How do you do I... am on the set of Rocky Horror, here at the
Springfield Contemporary Theater with a very important message: if you want you want
your research project to become a "wild and untamed thing", simply follow this horrible advice
It's astounding, time is fleeting, madness takes its toll, but listen closely...
don't bother consulting a statistician until your data are already collected!
Why talk to an expert before you begin? Jumping in without
knowing what you're doing can thrill you, chill you, fulfill you... okay, maybe not fulfill you,
but do you really want expert advice when you could just go it alone...
like a creature of the night?
Don't even think about collecting data until you know exactly what you're going
to do with it. Brad: "Dammit Janet" (slut)
I mean... go ahead get busy collecting data. You can figure out the rest later.
"You better wise up, Janet Weiss." It's never too early
to begin collecting data. Who cares if you don't have a research design?
Or even research questions? Your statistician can figure out all of that for you later?
In just seven days I can make you a man and you should always remember to make your own survey.
Validated surveys may look professional and scientific.
But don't get strung out by the way they look. Make your own!
You've never made a survey before? How hard could it be?
If the numbers don't tell you what you want to hear, torture them till they do.
"I've done my best, God knows I've tried to find the truth, I've even lied".
Is your test still not significant?
Just keep collecting data until it is.
Do the findings say that your program doesn't work?
Throw them out and substitute those that say that it does.
Or collect piles of data and only report the significant findings.
Why settle for Science, when you can have
science fiction...double feature.
Sometimes data goes missing. What a pity. Don't bother looking for them.
Don't get hot and flustered... all studies have limitations... just mention
them at the end as if there was nothing you could have done.
OR... if you don't want your project to be so rocky, then ignore this horrible advice.
Instead, talk to your statistician before you begin and know what you're going to do
from beginning to end, before you collect data. If you do that, then you won't be dealing
"cards for sorrow, cards for pain".
And if you dream of being a good researcher...
...don't dream it...
Be it.
Camera voice: "And, cut"
All right, good, you got it? "YEAH" Great...
"I'm going home."
Không có nhận xét nào:
Đăng nhận xét