Fall 1999 - Psych 3311 (Experimental Psychology) - Dr. Burton

There will be six quizzes. The quizzes will not take up the entire class period. The best five scores (including the score on the third quiz) will be entered into the grading formula, below, meaning that there is one drop quiz. Thus, there will be NO make-up tests in this course.

The final exam will be cumulative, and will count for twice the value of a quiz. On the final, you will definitely be required to a) critique a proposed research project, identifying flaws and confounds, and b) perform a two-way between-subject Analysis of Variance, using the method discussed in this class, without notes. Other sections will also be included - examples of previous years' finals are on reserve in the library.

Homework will be assigned almost every class including quiz classes, usually to be turned in the next class. Each homework is worth up to five points, minus one for each class session that it is late. The BEST twenty of the other homeworks will be added for your homework score; count on being able to skip (or drop) 3 assignments. Home-works should ordinarily be typed or word-processed; an exception is made for statistical homeworks. Illegible homework will not be graded.

One major homework will be the conduct of an experiment on November 17. Students will work in teams of two or three, and the other students in the class will serve as subjects. The experiments will be designed by the team based on a review of the recent literature on a topic of the students' choosing. More detail is provided on p. 3. The final product will be an APA-style research report to be submitted on December 6.

Finally, there is a score for class participation. Each student is expected to be ready to discuss assigned readings. As I will explain, each student will also be required to serve as "discussion leader" at least twice. Finally, teams will present their results to the class on one of the last days of the semester, and both team members must share in the presentation. Each student must ask at least one substantive question of his or her classmates during these presentations.

Thus, your final grade is the combination of nine items: Quiz 3 + 4 quizzes + 2*(final exam) + homeworks*1.5 + APA paper + participation/2.

There is no specific grade for attendance, but prompt attendance at each class is highly recommended. The quizzes tend to be fairly comprehensive, and each class that is missed is equivalent to a quarter of the material for the quiz on which you have less chance of success. In my experience, class attendance is a very good predictor of the final grade in the course.

There are three required texts:

  • Martin, D. W. (1999). Doing Psychology experiments. Stamford, CT: Wadsworth. 5th edition.
  • Gilovich, T. (1991). How we know what isn't so. New York: Free Press.
  • American Psychological Association (1994). Publication Manual of the American Psychological Association (4th ed.). Washington, DC: APA.

    Many of the class discussions will center around chapters in Gilovich, so please read this text promptly. Some of your homeworks will be to turn in a typed sheet with three questions you have about the assigned chapter. The questions should be of a challenging nature, such as objections you have to Gilovich's argument or new applications; they are not meant to be suggested essay questions; see the example on p. 4. Only students who have turned in their questions will be eligible to serve as spokesperson that morning.

    Course Objectives: By the end of this class, the student should be able to:

  • Develop a scientific view of how behavior can be investigated,
  • Analyze the structure of experimental designs (isolation of independent and dependent variable, between vs within-subject manipulations, etc.),
  • Understand the place and limitations of nonexperimental methodologies in psychological research,
  • Be familiar with the important organizations relevant to the psychological researcher (APA, APS, Regional Associations, NSF, etc.),
  • Appreciate the psychological literature and carry out significant library research using the PsychLIT database and other resources,
  • Understand, criticize and use APA format for research reports,
  • Analyze and criticize the design of experiments, with thorough appreciation of potential threats to internal validity, and recognition of alternative explanations,
  • Understand experimental control procedures for purposes of extraneous variable control,
  • Perform and apply inferential statistics, both manually and with computer assistance,
  • Design a valid basic experiment, and have first-hand experience in the conduct of experimentation, both as experimenter and as subject.

    The Experiment:

    You will participate in three or four experiments this semester. The first few will be for practice, but the last will be one designed by your team. You will carry out all of the steps of researching, designing, piloting, conducting, analyzing, and reporting this experiment. Although your idea will have to be authorized, you may conduct your experiment on almost any topic, with two restrictions.
  • Priority: Only one team may conduct an experiment in each general area. First come, first served.
  • Practicality: Your experiment must be of a scale that we could actually carry out in a two-hour session on relatively sophisticated human subjects. Thus, helping behavior is probably out. It might also be difficult to conduct an experiment on male-female differences or lefty-righty differences, since males and lefties might be underrepresented in our class. Some more viable ideas might be:
    	Illusions				Short-term Memory
    	Learning a concept or skill		Motor skill
    	Eyewitness memory			Mental imagery
    	Divided attention			Problem solving
    	Gambling				Feedback
    	Cooperation				Dominant hand-vs-Nondominant hand		
    	Perceptual acuity			Mathematical ability			

    1. Your first step is to decide on your teams and reserve your general topic for your experiment; do this by Class 3. Start now.

    2. Second, you and your teammate(s) will find a "keystone" article in the library that represents the state of the art in that area. This article must be recent (no more than four years old) and in a peer-reviewed journal (not a magazine). It must describe at least one original experiment (not a review or essay). It cannot be from Perceptual & Motor Skills, Psychological Reports, or Bulletin of the Psychonomic Society, which are not peer-reviewed journals. Each teammate must have his or her own photocopy of the entire article, which you must know thoroughly. On Class 6>, your team will give me a photocopy of the first page of this article, with a critique (details later).

    3. Next, you and your teammates will thoroughly explore the literature on your chosen topic, using PsychLIT and the reference list and introduction of your own article. Your team must identify three articles per team member plus three more, including the keystone--for each, how did they vary the basic procedure in your keystone article? What did they find? Each teammate will read thoroughly and summarize three different articles from this list. You will turn in an annotated bibliography (including photocopies of the front pages of the ones read first-hand) based on this research by Class 10.

    4. While you carry out this research, you will decide what sort of variation on this theme has not been tried yet, and why it might be interesting. Obviously, it must be a variation that you could achieve in our laboratory on relatively sophisticated human subjects. You and your team will then design in detail your variation of the experiment. It should be an experiment with a single true (not organismic) Independent Variable with two or three levels. You will submit a formal design proposal, in a format to be explained in depth later, on Class 14.

    5. Before Class 24, you and your teammate will conduct a pilot experiment with me as the subject. All equipment, stimuli, data sheets, instruction sheets, etc., should be ready at this time.

    6. Based on your design proposal and my comments on it, you will turn in (working individually) a justification for your experiment in the form of an APA-style Introduction section (a Reference List must also be included). You will use the keystone article and your three of the articles, and any others you are interested in, except for the articles that belong to your teammate. This will be turned in around Class 21.

    7. Class 24 is Experiment Day. You and your team will carefully conduct twelve of your classmates through your experiment. You will take turns conducting the study and serving as your classmates' subjects. ATTENDANCE ON EXPERIMENT DAY IS MANDATORY.

    8. You and your team will immediately organize the data and each turn in a graph of the means of your conditions for the next class. Every other assignment from this point on is to be done individually. For the class after that, you will turn in two versions of the analysis: One done by hand, showing all calculations, and one done in SYSTAT. Obviously, your results must agree.

    9. On a date to be determined, you and your team will jointly present a brief (about 10 minutes) report to the class, focusing on the general idea of the experiment, your results, and your conclusions. They will have the opportunity to ask some questions about your experiment, and likewise you will be able to ask questions of them when they present. Attendance on this class is also mandatory.

    10. Finally, in the week before the Final, you will turn in your own APA-format research report about your own experiment.


    EXAMPLE GILOVICH QUESTION

    On p. 26, Gilovich gives examples (such as the so-called "Sports Illustrated Jinx") of situations where regression to the mean is misinterpreted. Since a medium performance is always most likely, a good performance is likely to be followed by medium performance (which will look like a decline) and a poor performance is also likely to be followed by medium performance (which will look like improvement). In my opinion, what his examples have in common is that people interpret these situations as involving motivation or morale rather than chance (e.g., once he or she is on the cover of Sports Illustrated, the athlete feels so much pressure that he or she becomes distracted.) My question is, does Gilovich feel that morale has no effect on behavior at all? What about overconfidence effects, for example? Are these also illusions?


    SYLLABUS

  • Part 1: Three elements of experimentation. [DWM ch. 1, 13, TG 1-2, APA 1]
    Comparison
    Operational definition
    Control
  • Part 2: Initiating the research. [DWM 2, 3, 4, 11, TG 3-4]
    Hypotheses
    Literature Review
    Ethical constraints on experimentation
  • Part 3: Constraints on interpretation [DWM 6, 10, 12, A, TG 5-7]
    Statistical review
    The Selection confound
    Other confounds
    The varieties of validity
  • Part 4: Technical issues [DWM 13, APA 2]
    APA format revisited
    Analysis of Variance
    Review of computer applications
    Experimental paperwork
  • Part 5: Advanced designs. [DWM 7, 8, 9, 12]
    Experimental designs
    Counterbalancing
    Interactions
    Two-way Analysis of Variance
  • Part 6: Philosophy of Experimentation. [DWM 4, 10, TG 8, 10, 11, APA 6]
    Types of error
    Conflicts between ethics and interpretation
    Alternatives to experimentation