Your First-Year Project

[ The Happy Hour Archive | obereed.net ]

[C&P Happy Hour, Friday 4pm, Charley's - Be There! ]
-----------------------------

SPECIAL SEMINAR:  HOW TO COMPLETE YOUR FIRST YEAR PROJECT

Renown author Dan Horn will be the guest of honor on at a very special
brown bag seminar.  This Friday at 4pm at Good Time Charley's.  The topic
of the discussion will be his much anticipated article "How to Complete
Your First Year Project:  Tips From the Pros."  Please join us for this
informal get together.  The article has been attached below:


==========================================

How To Complete Your First Year Project:  Tips From the Pros

Dan Horn, Herb Simon, Ed Smith, John Anderson, 
Saul Sternberg, and William James


Abstract

As Many of you know, the third year students are just finishing up their
first year projects.  The process of developing a unique contribution to
the field is a long and arduous one.  In this brief article we attempt to
convey a number of useful strategies or "heuristics" (Tversky & Kahneman,
1974) that can be used in this endeavor.  

-----------------
	The authors wish to acknowledge the insightful and helpful comments of
Anne Treisman, Dave Meyer, Phil Johnson-Laird, Robert Sternberg, Gordon
Bower, Hal Pashler, John Jonides, Sara Purcell, Hazel Markus, Lynn Hasher,
Gus Craik, and Mike Gazzaniga.  
	Address all correspondence to Dan Horn, University of Michigan.
-----------------

	The first year project is often considered the most productive use of any
three years in graduate school (Brinck, 1996).  It is a time of great
excitement, intrigue, fascination, and boredom.  However, it is important
to keep in mind that, like any other time consuming rite of passage, the
first year project is meant to build character.  
	The reader may recall from his or her childhood that anything that is
intended to build character should be avoided like the plague.  However,
given the fact that we cannot simply brush off this responsibility, we have
developed a set of tips that should make completing the first year project
a breeze.  As Wilhelm Wundt so aptly put it, "[Dan ] ... should be ...
considered to be ... [a brilliant source of ] information, and ... [you
would do well to pay attention to his advice ] (1866, pp. 1-27)."
	
STEP 1:  Choosing a topic

	Many people make the mistake of choosing a topic that is too broad for the
scant three years that they have for their first year project.  It is
important to pick a topic that is within the scope of your abilities, and
permits a study of a single self-contained phenomenon.  This may sound like
a difficult task, yet it needn't be.  We have recently developed a time
tested procedure for choosing an appropriate topic.
	

	TIME TESTED TOPIC PICKING METHOD

	a.  Read a few of your advisor's articles
	b.  Ask your advisor what they would be interested in studying
	c.  Select a subset of that interest
	d.  Write down your idea carefully
	e.  Develop an example of this phenomenon
	f.  Declare it a special case, and make this your new topic of
		interest
	g.  Repeat steps d through f until you are left with a single word, 
		this is your topic.

STEP 2:  Designing the experiment

So, now that you have your topic, its time to devise an experiment.  You'll
want to make sure that your method is sound, your measurements accurate,
and your theoretical assumptions well represented.  Of course, we're just
kidding...  Seriously though, your goal is to have an experiment that:

	a.  is easy to set up
	b.  can be run by an undergraduate R. A.
	c.  is easily analyzed (i.e., has only one IV an one DV)
	d.  is of no theoretical interest (that would mean a lit. review)

STEP 3:  Analyzing your data

	It's been said that statistics are more an art than a science.  We're
inclined to agree with that, and would add that they're even more like
artists than art.  Specifically, statistics tend to be bizarre, complex,
annoying, and utterly useless.  Therefore, our biggest tip to you is to
simply make up any statistical values you'd like.  That's right,
psychology's dirty little secret is that everyone makes up their p values.
In fact, most statistical programs are simply complex random number
generators.
	What makes a good scientist a great scientist is the ability to rely on
statistical methods that confound and overwhelm the reader into accepting
any point of view (e.g., Tversky, Gilovich, & Vallone).  Now, you must be
careful in your use of statistics.  For example, if you want to claim that
subjects in one condition were significantly better than those in another
condition, yet the mean for the second condition is higher than the first,
you must rely on another useful technique.
	Bayesian statistics were created to allow psychologists to make all manner
of bizarre claims.  If in the preceding example, you were to claim that
group one was significantly better in a *bayesian* analysis, you could then
tell naysayers that they are simply (and stupidly) "ignoring the base rate."
	If you've mastered these methods, you may wish to try some advanced
statistical techniques.  The most effective is the invention of fictitious
measures (e.g., Fisher's Beta Squared, Tukey's Gamma Corrected Test for
Skewness).  The key is simply to pick the name of a famous statistician,
add a Greek letter or two, mention degrees of freedom or mean squared
variance error estimates or something, and you're set.  This time tested
method will surely enhance your credibility among your peers, as they will
feel inadequate in their meager statistical knowledge.  
	There is however a danger to this strategy.  You may be considered a
statistical "guru" (from the Latin *gurium* meaning "condescending
windbag").  If such a fate should befall you, remember these simple
phrases, "Geez, you can't do that with your variances," and "I'm pretty
sure that you're violating some assumptions."  When approached with a
statistical question, merely recite these magic phrases while shaking your
head incredulously.

STEP 4:  Writing up your project

The Title

	The first step in writing up your project is to come up with a clever
title.  The title is the most important part of the paper, except for the
acknowledgments (see below).  APA format is pretty specific in this regard,

	"A title should be no more than 45 words, and should
	 indicate to the reader what the topic of your paper
	 is in a succinct manner.  Each word in the title 
	 must be capitalized, and your title must contain at
	 least one colon [: ]. (APA, 1994)" 

	If your paper refutes an important scientific theory, you should try to
personalize it a bit.  So rather than, "An Alternative Theory of
Intelligence" you may want to go with, "Dual Facet Sub Theory My Ass:  The
Problem With Sternberg's Triarchic Theory of Intelligence."  It just has
more zing.

Authorship

	Perhaps the stickiest of situations in writing an article is the question
of authorship.  Again the APA has set specific criteria for authorship.  If
you really care you can look them up yourself.  As far as we're concerned,
the more the merrier.  We advise giving famous psychologists authorship to
inflate the importance of your work.  If you want, you can think of it as a
tribute, kind of like an honorary degree.  You're saying, "Thanks for doing
such a great job!"  For example, what paper on working memory wouldn't
benefit from Alan Baddeley's name on the author list?  We're sure he
wouldn't mind.  Even if he did find out, he should thank you for helping
pad his vita.

Acknowledgements

	Often overlooked, the acknowledgement is the secret weapon of the
successful researcher.  If you acknowledge famous people, your readers will
think that you are associated with, or at least know these folks, and you
will rise in their esteem.  Even if the extent of your relationship is
stalking him or her during trips to the grocery store, acknowledge their
insightful comments.  You're not saying that these comments were directed
to you, or even that you've heard any insightful comments from them.  All
you're doing is acknowledging, or admitting, that they may have made some
comments that may be considered insightful.

Content

	N/A

References

	More and more, a person is judged by the way he or she references.  There
are a few simple tricks here:

	a.  Reference one of the following dead people in your opening 
		paragraph:

		William James, Wilhelm Wundt, Sigmund Freud,
		B. F. Skinner, Emmanual Kant, Jimi Hendrix, 
		Charles Darwin, Hermann Ebbinghaus, Thomas
		Watson, Francis Galton, Paul Broca,
		Hans Eyesenck (new addition), or Socrates.

	b.  Pepper your text with references to authors who may read your
		paper.

	c.  As my advisor Gary Olson says, "cite relevant literature" 
		(Olson, personal communication, October 1997)

	d.  If you need to fill up space, be sure to include a large 
		number of long citations that have absolutely nothing
		to do with your paper.

	e.  We shouldn't even have to tell you to cite popular people.
		Whether their work is even remotely related is of no
		concern to you.  


CONCLUSION

This document is by no means meant to be comprehensive.  Our goal is merely
to provide you with a useful framework for approaching the problem.  We can
only hope that you will use this newfound knowledge for good rather than
evil.  Of course we take no responsibility for any actions that you take in
following these prescriptions.  But if you do follow these simple steps,
you will find that your first year project will be significantly (Fischer's
Modified Tau, MSE=3.1, P<.0001) better than your colleagues'.