MEMORANDUM
To: PS 585 (Intro. to Research Methods) Students
From: Instructor Ray Block, Jr.
Date: Fall, 2003
Re: Homework Assignment #5 (Mulitvariate Data Analysis)
Political Knowledge: Where Education meets Political Science
Last assignment, I asked you to calculate central tendency and variability
measures for the political knowledge scores that students received in this
class. We will build on the previous assignment here.
I ran a little experiment on you guys when I made the questionnaire.
In that experiment, I manipulated the political knowledge questions in
interesting ways. If you recall, I asked students 5 questions about
American politics and I took the number of correct answers (say 5 correct
out of 5 questions) to indicate a students’ level of political knowledge.
I wanted to see whether a student’s political knowledge score depended
on how the questions were formatted.
There is some research out there suggesting that some students do better
on multiple choice tests and others do better on open-ended test.
If you think of the political knowledge questions as being “test of political
knowledge”, then it is safe to assume that the format of this knowledge
test might possibly affect a students’ test performance. Accordingly,
I manipulated the questionnaires so that some students got open-ended political
knowledge questions and others got or multiple choice questions.
A breakdown of the experiment would look like this:
-
Dependent Variable: Political Knowledge Score (on a scale
of 0 to 5)
-
Independent Variable: Question Format
-
Group 1: Students who got the Multiple-Choice questions
-
Group 2: Students who got the Open-Ended questions
I recorded the results from this experiment in an SPSS data file.
You can obtain this data file by going to lecture notes section of my Web
Page and clicking the “Assignment 5 Data” link or by clicking the link
below:
Your Assignment: I want you to take this data file and run
an Analysis of Variance (ANOVA). To do so, I want you to follow these
steps:
-
State the null and alternative hypotheses in plain English
-
State the null and alternative hypotheses using notation
-
Compute the ANOVA in SPSS
-
Open the data in SPSS
-
In the main menu, click Analyze è Compare Means è One-way
ANOVA
-
Place the variable called “pkscore” in the box labeled “Dependent
List”
-
Place the variable called “format” in the box labeled “Factor”
-
Click the “Options” button and select “Descriptive” from the box labeled
“Statistics”
-
Interpret the results:
-
What are the means knowledge scores and standard deviations for each group?
-
What is the F-statistic?
-
Is the F-Statistic you computed bigger or smaller than 4.35 (this is what
the critical F-value would be if you did a significance test at the .05
level with 24 degrees of freedom)?
-
Can you reject the null hypothesis?
-
Make a conclusion:
-
Did format make a significant difference in how well students answered
the political knowledge questions?