EDF
5481 METHODS OF EDUCATIONAL RESEARCH
FALL 2002
ASSIGNMENT TWO:
TRYING OUT OPERATIONAL
AND NULL HYPOTHESES
EXPERIENCE EVALUATING
CAUSALITY AND EXPERIMENTS
|
DUE FRIDAY
SEPTEMBER 20 BY NOON. MY MAILBOX 307 STONE.
PLEASE DO NOT USE EMAIL ATTACHMENTS.
SEE THE
OVERVIEW FOR ALTERNATIVES IF YOU ARE UNABLE TO COME TO THE OFFICE.
ASSIGNMENT TWO is due IN MY MAILBOX
BY
NOON:
FOCI:
This assignment will give you experience
in
-
rewriting your research problem statement
-
constructing operational definitions
-
writing hypotheses and null hypotheses
-
examining a basic experimental design
|
THESE
DIRECTIVES ARE DESIGNED FOR FLEXIBILITY
DEPENDING
ON YOUR RESEARCH PROBLEM
For
example, if you only have one main independent variable, you do not need
to designate any additional independent variables. Space guidelines are
approximate.
|
PART ONE:
You will rewrite your
CONCEPTUAL research
statement taking into account feedback and other information:
All the materials for a rewritten research
problem statement from assignment 1:
-
introductory paragraph with justification
-
identification of key dependent variables
(no more than 3)
-
identification of key independent variables
(no more than 5)
-
TWO DISTINCT hypotheses
that CONCEPTUALLY link
your independent variables to your dependent variables.
At this point, your
rewritten research problem statement should only be about 1.5 double-spaced
pages. This is because you have eliminated
all extraneous material that does not relate to your research problem.
Suggested lengths below refer to double-spaced
print.
PART TWO:
You
will provide:
-
OPERATIONAL DEFINITIONS
of ONE or TWO dependent variables
-
OPERATIONAL DEFINITIONS
of NO MORE THAN THREE independent variables
(About .5 page.)
-
TWO OPERATIONAL
HYPOTHESES, each one linking one operationalized independent variable with
one operationalized dependent variable.
-
Each hypothesis must
be distinct.
-
If one hypothesis is
the opposite of the first hypothesis, you really have only one distinct
hypothesis. One way to guard against this problem is to vary either the
independent variable, or the dependent variable (or both) from hypothesis
one to hypothesis two.
About 4 lines.
-
TWO NULL HYPOTHESES,
each one linking only one operationalized independent variable with
only
one operationalized dependent variable.
Each hypothesis
must
be distinct.
About 4 lines.
PART THREE:
Below is an experiment which is in the
planning stages. Following it are a series of questions about this study
design. You will need to type or write out your answers to each question.
Turn your answers in with
the rewritten and augmented research problem statement on September 20,
2001 to my mailbox in 307 Stone Building.
THE PLANNED EXPERIMENT
Kim wants to see how introducing
new computer software will improve the performance of students
in WEB-based Distance Learning courses at the State University of Education
(SUE). She wants to conduct a pilot study first.
As part of her pilot study, Kim chooses
a Distance Learning class in intermediate level Information Studies. She
does this for two main reasons: (1) the instructor is on campus, so Kim
can consult with him easily as needed and (2) because this is an intermediate
level course, all students in the program for more than one semester already
have experience with WEB-Based courses.
Kim obtains the list of all students in
the class. She is able to examine prior grades, test scores, and similar
student histories. From the information on the list, it is clear that some
students are excellent academically but that others are poor. So she divides
the list into quarters, based on prior grades. She decides to use two of
these groups for this study, one from the very top quartile in academic
achievement and the second from the very bottom achievement quartile. Because
she wants to help the students with the greatest need, Kim assigns the
bottom quartile of her list to her new software program. The top quartile
is simply left alone.
Two months later, Kim examines course achievements
in this intermediate-level class. She finds considerably more improvement
in their course work among the bottom quartile group than in the top quartile
group. She concludes that her new software system is beneficial and now
plans to extend it to all of the Information Studies Distance Learning
courses.
|
The following are valid sources to use
to answer the experimental critique questions:
-
Both assigned textbooks for this course
-
All WEB study guides (and explicit links)
for EDF 5481, Fall 2002
-
Notes taken during class
|
EVERYONE MUST ANSWER THE FOLLOWING
QUESTIONS BASED ON KIM'S EXPERIMENTAL DESIGN
|
1.
Is Kim's design a TRUE EXPERIMENT? Please answer YES or NO, then explain
in a sentence or two:
2.
Do you think Kim's design in the box above has relatively high or
relatively
low INTERNAL VALIDITY?
-
Pick "high" or "low" and in then a sentence
or two, explain the basis for your decision.
3.
Regardless of your answer to Question 2 (since even the best designs can
have problems), briefly describe TWO DISTINCT
OR VERY DIFFERENT THREATS TO INTERNAL VALIDITY in Kim's
design.
4.
How is Kim able to tell if the students have improved over the two months
of the study?
5.
In
a few sentences at most, please give Kim ONE SUGGESTION of how she
can IMPROVE HER STUDY DESIGN.
September 10 2002
This page was built with
Netscape Composer.
It is best displayed in
Netscape Navigator,
600 X 800 display resolution.
Susan Carol Losh
EDF
5481 READINGS AND ASSIGNMENTS
EDF
5481 READINGS AND ASSIGNMENTS
