EXPERIMENTAL DESIGN 3021

Fall 2003

ROBERT L. LLOYD, PH.D.

 

 

Office Hours:  11:00-12:00 M, W, & F and by appointment.  330 Bohannon Hall

Tel. 726-6799; E-Mail rlloyd

Text: Text:  Experimental Psychology:  Understanding Psychological Research, Seventh Edition,

Kantowitz, B. H., Roediger, H. L., and Elmes, D. G., West Publishing.

 

 

Part I:  Review of Statistics

Characterization of a data set

            Measures of Central Tendency

            Measures of Heterogeneity

            Transformed Scores

 

Statistical inference

            Estimating Population Parameters

            Sampling Distributions

            Expectations of the Mean

                        Standard Error of the Mean

                        z-test

                        One-Sample t-test

 

Non-parametric tests

Two-group parametric analysis

Multi-group, Factorial, and Repeated Measures Analysis

Relationship between two continuous variables

 

 

 

Part II:  Research Proposals, Reports, and Hypothesis Testing

            Experimental vs. Quazi-experimental Designs

            Theory vs. Hypothesis

            Construct vs. Operational Definition

            Qualitative and Quantitative variables – Scales (Levels) of measure

Test Selection

            Chi-Square

            t-tests and Analysis of Variance

            Correlation

           

 

            Random Selection & Random Assignment

            Errors of Inference

            Statistical Power

            Experimental Control

Minimizing Within-Group Variance

Counterbalancing

Secondary Variance

           

           

Hypotheses and data variability

Data acquisition

Data reduction and descriptive statistics

One-group designs

Two-group designs

 

 

Part III:  Complex Designs and Interpretation

            Factorial Designs and Interactions

            Threats to Internal Validity

                        Secondary Variance

            External Validity

            Restriction of Range

            Floor and Ceiling Effects

Multiple-measurement designs

Regression and Discriminant Function Analysis

                        Sensitivity and Specificity

            MANOVA

Alternative Methods of Correlation

Factor Analysis

 

 

Exams:

Midterm I:                                            Friday, Oct 3rd

Midterm II:                                           Friday, Nov7th

Final Exam (non-cumulative):            Wednesday, Dec 17th, 8:00-9:55 am

 

 

Other Deadlines:

Friday, End of Second Week:            Tentative Research Proposal (Just the Topic)

Friday, End of Fourth Week:            Structure of Research Proposal (IV, DV, Instruments, etc.)

Friday, End of Fourth Week:  Test Critiques write-up

Friday, End of Eighth Week:  IRB Application

 

 

Grading:

            Your final grade will be based upon the two exams and a final (non-cumulative), one in-class report, one analysis of US Census data, plus a final paper.  The three exams will be weighted equally, the in-class presentation is worth one half of an exam, and the Census Data and Final Paper will be worth one half of an exam.  The format of each exam is short answer.  You will have the option of increasing your grade on each exam by taking a multiple-choice exam which covers the same material.  These multiple-choice exams are taken on TestPilot, a web-based program.  If you score 90% on the multiple-choice exam, your grade goes up one notch (e.g., from a C+ to a B-).  If you score 95%, your grade goes up two notches (e.g., from a C+ to a B).  You can take these exams as many times as you want; your score can go up, but it can never go down.  If you get 90%, you can go for 95%.  To be eligible to take the multiple-choice exam, you must make your first attempt within seven days of the exam.  Try to do well on the written exam, because this will determine the base from which you can elevate your grade.  A schedule will be made up with times to take the multiple-choice test (In Cina 22 or in the Psychology Office).

 

            For each of the exams, you will be required to submit three multiple-choice questions on the material which is on the exam (This includes the in-class presentation material).  You will submit these questions to a Web Crossings Page, under my name, over the internet.  Go to http://www.d.umn.edu/webx.  Click on my name and then click on the folder that says Experimental Design 3021 Fall 2003.  Then go to the sub-folder with the test number (e.g., Exam I).  You must check to see that your questions are unique from those already submitted by other class members.

 

  The final paper will be a formal research proposal using the format described in your text.  This proposal should describe a practical experiment which you should be able to carry out.  This proposal should form the basis for an experiment which you will, in fact, carry out in your next class in experimental design.

 

            Your proposal should begin with an introduction which reviews what others in the literature have found concerning this topic, and ends with a formal hypothesis.  Next, there should be a methods section which describes the population to be sampled, the procedures to be followed, and the statistical analysis to be used.  The next section will be the results section in which you use made-up data, which you expect to find when you ultimately do conduct your experiment, to compute statistical assessments of your experiment.  The last section of your paper will be the discussion section, in which you explain what your results mean in relation to what has already been reported in the literature.  Along with your research proposal and mock analysis, you will turn in your statistical analyses of your mock data.  This will consist of your data, computer printout of the statistical analyses, and graphical representations.  Finally, you will turn in the Human Subjects Approval Form which must be submitted to the University before any experimentation upon human subjects may be initiated.  This application is reviewed by an ethics committee on the Twin Cities campus.

 

 

Extra Credit:

You can get extra credit for turning in a scientific poster of your mock research paper (in addition to the required write-up).  You can also get extra credit for presenting your US Census analysis to the class.

 

Cina 22 Lab:

            The laboratory portion of the class will be conducted in Cina 22.  Here you will learn how to use the Statistical Package for the Social Sciences (SPSS), a computer program which allows you to create data files and analyze and graph your data.  This is the program which you will use to analyze your mock data (and your actual data in Experimental Design II).  In addition, you will be using the MEL Lab package, which contains computer programs for the execution of experiments which we will do in the lab.  This package both runs the experiment and collects the data.

 

            For the lab portion of this class you will need to purchase the MEL Lab package and two additional high-density (1.44 megabyte), PC-compatible three and one quarter inch floppy disks.  Bring your MEL disks and handbook, and your SPSS disk, to every lab meeting.

 

 

 

If you find that the material is not making sense to you, Come and see me.