THE MULTIVARIATE ANALYSIS OF CATEGORICAL DATA ON DOING A PRESENTATION Susan Carol Losh Department of Educational Psychology and Learning Systems Florida State University |
|
|
|
|
You will prepare a 15-20 minute (MAXIMUM)
oral presentation based on your analytic paper.
Virtually everyone does a Power Point
presentation and the program is on our class machine.
Place your presentation on an external device: e.g., a CD or
usb drive and bring it to class. I will upload it to Blackboard.
Alternatively, you can send it to the
Discussion
Board two days in advance!
Bring handouts! Some people simply print their Power Point presentation.
Others have a blank lined front page (a lot of audiences like to take notes)
with some references at the end.
This gives people something to hold in
their hands and look at! That way people are not staring at you all the
time.
PLEASE INCLUDE BIBLIOGRAPHIC REFERENCES IN YOUR HANDOUT!
REHEARSE your presentation. This helps in several ways:
|
|
.
|
Topic | |
Addy: Tuesday 4-25 | |
Brian Tuesday 4-18 | |
Chena Tuesday 4-25 | |
Fanny Tuesday 4-25 | |
July Tuesday 4-18 | |
|
Know your audience and your purpose. Since we are an analysis course, the bulk of your presentation should be directed at the logic and justification for your choice of an analysis and the results and interpretation of your analysis.
Set the stage.
Briefly
describe your "research problem" in a couple of paragraphs (your topic,
its importance, and what your analysis can add). Briefly describe your
major concepts (another couple of paragraphs?).
Will
you be using a causal model? If so, before you begin the description
of the empirical dataset and how you will operationalize these concepts,
sketch out your causal model with the conceptual variables and the links
among them.
Methodology.
Describe
your choice of dataset (perhaps it's the only one available that deals
with the research topic; perhaps it is the best choice known; perhaps it
is an underexplored dataset, etc.) and why it is a good choice for
your topic (another couple of paragraphs).
Describe
how you will operationalize your constructs. What are the specific
measures you will use? Give enough description so your audience knows
what you did or how they were concretely measured. Pay particular attention
to the measurement level of each variable (nominal? potentially
ordinal?) and the number of categories in each one. If you have
recoded to collapse categories, you must mention that and how you recoded
the categories.
What
kind of analysis did you do? Why was this the best choice for your
analysis for this research problem and these data?
This
is a good time to mention any particular problems you have found
with the data. (There are ALWAYS problems!) Was it the nature of the sample?
The number of cases? The way variables were measured? The amount of missing
data? Or whatever was the case for you. With a couple of exceptions, most
of you are working with secondary analyses--i.e., sets of data collected
for an original purpose that may have differed from yours and by another
individual or institution. Therefore, it can be frustrating because you
have different intentions for the data, there may be too few questions
on your topic or the sample may be the wrong age, or there may have been
biases in collecting the data.
Results.
What
were your analytic results? These are NOT your output (in your paper,
please append output at the end for your paper; for your presentation,
the output is basically invisible!) Analytic results include:
Discussion, Limitations,
Implications.
What
were your SUBSTANTIVE conclusions? This is where you put your findings
into WORDS and you say what the results MEAN.
Gosh,
is there ANY time left? Well, if there is, this is the space for your limitations
and implications. You already touched on some of the empirical limitations
above. Maybe there are some issues it is tough to surmount (try to figure
out the gender in a drawing by a first grade boy...could another measure
be used instead?) Do your results have any policy decisions (make exercise
more appealing; help parents get their chikl ready for school, relax math
anxiety.)?
Call
your audience's attention to the reference list you have provided for more
information and ask for questions.
Wait
for the applause!
![]() |
|
READINGS |
![]() |
This page created with Netscape
Composer
Susan Carol Losh
Aril 13 2017