PSYCH 136A_Midterm Review Presentation .pdf

PSYCH 136A_Midterm Review Presentation .pdf - 136a Midterm...

  • No School
  • AA 1
  • 26

This preview shows page 1 out of 26 pages.

Unformatted text preview: 136a Midterm Review Patrick Rock Feb 7/8, 2015 Someone remind me to record this review!! General Remarks •  If I’m not recording, someone tell me to start •  Will go through by topic, not chapter or lecture (will idenMfy which chapters/lectures fall under each topic). Will take quesMons at end of each topic (every few slides) •  Do need to know –  Concepts from reading/lecture –  General importance of studies emphasized in reading (how it demonstrates an important concept) •  Don’t need to know –  Specific details from studies (e.g., how much shock was administered) –  Terms or concepts menMoned in passing in the reading (e.g., in a parentheMcal, or an explanaMon, but without a definiMon given or any real discussion of the concept) Today vs. Tomorrow’s Review •  Same material •  Class will ask different quesMons •  Different desserts –  Lemon bars today. If any leYovers, will bring them to Monday’s session –  Bomb-­‐ass-­‐brownies tomorrow (with mints! And walnuts!). If any leYovers, will bring them to class. •  Come to both, if you want, but no need Topic 1: Basic Research Concepts •  Lecture 1 •  Chapters 1 and 2 •  General Terms/Concepts –  Law –  Theory –  Hypothesis –  OperaMonal definiMons 1A: What makes good research •  Goals: Describe, predict (correlaMon), explain (causaMon), change behavior •  A study is important if it –  Challenges exisMng theories –  Integrates exisMng theories –  Challenges intuiMons/expectaMons –  Large effects from small manipulaMons –  Speaks to issues of social importance –  The more of these things it does, the be`er a study it is. 1B: How do people “know” about the world? •  •  •  •  •  IntuiMon (lay people) Authority (religion) ObservaMon (science) Logic (philosophy, math) What approaches have people used to explain behavior other than science –  Metaphysical (religion, animism, astrology, etc.) –  Logic-­‐based (philosophy) –  Biological explanaMons (physiology) 1C: Canons of Science •  Determinism: Events have causes –  Avoid illusory correlations; seek out actual causal relations •  Empiricism: Use observations •  Parsimony: Start with the most simple explanation •  Testability/Falsifiability: Seek out tests that refute your theory •  Common Problems –  ConfirmaMon bias: the tendency to just seek out confirming evidence or just noMce confirming evidence (also called posiMve test bias in the textbook) –  Behavioral confirmaMon (a form of experimenter bias): Tendency to elicit behavior that you expect from a parMcipant 1D: Methods of Developing ScienMfic Knowledge (Chapter 2) •  InducMon (make observaMons, conclude based on them, –  Case studies, surprising incidents, looking at experts’ behavior –  Hard to know with any confidence what to conclude •  DeducMon (construct theories, make specific predicMons, test those predicMons) –  Hypothesis tesMng (validate hypothesis, falsify hypothesis, qualify the boundary condiMons of a hypothesis –  Can’t make absolute statements of a theory being “proven” but can demonstrate confirming evidence more effecMvely than inducMon Topic 2: Different Research Methods •  Lecture 2 (General methods) •  Chapters 6 and 7 (General methods) •  Important Terms –  CorrelaMonal studies •  CorrelaMonal fallacy •  Use random sampling (produces generalizability) –  True experiments •  Use random assignment (produces validity) –  Archival research –  ObservaMonal research (obtrusive, unobtrusive) –  Pseudo-­‐experiments 2A: CorrelaMonal Methods •  Lecture 2, Chapter 6 •  Tell you if two variables co-­‐vary (change together) •  Advantages: Cheap, easy, quick. OYen use surveys –  (Ideally) Random sampling of representaMves from populaMon of interest –  Can use cluster sampling (less random, more feasible) –  Produces sampling error (difference between results found by measuring populaMon sample and “true” results in total populaMon •  Disadvantage: Can’t show causality (confounds) or direcMonality (though longitudinal research helps) •  Want to avoid –  SelecMon bias –  Non response bias 2B: True Experiments Lecture 2, Chapter 7 Have (at least one) IV and (at least one) DV Random assignment to condiMons of IV Advantages: Minimize effect of individual differences, confounds; lots of control over variables; can isolate subtle phenomena; can prove causality and demonstrate interacMons; can minimize random error •  Disadvantages: ArMficiality. Can address by enhancing •  •  •  •  –  Mundane realism (how similar task is to real world task; not always feasible as an opMon) –  Experimental realism (aka psychological realism): How similar psychological experience is to the one in the real world; oYen more feasible than mundane realism) 2C: Quasi-­‐Experiments •  No random assignment to condiMons •  Good in situaMons where manipulaMon is impossible/ difficult or unethical •  Types –  Person by treatment (use pre-­‐screening + extreme groups or median split to idenMfy person variable; manipulate a different variable) –  Natural experiments (natural event disMnguishes people; want to look for control that has comparable individuals not affected by event); can do pre-­‐test post-­‐test comparisons or Mme series data –  Patched up designs (keep adding control groups to study unMl you have eliminated obvious confounds) Topic 3: Experimental Designs •  Lecture 6; Chapter 8 (factorial designs) •  Lecture 7; Chapter 9 (between, within and mixed designs) •  Important terms –  Between vs. within vs. mixed designs –  Main effects and interacMons; simple effects –  Counterbalancing (full, parMal, laMn square) 3A: Different Experimental Designs •  One-­‐way vs. Two-­‐way vs. Three-­‐way = # of independent variables •  Two groups vs. three groups vs. more groups = # of levels in the independent variable •  Factorial designs (most oYen: 2x2) –  Main effects (2) –  InteracMons (1) (see chapter 9) •  Ordinal/spreading •  Disordinal/crossover –  Simple effects tests 3B: Between, Within and Mixed Designs •  Lecture 7; Chapter 9 •  Within subjects: Each parMcipant is given all levels of a given IV –  Pros: reduces random error, easier to detect small changes, fewer subjects needed –  Cons: ContaminaMon effects (pracMce effects, interference effects); order effects (Mme/Mredness affects responding) •  Control for these using counterbalancing –  Complete counterbalancing –  Incomplete counterbalancing •  Reverse counter balancing •  ParMal counter balancing •  LaMn square counter balancing •  Mixed-­‐Model Designs: Have both between and within subject variables Topic 4: Validity (and threats to it) •  Lecture 3; Chapters 3 (validity generally) and 5 (threats to validity) •  Important terms –  Internal validity –  External validity –  Random vs. SystemaMc error –  Common threats to validity (know them, covered on next slide) 4A: Kinds of Validity (Chapter 3) •  For a study: External vs. internal validity (generalizability vs. truth) •  For a measure –  Convergent validity –  Discriminant validity –  Construct validity (how good are the operaMonal definiMons) •  For a hypothesis: conceptual validity (how well does the hypothesis relate to the theory?) •  To be valid, study or measure must be reliable –  Test-­‐retest –  Interobservor/interrater 4B: Common Threats to Validity (chapter 5) •  People change over Mme –  History (fix with control group) –  MaturaMon (fix with control group) –  Regression to the mean (fix with control group, not selecMng based on extreme scores) •  Studying people changes people –  TesMng effects (use different tests, use control group, don’t give a pretest) –  Heterogeneous a`riMon (minimize dropout, check for this) –  ParMcipant reacMon biases (expectancies and reactance; evaluaMon apprehension) (give false hypotheses, use behavioral measures, make non-­‐ obvious hypotheses) –  Hawthorne effect (mere measurement effect) (control group asked same quesMons) •  Confounds and ArMfacts –  Confounds (measure possible ones; do a study to rule them out) –  ArMfacts (something about lab affects results; hurts external validity) –  Experimenter bias (use double blind design; limit experimenter-­‐parMcipant contact) 4C: External Validity Threats •  External validity = generalizability •  Threats –  Homogeneous a`riMon (minimize dropouts, compare those who do to those who don’t dropout on key variables) –  SelecMon bias (non representaMve sample) •  Increase external validity with mundane realism and experimental realism •  Experiments trade off between external and internal validity •  A`riMon: –  Homogeneous threatens external validity –  Heterogeneous threatens internal validity Topic 5: ManipulaMon •  Lecture 4 (no associated reading) •  Important terms –  ManipulaMon check •  4 types of manipulaMons –  Environmental –  SMmulus –  Social –  InstrucMonal •  Can hide manipulaMons •  Compliance: Foot in the door, door in the face, low-­‐ ball, even a penny Topic 6: Measurement of DVs •  Lecture 5; Chapters 3 and 4 •  Know the types of DVs (self-­‐report, behavioral, implicit) •  Know measurement scales (nominal, ordinal, interval, raMo) 6A: Types of DVs •  Types –  Self-­‐report: easy, cheap, anonymous, but not necessarily honest and people might not know true feelings; sensiMve to biases based on answer choices, quesMon order, context within a larger survey –  Behavioral measures (obtrusive, unobtrusive): More subtle, can be more engaging, might be more true to real world scenarios; but harder, same behavior in different people can mean different things. –  Implicit measures: Can get at unconscious artudes and evaluaMons; not influenced by self-­‐presentaMon concerns but not always clear how to interpret; sensiMve to contextual factors; lack convergent validity 6B: Types of Scales •  Nominal (categorical, no meaningful order) •  Ordinal (meaningful order, not sensiMve to magnitude of dierence between points) •  Interval (equal difference in magnitude from one point to next) •  RaMo (true zero point; allow comparison of raMos) 6C: How to create a scale •  Chapter 4 •  Start with pilot tesMng (focus groups, open ended quesMons) •  Create quesMons (use quesMons that communicate clearly what you want to know, so everyone is on the same page; see book for details) •  Provide meaningful responses –  Give good anchors on numerical scales to communicate what each value means Topic 7: StaMsMcs •  Lecture 8 (no associated chapter) •  To know how unusual an observaMon is, need to know –  PopulaMon mean –  PopulaMon variance (measured by standard deviaMon) •  To compare two groups, look at the raMo of the difference between the means and the variance of each group –  If difference between means is big but variance within groups is small, it means the groups are likely different in reality –  If difference between means is small while variance within groups is big, it means the groups are not likely to be different in reality •  People lie with staMsMcs –  Using dishonest scales for graphs –  Emphasizing small differences by adjusMng scale in strategic ways –  Do good stats reporMng: Start scale at 0, show SD or some form of variance Topic 8: Ethics •  Lecture 9; Chapter 2 •  APA Guidelines –  Informed consent –  Freedom from coercison –  ProtecMon from physical and pyschological harm –  Risk-­‐benefit rule (in book, called risk-­‐benefit analysis) –  Debriefing •  IRB evaluates new studies ...
View Full Document

  • Fall '19

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern