Study Guide Chapters 6-9
Study Guide Chapters 6-9 Psych 2300
Popular in Research Methods in Psychology
Popular in Psychlogy
This 10 page Study Guide was uploaded by Emma Dahlin on Sunday October 25, 2015. The Study Guide belongs to Psych 2300 at Ohio State University taught by Seth Miller in Fall 2015. Since its upload, it has received 206 views. For similar materials see Research Methods in Psychology in Psychlogy at Ohio State University.
Reviews for Study Guide Chapters 6-9
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: 10/25/15
Include homework assignments and know vocabulary terms O O 0 0 0 O O Surveypollmethod of posing questions to people on the phone in personal interviews on written questionnaires or online Question Formats Openended questionsljallow respondents to answer any way they like spontaneous rich info difficulttimeconsuming to codecategorize responses Forcedchoice formatljlpeople give their opinion by picking best of 2 or more prespecified options limits type of responses Likert scale people presented with statement and use rating scale to indicate degree of agreement 5 response oonns strongly agree agree neither agree nor disagree strongly disagree 15 minor variations are called Likerttype scales Semantic differential formatrespondents rate target objective using a numeric scale that is anchored with adjectives ex 5star rating format Writing WellWorded Questions The way a question is worded can make tremendous difference in how people answer Crucial that each question is clearstraightforward Leading questionljlquestion that promptsencourages the desired answer through framing certain words ex quotWhat did you think of that amazing concertquot Doublebarreled questionsljasks two questions at once have poor construct validity bc people might be responding to the rst half second half or both and you don t know which item is being measured Negatively worded questionsljlcan make survey items unnecessarily complicated and reduce construct validity of survey Ex quotWas that not the best fourstring quartet you have ever heardquot Goal simplicity even one negative can increase complexity 0 Question orderl Preceding questions can frame subsequent interpretations and encourage particular response patterns on later questions o Every multiitem scale has an order so what should we do Produce multiple versions and then compare results If there is no difference between versions then we can have greater con dence in the results If there is a difference between versions then we should report both outcomes Try to identify a set of questions that does not frame the interpretation of later questions Alternatively nd questions that are not framed inappropriately by earlier questions If there is one quotmostimportant question on your survey you may wish to ask that question rst Encouraging Accurate Responses o People may give inaccurate answers bc they don t make an effort to think about each question they want to look good or they are simply unable to report accurately about their own motivationsmemories 0 Response set a shortcut approach to answering related survey questions Rather than reporting sincere thoughtful and questionspeci c responses a person may develop a pattern for responding o Acquiescence Yeasaying this occurs when people say yes or quotstrongly agreequot to every item instead of thinking carefully about each one Affects construct validity bc instead of measuring true feedback the survey could be measuring tendency to agree or lack of motivation to thinkrate carefully Try reversewording some questions Be careful about double negatives 0 Fence sitting playing it safe by answering in the middle of the scale especially when survey items are controversial Can weaken construct validity bc suggests that some responders do not have an opinion when they actually do One way to avoid this is to remove neutralmiddle responses only drawback is that sometimes people really do have no opinion or answer o Socially desirable responding faking good and faking bad giving survey responses that make one appear better or worse than is actually the case Researchers can avoid this by Guaranteeing that responses are anonymous Including items to identify socially desirable responding and then discard this data if they suspect someone is exaggerating or not paying close attention Using implicit measures Implicit Association Test 0 Saying more than we can know People will report opinions and the reasons on which those opinions are based even though they re effectiver making it up as they go along 0 Ex For identical nylon stockings people generally preferred the rightmost one and claimed their preference was based on quality 0 There may be many empirical forces in uencing our behavior that we are unaware of 0 When asked we attribute reasons to our actions and these inferred reasons might be biasedwrong o Recalling Information we often don t recall our past behaviors accurately I Fashbub memories Vivid memories of major events Con dence is poorly calibrated with accuracy Retrospective selfreports may be inaccuratehave poor construct validity as measure of past behaviors Solutionobserve directly rather than rely on self reports Observational research when a researcher watches people or animals and records how they behave or what they are doing 0 Many scientists believe observing behavior is better than selfreports bc people cannot always accurately report on their behaviorpast events 0 Examples observing how many cars stop for pedestrians at crosswalk listen to comments of parents watching hockey game recording how much people eat in fast food restaurants 0 Researchers can use direct means such as sitting in the stands during a hockey game or by using technology such as an EAR or video camera o Observer bias occurs when observers expectations in uence their interpretation of the participants behaviors or the outcome of the study Rather than rating behaviors objectively researchers might rate behaviors according to their expectationshypotheses Observers can see What they expect to see 0 Reactivity participants may act differently when they know they are being observers may change their behavior Observer effects expectancy effects participants behavior changes to match the observer s expectations 0 How to PreventDiminish Observer BiasObserver Effects 0 Masked blind design Observers are unaware of the conditions to which participants have been assigned and are unaware of what the study is about 0 Use multiple coders allows you to assess interrater reliability 0 Use cleardetailed codebooks allows observers to make accurate records mediates disagreeing observations and allows others to assess quality of operationalizations o How to Prevent Reactivitv o Blend in Unobtrusive behavior make yourself less noticeable one way mirrors 0 Wait it out allow acclimation to one s presence allow people to adjust to your presence and forget that you are there ex Jane Goodall spent several months letting chimps get used to her before they started behaving normally around her 0 Measure the results of the behavior traces behavior leaves behindunobtrusive data indirect measures of constructs ex smudges on windows wearandtear of ooring empty liquor bottles 0 Always observe ethically 0 Must obtain permission in advance to watchrecord people s private behavior Generalizabilitv Does the Sample Represent the Pooulation 0 External validityljrefers to our ability to generalize our ndings of a particular study to a larger populations of interest 0 Concerns both samples and settings 0 Populationljentire set of peopleproducts in which you are interested o Sampleljlactually observed group taken from that population that you can describe directly 0 Use to draw inferences about the population 0 Censusljlcollecting information about the entire population 0 Population ofinterestpopulation to which the researchers want to generalize information 0 Just bc a sample comes from a population does not mean it is representative ofthat population 0 Biased samplelunrepresentative sample some members of the population of interest have a much higher probability of being included in the sample compared to other members 0 Differs systematically from population you d like to represent 0 Convenience samplingusing a sample of people who are readily available to participate o Selfselectionsample contains only people who have volunteered to participate o Purposive samplingrecruiting and studying certain types of participants ie participants w speci c qualities in a nonrandom manner 0 Snowball samplingexisting participants recommend or recruit other participants 0 Quota samplingstrategy in which researchers identify subsets of the pop of interest and then recruit a certain number of participants from each subset in a nonrandom manner 0 Representative samplelunbiased all members of a population have an equal chance of being included in the sample 0 Probability random samplingevery member of population of interest has an equal chance of being selected for the sample using random selection process 0 Simple random samplingchoosing a certain number of participants from the population of interest through random selection most basic version of probability sampling a Cluster samplingclusters of participants within a population of interest are randomly selected and then all individuals in each selected cluster are used 0 Multistage samplingtechnique in which random clusters are chosen and then participants are randomly chosen from those clusters 0 Strati ed random samplingresearcher chooses different subgroups that comprise the population of interest and then randomly selects people from each of the groups 0 Oversamplingresea rcher intentionally overrepresents one or more groups 0 Systematic samplingprocess in which researcher randomly chooses a starting point and then includes every nth person nrandomly chosen number 0 Random sampling vs Random Assignment 0 Random sampling researchers create a sample using some random method such as drawing names from a hat or using a randomdigit phone dialer so that each member of the population has an equal chance of being in the sample Enhances EXTERNAL VALIDITY 0 Random assignment only used in experimental designer when we randomly assign people to groupsconditions so that each group will be as similar as possible on all nonmanipulated variables Enhances INTERNAL VALIDITY 0 External Validity o ESSENTIAL for frequency claims 0 May not be priority for associationcausal claims 0 Larger Samples alone are not sufficient for external validity 0 External validity deals with how participants were recruited not how many 0 Sample size deals with statistical validity Covariance and correlation represents the degree to which two variables vary together Bivariate correlation association an association that involves exactly two variables 0 Variables may be continuous such as age in years or categorical o Scatterplots are good for representing bivariate correlations two continuous variables 0 We can use a correlation coefficient called rfor continuous variables to examine the strength of the association 0 Analysis of bivariate correlations only looks at 2 associations at a time Cohen s guidelines for estimating the strengtheffect size of r o rhas 2 qualities direction and strength how close ris to 1 or 1 O 10 smallweak 30 mediummoderate 50 largestrong When interrogating association Claims the two most important validities to assess are CONSTRUCT VALIDITY amp STATISTICAL VALIDITY 0 Construct Validity How well was each of the 2 variables measured operationalizations What kind of measure was used for each variable Does measure have good reliabilitymeasures what it was intended to Statistical Validity how well do the data support the conclusion Questions about the factors that might have affected the scatterplot correlation coefficient r bar graph or difference score that led to your association claim Effect size describes the strength of an association larger magnitude of rbigger effect sizestronger association 0 0 Larger effect sizes give more accurate predictionsare usually more important However depending on context small effect size can be important ex extreme outcomes Statistical signi cance is the correlation statistically signi cant How likely is it that this correlation size would be reached just by chance 0 O O O p value smaller p value means more statistically signi cant ex plt05statistically signi cantunlikely to have come from quotzeroassociation population while higher p means that the result is quotnot statistically signi cant p value does NOT tell you how big the association actually IS Larger effect size usually means more statistically signi cant but we have to look at p values associated w it Larger sample size improves our chances of nding statistically signi cant association Outliers O 0 An outlier is an extreme score that stands out far away from the pack Can have a strong effect on r Restrictions of range 0 In a correlational study if there is not a full range of scores on one of the variables in the association it can make the correlation appear smaller than it really is Curvilinear relationships 0 Curvilinear association relationship bt 2 variables that is not a straight line 0 Nonlinear relationships are poorly described by r bc r only measures strength of linear relationships 0 Not relevant for association claims but we always have to guard against causal temptation the powerful tendency to make a causal inference from any association claim we read 0 Correlation is NOT causation o The Three Causal Criteria 1 Covariance of cause and effect must be correlationassociation bt variables 2 Temporal precedence causal variable must precede effect variable 3 Internal validity must be no plausible alternative explanations for relationship bt variables 0 Possible third variables can cause spurious association bt variables To whom can the association be generalized Generalization extending a claim beyond the original sample situation or time in which it was observed Size of sample does not matter as much as the wayin which the sample was selected from the population of interest Moderator a quotthird variablequot whose effect on the two focal variables is to change the strengthdirection of their association Multivariate designs involve more than two measured varialbes Ex longitudinal designs multipleregression designs and pattern and parsimony approach Can provide evidence for temporal precedence by measuring the same variables in the same people at several points in time 0 O O Crosssectional correlations Two variables measured at the same point in time Autocorrelations One variable measured at two different points in time correlation with itself Crosslag correlations relationship between an earlier measure of one variable and a later measure of another variable 0 Bene ts helps with third variable problem Criterion variable dependent variableva riable we are most interested in understandingpredicting Predictor variable independent variable 0 By conducting multivariate design researchers can evaluate whether the relationship bt two key variables still holds when they control foranother variable 0 Beta vs b 0 Higher betastronger relationship bt predictor amp criterion variable 0 Smaller betaweaker relationship 0 quotbquotUnstandardized slopes that keeps measurement in its original scale rather than using standard deviation scale 0 Interpreting regression equations and using them for predictions 0 Regression equation Yab1X 3602733X 0 Least Squares regression line minimizes squared distance bt each predicted value of Y and each observed value of Y 0 Does regression establish causation 0 May not be able to establish temporal precedence 0 You can only control for variables that you thought to measure internal validity Multiple regression can often be detected by looking for phrases such as quotcontrolled for taking into account correcting for adjusting forquot 0 Why might we rely on correlational research rather than conducting experiments 0 Sometimes experiments are not feasible 0 Sometimes unethical or dif cult to use random assignment ex smokers Pattern and parsimony for correlational research 0 Parsimonyljldegree to which a good scienti c theory provides the simplest explanation of some phenomenon Ex all ve statements can be explained by single principle smoking as cause of cancer 0 Pattern lis there a consistent association bt the variables 0 Mediator intermediate variable that helps to explain why or how two other variables are associated Mediator variables are internal to the causal variable and often of direct interest to the researchers rather than a nuisance 0 Third variable external to bivariate correlation in question two variables are thought to be correlated but only bc of their relationship with the third variable Moderating variables when for whom or under what conditions are two variables related Mediation moderation and third variables mediation Definition Why are two variables related Example is related to because Esampte Becoming desensitized to viiollenoe i View ting i violent 1 v Sentence Level of desensitization mediates the relationship between Tv violence and aggressive behavior Figure 913 Viewing r 35 violent TV i i39 Viewing violent TV Aggressive v I behavior 3 Moderation When tor whom or under what conditions are two variables related y is related to far type of but not for the other type of 3 EHE Gender moderates the relationship between W violence and aggressive behavior Egg res s i ve behavior r 1IEI Aggressive behavior Thirdvariiable Problem Two variables are eorrelletedi but only because they are both limited to a third variable El is related to but on I because e a F iewiing V violent TV Having quot lienriienlt plrents Aggressive behavior The relationship between viewing violent TV and aggressive behavior may r be attributable to the third variable of parental Ienienov Eopyright it EM 5 W W Norton 3 Company llntt
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'