor Diff Factor Diff factor Alone 60 Alone 40 Aud 90 Aud Do this for American Idol Example Simple main effect for love song Simple main effect for rock song simple main effect for females simple main effect for males Arousal and MoodPerformance As Extraverts less distracted by noise College students each take 2 differ versions of reading comprehension test once with TV on in background and once in silence What type of measurement Repeated Measure Why Filed out personality test first and selected 10 introvert and 10 extraverts predicted an interaction effect extraverts would be less distracted than introverts Personality and Distraction lntrovert Extrovert Silence 910 820 TV 660 780 subtract introverts 910660 Subtract extroverts 820780 OR make line graph Line parallel no so interaction effect Example Q on Factorial Designs Researcher is interested in effects of music amp lightening on exam performance in 2X2 Independent groups factorial design subjects are assigned to take same exam with music being played bs no music and in brightly lit room vs dimly lit room Following data Music amp Bright light9 music and dim5 no music and bright5 no music and dim lights 1 Main effect of music musicgtno music main effect of lighting brightgt dim interaction between music and lighting lines are parallel no interaction Music no music Bright 9 5 Monday February 24 2014 Dim 5 1 3 3 Effect of music doesn39t depend on lighting Lecture Thursday Feb 27th Mixed Designs o mixed factorial designs have at least one withinsubjects factor and one betweensubjects factor ln withinsubjects factors the same individual appear in all levels of the factor ln betweensubjects factors different individuals appear in each of the levels of the factor o EX Non tactile exposure to females cause male rodents testosterone to increase Measure test saliva female measure testoterone saliva Measure test saliva male measure testoterone saliva We can analyze this design as mixed factorial design what are the two factors 2 Man vs women X 2 Time of measurement pre vs post mixed factorial design pretest woman posttest pretest man posttest 2 Pretestposttest X 2 manwoman Hypothesis Testosterone change from baseline will be more positive in female condition than in male condition How do we test this hypothesis in terms of main effects and interaction effects Is the hypothesis about a main effect or interaction Hypothesis is an interaction effect does the change from pretest to posttest depend on te experimental group That is is the change from pre to post more positive in female group than male group Hypothesis testosterone change will be more positive in female than male condition Say want to know whethertestosterone also went up in male condition even if less than in female conditoin what type of effect Notice in this example no main effect of pre vs post Notince in this example no main effect of male vs female Repeated Measures Factorial Design Driving simulation study with 2 factors 2 drunk vs sober X 2 tired vs rested Say want 20 subjects in each cell in repeated measures experiment How many cells how many subjects How many times is each subject tested How many possible orders of testing HOw should we deal with order effects Best way in this case Drunk Sober Tired A Rested C Um Latin Square 1 Each condition appears at each ordinal position Each condition precedes and follows each condition on time Monday February 24 2014 UOW gtUOWN 00139000 wgtUOA Drunk Sober Tired 1O 7O 4O Rested 7O 90 80 4O 80 Third Factor sex of subject 2 male vs female X2 drunk v sober X2 tired vs rested Factorial design Main effect of sex male mean20 femal30 yes main effect Main effect of sobrietygt drunk20 sober30 yes main effect Also main effect of rest restedgttired 3 main effects in this study I males Drunk Sober Tired 1O 20 Rested 20 3O 15 25 Females Drunk Sober Tired 20 3O Rested 3O 4O 25 35 3 way interaction test whether a 2way interaction differs at levels of a third factor n this example there is no 3way interaction there is no two way interaction rest and sobriety for either males or females Main effects with gt 2 levels Say in math problems example from last time 2 alone vs audience X2 easy vs hard design have a 2 alone vs audience X 3easy vs medium vs hard design 6 total cells and want to assess main and interaction effects Easy Medium Hard Alone 60 60 4O Audience 90 90 3O 7 75 35 5 Chapter 11 Reading High internal validity DV must be effect of IV Singlecase experimental design single subject design used to detwemine whether experimental manipulation had effect on single research participant Subjects behavior is measured overtime during baseline control period Manipulation then intro during treatment period behavior observed Simple reversal design to determine manipulation of IV had an effect ABA design Monday February 24 2014 Improve by ABAB design experimental treatment intro 2nd time sometimes can be impossible or unethical multiple baseline design effectiveness of treatment is demonstrated when behavior changes only after manipulation is introduced across subjects behavior of several subjects is measured over time manipulation intro at different times across behaviors differ behaviors of single subject measured over time maniple at differ times ex reward system increase socializing grooming and reading behaviors applied to each behaviors at differ times across situations same behavior measured in differ settings such as home and work Advantages of single case experimental design useful for studying many research problems and powerful alternative to more traditional research designs Espec baluable for someone applying some change technique in natural enviro ex teacher trying new technique in classroom program evaluation research on programs that are implemented to achieve some positive effect on group of individuals In schools work entire communitites ex DARE drug abuse resist edu Outcome evaluation did program result in positive outcome for which it was designed to Needs assessment studies ask whether there are in fact probekms that need to address target population ex drug abuse by children and adolescents in community Program theory program designed to address them collaboration of researchers Process evaluation impact assesment intended outcomes of the program being realizedgt Efficency Assessment Must determine if it is worth the resources it consumes needs assessment Program Theory Assessment Processl Evaluation Outcomle Evaluation Efficiency Assessment Quasiexperimental designs address need to study effect of IV in setting in which control features of true experimental designs cant be achieved Examine impact of IV on DV be there soon causal inference is much more difficult one group posttest only design one shot case study lacks control or comparison group can39t draw causal inferences about effect of IV and DV one group pretest posttest design measure before and afterward Threats to Internal Validity History any event that occurs between 1st and 2nd measure any confounding event Maturation any change that occur systematically over time Testing if the pretest changes participants behavior Instrument Delay basic characteristics of measuring instrument change over time Regression toward Mean high or ow score tested again and score change in direction of mean Monday February 24 2014 nonequivalent control Group Design employs separate control group experiment and control group rant same selection differences or selection bias when participants wo form 2 groups in experiment chosen from existing natural groups nonequivalent control pretest posttest not RA to groups groups may not be equivalent propensity score matching method in pairing individuals for assignment to treatment and control condition based upon combination of scores on participant variables interrupted time series design design which effectiveness of treatment is determined by examining series of measurements made over extended time period both before and after treatment intro at random point in time control series design extension of interrupted time series quasi experimental design which comparison or control group QuasiExperimental Design QuasiExperiments Designs that lack all of the control features of true experiments Research design in which an experimental procedure is applied but all extraneous variables are not controlled Definitions vary from person to person Inability to RA participants to conditions is most common reason for use of quasiexperiments Lower internal validity Do not allow same confidence in causal inferences as true experiments One Group PretestPosttest Design Same as pretestposttest design but no RA to multiple conditions no control condition Hot Dogs Increase IQ Scorequot Take mini IQ test Eat a hot dog Retake Mini IQ Test Design is quasiexperimental because of lack of control condition Empirical study to test effectiveness of a relationship improvement program Martial satisfaction share feeling program martial satisfaction Testimony celebrity says marriage is better Problems History any event that happens between pre and posttest is confounded with manipulation Maturation systematic changes in individuals that tend to occur over time are confounded with manipulation Ex study to test effects of vitamins on kids growth height measurements at start and end of 3rd grade with no control group Problem ex drug study to treat symptoms prepost symptoms with no control group maturation effect Maturation effect problem for share your feelings program Testing any effect of pretest on posttest ex hot dog Instrumental Decay posttest meaurement of DV has changed since pretest maturation Regression to mean occurs when scores are unusually High or Low due to Random factors at one testing occasion on next occasion random factors likely to be absent and score return to mean ex basketball player shoot 40 on average one game 90 9 out of 10 What is player likely to shoot in next game 40 change back regresison to the mean Monday February 24 2014 Problem for one group prettestposttest if choose subjects bc scored very low on prettest part of low scores may be due to chance such expect improvement on posttest even without manipulation People often misperceive instances of statistical regression Belief that punishment improves behavior praise hurts it Then why use this design Paleolithic Diet One Group PrePost Design Most of human history before agriculture humans thought to have consumed very different diet than today Lean meatsalso bone marrow internal organs uncultivated plan material fruits vegetables nuts roots No dairy no grains no processed foodsrefined sugars Modern diets in industrialized countries many new components absent until recently Cereal grains and their products nonhuman milk dairy refined carbs lots of sugar separated fats and oils Human taste for sugar and fat mean reduction in fruit and vegetable consumption Diseases of civilization in wealthy countries like US high rates of clogged arteries and associated heart disease type 2 diabetes and high blood pressure risk of stroke Study Design n9 subjects healthy nonobese mean age386 male Subjects ate usual diets for 3 days and had baseline measures taken of health outcomes dv Subjects ate palolithic diet for 10 days then measure same health outcomes Subjects maintained constant weight DV cholesterol insulin response to glucose blood pressure Pretest paleo diet posttest Results LDL Cholesterol dropped by 22 Reductions usually seen in at least 8 out of 9 subjects for each variable Results seem convincing Any problems Nonequivalent Control Group Design participants treatment DV Participants no treatment DV No RA but there is a control group CTEamp nonequivalent control groups CTEbrain disease POstmortem brain examinations find specific markersabnormal tau proteins atrophy of specific brain regions etc Symptoms inclue dementia depression abnormal aggression etc Thought to be caused by multiple bunt force traumas to head How can one design a study to test whether playing football causes CTE Randomized experiment probably precluded by ethics What is next best design Anterograde amnesia loss of ability to form new long term memories after brain injury retrograde amnesia loss of long term memories that occurred before brain injury Experimental Design Problems in Testing Brain Damaged Patients Monday February 24 2014 Nonequivalent control group Subjects matched for age education level and intelligence Flu Shot Study Background Testosterone appears to regulate energy distribution Promotes fat breakdown and energy investment in muscle construction Decreases energy investment in immune function Nonequivalent control group pretestposttest participants pretest treatmentflu shot posttesttestost Participants prestesttest controlno flu shot posttesttestost lf assignment is random to flu shot vs no flu shot then true experiment pretest posttest design If assignment is nonrandom then have quasiexperment nonequivalent control group pretestposttest design Which design has greater internal validity why True experiment THURS March 6 Hypothetical ABA Design 16 polo shirts 50 10 fancy shorts 60 hair gel priceless Opertation def Hell get more high Example of ABA Design Students behavior in a classroom Limitations of Reversal Designs Some effects of treatments are difficult to reverse reversak designs most useful for seeing shortterm effects of treatments Multiple Baseline Design ntroduce treatment at different times for different subjects How much kid is responding to panic attack Introduce treatment for first scene Panic attacks are declining Maybe maturation efffect Just getting better over time But its treatment itself because of multiple test Cant use reversal because ended response so it wont reverse Statistical Review Many research designs intended to provide evidence that 1 variable caused another In true experiment does mean score in one experimental group differ from another group Statistical significance assesses probability that results could be due to chance rather than hypothesized cause corresponds to pvalue eg could difference between 2 means be as large as its by chance Could of outcomes be as large as it is by chance alone Say we have hypothesis that someone has ESP can predict future Null hypothesis Person has no special abilities and cant predict future Research hypothesis person does have ESP and can predict future Monday February 24 2014 Data collection we have person predict outcome of one coin flip Say they get it right Does this prove they have ESP What is value for our results probability of getting this result if null hypothesis is true p50 Suppose person predicted 10 consecutive coin flips correctly what is pvalue for this result p0098 tdiffernece between groups means variability within groupssample size Analysis of Variance ANOVA computes pvalues for main effects and interactions in factorial designs Ftest is extension of ttest Computes ratio of systematic variance deviation of group means from overall mean to error variance deviation of individual scores in each group from group mean Type amp Type II Errors When assessing statistical significance two types of errors can make based on yesno significan decision Type Error incorrectly rejecting null when it is in fact correct ex psychic when not based on coin film committed type I error pvalue5 Lower pvalue considered significan less likely to reject null and less likely to commit type I error Type II error incorrectly accepting null hypothesis when its in fact false blindness to relationship Say someone is psychic and correctly predicts 3 straight coin flipspvalue p125 may not be improbable enough to say statistically significant dont reject null How could we avoid type II error in this case while finding acceptable value AS sample size increases probability of type II error decreases larger sample size have more power to detect effects that are present low power in 3 coin flip example Sample size wasn39t big enoughType ll Error Alpha is pvalue at which decide to reject null hypotheses as alpha gets larger probability of type I error increases and probability of type II error decreases Alpha05 often used in psych research as compromise is between offs of type and type II errors Pvalues are not same as effect sizesquot Correlation of r80 indicates strong relationship between 2 variables but r1O is a weak relationship Pvalue for correlations indicated probability of finding correlation as large as observed if correlation is Wiith very large seize may have very small value for r1O but still weak relation Hierarchy of Internal Validity 1 True Experiments 11 Monday February 24 2014 Posttest only Pretestpost test repeated measures matched pairs Factorial Designs 2 Quasi Experiments 1 One group prestestposttest 2 nonequivalent control group 3 nonequivalent control group presttest posttest 3 Correlation Designs Measure variables and test their association No manipulation U PP NT Practice for Final 1 What is internal validity conclude from one design that one causes another 2 Students learn better from chalkboard or power point instructor uses powerpoint in psych 7 in one quarter but chalkboard in next quarter what type of design Quasi Experiment Different group of students Say you surveyed GRE prep instructors who use 39 quot 39 39 vs 39 then 39 average GRE scores of their classes what type of design Correlation Design Say instructors who used powerpoint had classes that second higher on average Good evidence Alternative explanation Its not the chalk or powerpoint but rather the instructor Ton of lurking variables Was quasi experimental evidence better Simple random sample of psych 7 and RA half quarter tutoring session with Alex using chalkboard and half with differ TA Ture experiment Posttest Only Give example of nonequivalent control group pretest posttest design test whether scores on psych 7 exams greater with chalkboard or powerpoint 2 classroom one using chalkboard one uses powerpoint NO RA Measure knowledge of Psych 7 in beginning and end Interpretive problems with this design RA half subject watch video with powerpoint half with chalkboard take 2nd version of test on interpreting factorial design Pretestposttest true experiment Internal validity of this design possible problems or limitations Say chalkboard was better does this mean should end powerpoint in classroom External validity problem Different design have differ cost and benefits true experiment lab have highest internal validity Chapter 1214 Reading Review nominal ordinal and intervalratio scale Monday February 24 2014 3 basic ways of describing results 1 comparing group percentages 2 correlating scores of individuals on two variables 3 comparing group means Comparing group percentages ex compare males and females if like or dislike travel First calculate of females who like to travelampcompare this with of males who like to travel Then report 80 females 60 males thus relationship between gender and travel variables appears to exist then perform statistical analysis to determine if signif difference Correlating scores needed when dont have distinct groups of subjects measured on 2 variables ex location of classroom amp grades sit near front higher grades Comparing group means compare mean responses in 2 or more groups ex experminet designed to study effect of exposure to agressive adult child in 1 group observe adult model behaving aggressively while child in control group don t then children play alone for 10 mins o compare mean number of aggressive acts by children in 2 conditions to determine whether children who observed model were more aggressive than children in control condition o Frequency distribution indicates if individuals who receive each possible score on variable o Pie Charts divide whole circle or pie into slices that represent relative bar graphs separate and distinct bar for each piece of info Freq polygons line to represent distribution of freq of scores o Histograms use bars to display a freq distribution for quantitative variable o Descriptive statisticsL allow researcher to make precise statements about data o central tendency statistical tells us what sample as whole or on average is like mean indicator for central tendency variability characterizes amount of spread in distribution of scores Standard deviations indicated average deviation of scores from mean Variancesquot2 square root of variance Correlation coefficient statistic describes how strongly variables are related to one another Pearson productmoment correlation coef used when both variables have interval or ratio scale properties the rvalue linear relationships restriction of range occurs when individuals in your sample are very similar on variable you are studying o effect size strength of association between variables regression equations calculations used to predict persons score on one variable when person score on another variable is already known criterion variable predictor variable ex college grades scores on apt test multiple correlationcombing of predictor variables to increase accuracy of prediction of given criterion or outcome variable partial correlation provides way of statistically controlling third variables correlation between 2 variables of interest with influence of 3rd variable removed or partialed out of original correlation Structured equation modeling SEM techniques examine models that specify set of relationships among variables using quan non experimental methods Chapter 13 inferential statistics used to determine whether results match what would happen if we were to conduct experiment again again with multiple samples null hypothesis simply that population means are observed difference is due to random error IV has no effect research hypothesis population means are in fact not N has effect aocepting means IV has effect on DV Monday February 24 2014 null hypothesis is incorrect accept research hypothesis is correct if null is incorrect accept research hypothesis as corre null used because very precise statementpopulation means exactly statistical significance result is one that has low probability of occurring if pop means are Alpha level probability required for significance o common alpha level used is 05 o 5 out of 100 results due to random error in one sample from population T test commonly used to examine whether 2 groups are significantly different from each other F test if more general statistical test that can be used to ask whether is is difference among 3 or more groups or to evaluate results of factorial designs To use statistical test 1 specify null hypothesis amp research hypothesis 2 specify significance level you will use to decide whether to reject null hypotheses alpha level 3 t group differencediffer between obtained means under nulwithin group variability amount of variability of scores about the mean lft low prov of occurrence 05 or less nullrejected F test extension of t test ratio of 2 types of variance systematic amp error Type I error reject null but null is actually true if we use lower alpha level 01 less chance of making type 1 error Type II null is accepted although research is true 1 significance alpha level low alpha level to decrease error We increase chance of type II error 2 sample size is large effect size is large type II is unlikely Chapter 14 EXTERNAL VALIDITYgeneralizing data College students used in most studies highly restricted population Volunteers nternet Research sampling particular demographic higher wealth college education younger Gender Considerations using only males or only females for convince Locale recruited from location impact on external validity ex UCLA students vs lovva vs NYC Researchers must understand human behavior across and among cultures different results with different experimenters ex warm nice vs cold unfriendly or different sex 39 Solve by using 2 or more experimenters 39 pretesting may limit ability to generalize to populations that didn39t receive pretest 39 enables to asees mortality effects ppl different who withdrew vs who finished 39 Generalizing in lab settings impact of IV under highly controlled conditions 39 generalization to reallife settings is not relevant when purpose of study was to investigate causal relationships under controlled conditions 39 replication overcoming any problems of generalization that occur in single study 39 exact replication attempt to replicate precise procedures of study to see whether same results are obtained 39 find is findings are reliable builds on findings of proper study 14 Monday February 24 2014 conceptual replication use of differ procedures to replicate research finding understand relationships among abstract conceptual variables by using new or differ operational def of those variables Key goal in research relationship between conceptual variables exist conceptual replication same IV is opertionalized in differ way DV may be measured differ too may involve alternative stimulus or Dependent measure literature review researcher reads number of studies that address particular topic then writes paper that summarizes and evaluates literature metaanalysis combines actual results of of studies set of statistical procedures that employ effect sizes to compare given finding across many differ studies LECTURE TUES MARCH 11th External Validity external validity measures extent to which research findings can be generalize to other participants setting or ways of measuring variables Poor external validity results might still be valid butjust don t generalize Experiment Culture of Honor Study Read article on Gauchospace Cohen hypothesized that men from southern US would respond differently to an insult that men grew up in northern US Theory South has culture of honor in which expected to retaliate strongly against threats or insults or they lose status with their family friends peers economy in south historically based on herding culture of honor passed down through some cultural transmission process possible explanation for higher rates of reliatory homicide in South Hypothesis interaction effect of regional origin and insult with southerners reacting with more aggression after insult than northerns not interested in main effects Study 1 design Experiment involved in 2X2 design with participants region of origin north vs south as one variable and condition insulted vs not insulted as other variable participants 83 university of Michigan Procedure Subject fill out form in beginning of study now take form and hand it end of the hallway Confederate blocks hallway and subject has to squeeze pass him Subject comes back and confederate bumps it and says asshole Control group no insult or insult Results Word completion no significan effects observer rating no sig effects for angry ambulence story Study 2 see basic procedure as study 1 but with additional dependent measures dependent variables cortisol response stress hormone saliva samples collected results testosterone results increases cortisol results increases 1 Southers more affected by insult than northerns for measures of 1 using violence to end story about man hitting on someones fianc 2 results provide evidence for cultural differences in reactions to insults 3 alternative explanations for result Specific studies are performed in specific places amp tumes with particular operational definitions of variables replication of results in different settings with different participants or differ operational definitions increases generalizability external validity 15 Monday February 24 2014 Exact replication do exact study might not get exact results significant conceptual replications differ op def can establish boundary conditions Conceptual vs Exact Study looking at extrinsic rewards like on math performance Study uses specific rewards specific types of math prob in specific sample of students operational def of each variables Exact replication employ same operational def but use new subjects tests how robust consistent finding of this specific study are Conceptual replication will change aspect of design while still testing same general idea estrinsic rewards affect math Eg may change size of rewards types of probles length of retention Review articles of assessments of external validity Qualitative reviews meta analyses eg gender amp agression Quantitative compute average effect sizes across many studies Can compute differ effects under differ circumstances verbal vs physical aggression File Drawer Problem My experience with external validity testing results with different subject populations Mae Rats study Testestorone goes up in male rates with female present Study Design Male university of Chicago students mean age2136 UCSB vs Chicago UCSB male students 1890 yearsmean randomly assigned to male vs female experimenter condition Male don t respond to male Significant increase with female Thursday Review Session Predictable validity construct validity concurrent validity criterionoriented validity divergent no scoring high on what it should convergent similar construct discriminant Predicting behavior positive neg no relationship rO ex weight and voting curvilinear Clarifying vs alternative explanations ex ice cream amp drawing clarifying ice creamcrampsdrowning alternative temperature Basic experimental Designs posttest only prettestposttest repeated measures matched pairs Dealing with order effects in repeated measures Designs Comple counterbalancingpresent the different conditions in every possible order for 2 condigtions controlexperimental only 2 possible orders For 4 conditioons 24


Buy Material

Are you sure you want to buy this material for

25 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

Jim McGreen Ohio University

"Knowing I can count on the Elite Notetaker in my class allows me to focus on what the professor is saying instead of just scribbling notes the whole time and falling behind."

Amaris Trozzo George Washington University

"I made $350 in just two days after posting my first study guide."

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.