Study Sheet for PSY 311; Psychological Research Methods; Final Examination; Glenn Geher, instructor

 

The examination will consist of two sections; a multiple choice section (addressing class notes and the textbook chapters covered in the syllabus) and an essay section.  A list of concepts that will be stressed on the multiple choice section is included here.  Also included are essay questions corresponding to several important themes from class.  Of these essays, four will be on the exam.  You will be asked to complete three of those four.

 

--------------------------------------------------------------------------------------------------------------------------------

While all material from class notes and the textbook may be represented on the exam, you may want to focus your studying on the following concepts:

 

-- Definition of science in the context of psychology (e.g., general rules, objective evidence, verifiable, skeptical, open-minded, creative, public, productive)

 

-- Kuhn’s notions regarding paradigms in science (know the different stages and know the role that psychological research plays)

 

-- Operational definitions

 

-- Internal validity, construct validity, external validity

 

-- independent versus dependent variables

 

-- issues surrounding ethics in psychology

 

-- the distinction between “mediator” and “moderator” variables

 

-- defining features of different kinds of measurement in psychology (e.g., self-report, nonverbal behavior, etc.)

 

-- sources of error due to: observer bias, administration of measures, participant error, reactivity, experiment expectancy effects; what are these issues and how can they be addressed?

 

-- ways to assess reliability of measures

 

-- the defining features of different kinds of validity (e.g., convergent validity)

 

-- the basic logic of establishing internal validity

 

-- advantages and disadvantages of within versus between subjects designs

 

-- Campbell and Stanley’s (1963) 8 kinds of threats to internal validity

(selection, selection by maturation interactions, regression, maturation, history, testing, instrumentation, & mortality)

 

-- Problems with different kinds of assignment to conditions

 

-- The features that distinguish experimental research from correlational research

 

-- General reasoning of hypothesis testing

 

-- Type 1 and Type 2 error

 

 

--  What statistical significance really means

 

-- the relationship between effect size and statistical significance

 

-- Reasons to conduct multigroup experiments

 

-- The general reasoning of ‘post-hoc’ tests

 

-- Statistical Power

 

Essays: 

 

1.  Briefly describe what ‘validity’ means with regard to scientific research in psychology.  Next, define ‘construct,’ ‘internal,’ and ‘external’ validity.  Next, briefly describe an experiment that one could conduct which tests whether eating vitamins increases alertness during daytime hours.  In your description, clearly address how you would operationally define your independent and dependent variables.  Further, be sure to explicitly address how you design the study to increase construct, internal, and external validity.

 

2.  Describe the conceptual difference between ‘mediator’ and ‘moderator’ variables.  Next, give a brief, hypothetical conceptual example of a mediator variable.  Finally, give a brief, hypothetical conceptual example of a moderator variable.  Make sure that these examples are ones you make up yourself!

 

3.  Suppose you wanted to design a study to establish the reliability and validity of a new measure of social competence.  Describe the different steps you would take in implementing this research.  Be sure to outline how you would address each of the following: internal reliability, test-retest reliability, discriminant validity, criterion validity, and predictive validity.

 

4.  Briefly describe four of the threats to internal validity that Campbell and Stanley (1963) outlined.  For each threat you describe, outline a hypothetical example of a study in which this threat is clearly present.

 

5.  Explain the differences between Type I and Type II error.  In your explanation, be sure to address what is meant by the “real state of affairs” regarding the treatment effect, and the statistical significance decision.  Additionally, provide one hypothetical example of Type I error in addition to one hypothetical example of Type II error.  Finally, describe one method designed to reduce Type I error as well as one method used to reduce Type II error.

 

6.  Suppose that you were going to conduct a study to examine if three methods for teaching statistics differ in terms of how well students using each method understand the material.  In your first group, participants are encouraged to read the textbook but not attend lecture.  In the second group, participants are encouraged to attend the lecture but not read the textbook.  In the third group, participants are not encouraged to read the textbook and they are not encouraged to attend the lecture.  Suppose you had 20 participants in each group.  Further, suppose that scores on the DV were represented by scores on the final exam, varying from 0-100.  Explain how you would analyze data from this study using a One-way ANOVA.  In your explanation, address the concepts of “between groups variability,” “within group variability,” and the “F ratio.”  Further, describe why you would use a post-hoc test such as the Tukey HSD; also, explain the function that this post-hoc test would serve.

 


7.  Suppose I conducted research to examine if grades are positively related to life satisfaction among college students.  To implement this study, I ask 100 undergraduate students to report their GPA’s and to complete the GLSS (Glenn’s Life Satisfaction Survey).  I find a Pearson correlation of .9.  I further find that the “p value” associated with this correlation is .0002.  Based on this finding, explain the implications in terms of the following: (a) the strength of this relationship, (b) the direction of the relationship, (c) the r2 associated with this relationship (including how to interpret that r2), (d) and the meaning of this p value in the context of traditional alpha levels (e.g., .05).  Finally, (e) indicate what implications I can and cannot draw regarding the possibility of a causal relationship between these two variables.

 

8. Research reports typically include the following sections: (a) an abstract, (b) an introduction, (c) a “Method” section, (d) a “Results” section, and (e) a “Discussion.”  Suppose that your student Frankie just handed in a research report on an experimental study he conducted in which he examined whether people who were randomly assigned to an intense exercise condition differed from subjects in a control group on the dependent variable of happiness (as measured by the SUNY Happiness Survey).  Briefly write out what specific elements you would expect to find in the aforementioned five sections of his report if Frankie were going to receive an A on his paper.

 

9.  Describe the general purpose of SPSS.  Next, consider the three kinds of SPSS files that you learned about in this class: data files (.sav), syntax files (.sps), and output files (.spo).  Briefly describe the function of each of these different files. 

 

Finally, write a sentence or two outlining how much you love SPSS … just kidding about this last point …

 

10.  Megan wanted to conduct a study to examine if her three favorite characters, Elmo, Ernie, and Cookie Monster, differed in terms of how much amusement they bring to other babies.  To examine this question, she randomly assigned five babies to one of three conditions.  In the Elmo condition, babies watched an hour-long video of Elmo (Group 1); in the Ernie condition, condition, babies watched an hour-long video of Ernie (Group 2); in the Cookie Monster condition, babies watched an hour-long video of Cookie (Group 3).  Next, Megan and her mommy carefully administered the “baby amusement” survey to each baby as the DV.  Here are the scores on the amusement variable for all participants (higher scores mean more amusement):

 

Group   Amuse

Group   Amuse

Group   Amuse

1.00      4.00

1.00      3.00

1.00      7.00

1.00      7.00

1.00      2.00

2.00      7.00

2.00      8.00

2.00      9.00

2.00      9.00

2.00      7.00

3.00      2.00

3.00      7.00

3.00      1.00

3.00      4.00

3.00      1.00

 

Here is the .sps file that Megan types up for herself:

 

ONEWAY

  amuse BY group

  /STATISTICS DESCRIPTIVES

  /MISSING ANALYSIS

  /POSTHOC = TUKEY ALPHA(.05).

 


Finally, click on the following for Megan’s dandy results:

 

Descriptive Statistics

 

ANOVA Summary Table

 

Tukey Post-hoc tests

 

 

Using the standard .05 alpha level, write up a summary of Megan’s findings including the implications of the F and the implications of the Tukey HSD results.