| Psych 290 | |
Graduate Research Methods: How to do stuff
|
back to course homepage
The Motivation:
Once upon a time, being able to do statistics was the only technical
skill
required of a psychologist. All psychology departments require their
students to learn statistics. The logic behind the stats requirement is
that it provides us with a set of tools that may be useful at different
times down the road (even if they're not now). Having more tools in
your
conceptual tool-box makes you a better scientist - it affords more
flexibility and creativity to how you approach questions, and in general
is an all-around good thing.
So far so good. But statistics is no longer the only technical skill
that
psychologists use. We write programs to run our experiments, we use
eye-trackers, fMRI, EEGs, and physio equipment, etc etc. All of these
are
tools. Just as with stats, having a hands-on familiarity with these
tools
affords greater flexibility in how you as a scientist can approach
questions. I know several people for whom something as simple as
learning
how to use PsyScope to run their experiments has revolutionized their
research programs (it helped them overcome what before seemed like a sea
of technical impossibilities).
This graduate methods course was created this year on the request of the students.
It will be organized by a senior graduate student with a more junior
student assisting and preparing to continue the tradition in the
following year. Most of the teaching will be done by guest-presenters
(many of whom are themselves students and may even be students in the
class). Here's some hype....
Topics to be covered:
Basics of experiment design
-
What are _all_ the ways you can use to get at any given
question? What are the trade-offs? What are the potential
problems/confounds of each method?
Computer-based experiment design - PsyScope
-
Still using pen and paper studies? Or maybe you're programming
everything from scratch in C? Come learn PsyScope - it's free,
and downloadable from the web. PsyScope is a quick easy tool
for
making computer experiments. It can counterbalance and
randomize,
take all kinds of input from your subjects and accurately
measure
reaction times. All the data is stored in a file that's
readable
by data analysis programs (no more tedious data entry!)
Web-based experiment design
-
It is now possible to run experiments on the web that are almost
as sophisticated as anything you can do in the labroom. Running
your experiments on the web means (1) you never have to spend
time
running subjects ever again, and (2) you could reach all kinds
of
populations that are hard to actually get to come to your lab -
different language/culture groups (imagine if you didn't have to
go to japan every time you wanted to run a cross-cultural
study),
older adults, etc.
Data analysis - Excel & SPSS
-
OK, so you've learned minitab, now how do you actually analyze
your data?
Subversive Statistics
-
Time to go beyond simple hypothesis testing. How do you analyze
and conceptualize large amounts of data? Josh & Tom will
introduce us to Cluster analysis, Multi-Dimensional Scaling, and
other techniques for getting structure out of data that go beyond
just comparing two columns of numbers.
Data presentation-
What are the best ways to present your data for papers or
conference presentations? Research that comes with a nice
illustration makes it into textbooks, and other equally
interesting research doesn't. How do you make sure your research
makes it into textbooks and gets talked about at cocktail
parties?
Psychophysiology
- James and his students will open up their lab to teach us about
psychophys methods and data analysis. What can heart rates and
skin conductances tell you about what your subjects are
thinking?
Eye-tracking
- We can learn all kinds of things by watching where people
look. Zenzi and lab will introduce us to the basics of
eye-tracking and what it can do for you.
Developmental methods-
Developmental psychologists have developed the most ingenious
methodologies in all of behavioral science. Not only will
you learn about some of the technical challenges of working with
kids, you can also use all of these methods with adults.
Survey Construction-
How do you make an effective survey. What kinds of things
should
you worry about? Should you always use a 7-point scale?
Plus... How to become a famous Psychologist: a wine and cheese
reception with Phil Zimbardo