01. Social Psychology: Introduction and MethodsThe Scientist-Practitioner ModelOutline1. The value of social psychological research1. The value of social psychological research2. How social psychologists test ideas2. Observational research2. Archival research2. SurveysTest your knowledge!2. Correlational researchSlide 122. Correlational research2. Experiments2. Experiments2. Other distinctions3. Evaluating study quality3. Experiments and validity3. Other concepts for understanding research3. Replication4. Ethical concernsSummary: Methods02. The Social Self01. Social Psychology: Introduction and Methods1The Scientist-Practitioner Model•Stats classes, methods components, reading original articles, writing research papers, participating in research, doing your own studies, …•Why?•Understand others’ claims•Verify others’ claims•Criticize others’ claims→ have you not be dependent on other people’s interpretations2ConclusionConclusionStudy Analysis InterpretationOutline1. The value of social psychological research2. How social psychologists test ideas3. More concepts for understanding research4. Ethical concerns in social psychological research31. The value of social psychological researchTwo groups (Schwarz et al., 1991):•Group A: Report 6 instances of assertive behavior•Group B: Report 12 instances of assertive behaviorWhich group later reports the highest level of assertiveness?A. Group AB. Group BC. Both groups report similar levels of assertiveness41. The value of social psychological research1. Hindsight bias2. Many unexpected and unobservable things influence behavior3. Quantification and effect sizes → how much is it so?4. Behavior is multiply determined52. How social psychologists test ideas•Hypothesis = prediction about what will happen under particular circumstances•Theory = a body of related propositions to describe something1. Observational research2. Archival research3. Surveys4. Correlational research5. Experiments62. Observational research•Observation•Participant observation (cf anthropology)•Social psychologists typically add quantitative data and formal studies72. Archival research•Census reports, police records, sports statistics, newspapers, hospital records, …•e.g., more insult-related homicides in the Southern US (Nisbett & Cohen, 1996)•e.g, predicting gonorrhea and other STIsfrom tweets (Chan et al., manuscript under submission)82. Surveys•Interviews•Questionnaires•Entire population•Sampling•random•Convenience•Example of biased sample – 1936 presidential election•Literary Digest predicted landslide win of Alf Landon over Franklin D. Roosevelt•Selected participants from telephone directories, automobile registrations9Test your knowledge!•What is NOT a basic tenet of social psychology?A. The situation can have strong influences on our behavior without us being aware of itB. There is little variation in the most important psychological processes across culturesC. Evolutionary processes have shaped many of our psychological processesD. We usually react to how we think the world looks like, not how it actually looks like102. Correlational research•Are X and Y related?•Variables usually just measured, not manipulated•correlation ≠ causation1112CC-BY-4.0http://tylervigen.com/discover2. Correlational research•Spurious correlations = a third variable is causing the relationship•E.g., self-selection: “Baby swimming will increase your baby’s IQ!”•Reverse causation / bidirectional relations•E.g., “Study finds that happy people help others more!”•E.g., “Watching violent TV makes kids more aggressive!”132. Experiments14random assignmenttreatment groupcontrol groupindependent variable (IV)outcome 1outcome 2dependent variable (DV)manipulate observe2. Experiments+Keeping many factors constant → more precision on what exactly led to result+No self-selection effects; differences should, on average, balance out+Establish temporal order+Much better evidence for causality-Some things can’t or shouldn’t be manipulated-Sometimes things should be observed as they naturally occur152. Other distinctions•Lab vs. field research•Cross-sectional vs. longitudinal research•Basic vs. applied research161. Objectivity – Do you get the same result when someone else measures it?2. Reliability – Do you get the same result when you measure multiple times?3. Validity – Are you measuring what you want to measure?1. Internal validity – Did the IV really cause the what you’re measuring?2. Measurement validity – Is the measure related to an outcome?3. External validity – Is the result generalizable?1. Objectivity2. Reliability3. Validity3. Evaluating study quality17reliable and valid reliable but not validunreliable3. Experiments and validityGet together with your neighbor. One of you does the task on the left, the other the task on the right!(briefly introduce yourselves if you don’t know each other yet)•Explain what an experiment is and what its crucial characteristics are, to check if you understood the concept•Explain why experiments usually have better internal validity than correlational studies, to check if you understood the concept183. Other concepts for understanding research•Regression to the mean – all things must come down•Unusual values often produced by unusual circumstances or an unusual combination of factors that are unlikely to occur again•Kahneman & Tversky (1973): misleading perceptions about the effectiveness of praise and punishment•The sequel is worse than the original•Statistical significance•p-value: probability to obtain this result or a more extreme one if the null hypothesis is true•Significance usually defined as p < .05•Depends on effect size and sample size193. Replication•Direct replication•Conceptual replication•Problems•One single study cannot give an ultimate answer•One single replication cannot give an ultimate answer204. Ethical concerns•Institutional review boards (IRBs)•Risks vs benefits•Informed consent•vs deception research•Debriefing21Summary: Methods221. Objectivity2. Reliability3. ValidityInternal validityExternal validityRegression to the meanSignificanceReplicationEthical research02. The Social Self23NEXT