They say everyone lives their last year of grad school on perpetual edge, precariously balanced between triumph and blubbering insanity.
Today I got my first taste of the darker side of this liminal state.
It all started when I began doing the recodes for Wave 2 of my dissertation data. I had finally finished merging the two waves--this is harder than it sounds, I might add. Sure, there is a fancy tool in the software that allows you to merge datasets. All you have to do is give it a variable that will be the same in both datasets, then the computer does the rest.
Riiiiight.
The variables I had in common were respondent's first name, last name, and email. Email seemed the best option. So, I merged. But only about 3/4 of the cases merged this way. The rest the computer couldn't match. Some students mistyped their email address (name@school.ed - WHERE'S THE U?!?). Some inexplicably decided to capitalize their address. Some gave me a different address each time. I had to go through manually and match those up.
I still have three respondents for Wave 2 that I can't match to anyone in Wave 1. Which is weird, since the survey request only went out to those who had taken Wave 1. Haven't decided what to do with them.
So, once merged, I went to do the recodes for Wave 2. A lot of it is exactly the same as Wave 1 was, since most of the questions are repeated. So, I brought up my Wave 1 syntax file (SPSS, all I got at home). And I'm going through the file just tweaking the variable names to work for my Wave 2 variable names, and lo - there's a repeated line. An entire set of variables recoded twice! Ack!
So, I had to go back to the raw data for Wave 1 and compare/contrast, see if my variables were actually coded, then coded back again. And they were. So, fixed that. Luckily, I haven't used Wave 1 for much of anything yet.
Finished recoding Wave 2. Ran descriptives. Noticed something funky with the international students. My old descriptives report said I had 73 international students. My newly minted run says I have 83. I am still not sure what happened there. At this point, I am chalking it up to an initial mistype. I mean, the second digit is still the same, right? Still, it's disconcerting.
Other than that, my descriptives look great. I am feeling good. I want to hold my descriptives in my hand, walk around with them, look lovingly at them, and maybe sleep with them under my pillow for good measure. So, I go to print them out.
The. Printer. Is. Out. Of. Ink.
I go to get some ink.
But it's Sunday, and it is after 8pm.
After trying two of my preferred places, I end up at a Big Box Store. Those things are always open. I find ink. I wander around the aisles a bit, in something of a daze.
Pretty towels. Shiny camera. Oooooo.
I pay for my ink. The cashier gives me my receipt, AND a coupon for Starbucks. There's a Starbucks in the store. I think, "OOOooo a fancy caffeinated beverage is EXACTLY what I need to put the mess that was this day behind me!"
The store was closed.
I almost cried then.
Sunday, July 5, 2009
Tales from the Dissertationside
Posted by
Anomie
at
8:45 PM
Labels: a day in the life, academia
Subscribe to:
Post Comments (Atom)



3 comments:
Congratulations on entering the dissertation netherworld!
Congratulations, indeed! Sorry about the caffinated beverage -- I HATE when that happens!
Also, I know that this is of little help now, but in case you have to do something similar in the future, one trick that I learned is to lower-case all of the strings before you try to merge. In Stata, you can make a new variable using:
gen emaillc = lower(email)
I am sure that there is something similar in SPSS. Good luck!
Haha, I love that people are congratulating me on my horror story :) . I feel like I've passed some sort of academic hazing ritual.
Post a Comment