It’s been a little while since my last post. I took a vacation…actually, my honeymoon…last week to New Orleans. Getting away in the middle of the semester…and in the middle of data collection…was exactly what I needed. The problem with “getting away” is the “getting back.” No matter how wonderful and relaxing a trip was, stuff piles up while you’re gone and…well, let’s just say it’s a good thing you just rested up for a bit because the stress is going to be a little higher for a while. This week’s stress could be filed under any of the following labels:
Data Issues
Preliminary Analyses are No Joke
Proof. Read. Everything. (At least 25 times)
Save Your Data File Before Trying Something New
But…probably fits best under “Lessons Learned”.
Regardless of what we label it…I’m going to open the file and share a little bit about my data stress this week.
As I said in my last post, it is time for a mid-data collection hiatus to analyze the data I already have. This includes over 200 surveys (at varying levels of completion) and 7 interviews. I’ve been cleaning the survey data in SPSS to prepare it for analysis. I am privileged to have two students working with me and so I saved some of the cleaning to do with them so they could learn how to do it. After the last student lesson, I had to rush out for a meeting. I knew I’d be coming back to work in SPSS again so I just left it open. A couple hours later, I’m back and finish the cleaning–days worth of work–hundreds of rows of variables cleaned. I had a little time left so I decided to explore the data a bit. What happened next is a little fuzzy (or maybe I’m just embarrassed to admit it)…but it involved something along the lines of this:
Me (to myself): Huh, I haven’t seen this before…I wonder what this command does. (Clicks it…doesn’t understand what happened…goes to “undo” and gets a message similar to “Undo did not undo previous step”)
Me: Huh? (clicks undo again, gets same prompt, realizes it’s “undoing” the last 2 steps of cleaning, and not the other prompt)
Me: (Not sure how to make the prompt go away, comes up with a genius plan) I’ll just exit and not save because I saved after cleaning earlier and it will go back to that state. (Engages in said “not saving” and re-opens. Realizes file was not saved after earlier cleaning. A decent amount of work is now undone. Cue tragic music.)
Lesson learned: Always save. Constantly. Especially before trying something new you don’t understand.
So, I re-cleaned the data. It was easier and faster the second time around…and there’s still more to do, but everything is labeled and has the appropriate variable type, etc. As I finished this second round of cleaning, I remembered that one of my measures has both positively-worded and negatively-worded items. I looked up the original measure to check the scoring so I could make sure it’s scored correctly in SPSS. The original measure has 20 items. My survey had 19. <Insert Panic Here> Not only had I somehow forgotten an item, but I had also repeated one making my measure 18 instead of the full 20. Now, if you’ve used validated measures you know this is not okay. It’s important that the measure matches (as closely as possible) the original measure. I immediately went to the place where I start exploring alternative career options (Elephant Handler) because I’m sure it’s all over. I’ve screwed it up completely.
I may have jumped the gun a bit…As luck would have it, I had a meeting with one of my advisors the next morning. The measure is an index (it’s a summed total–the higher the number the “more” the person reports whatever that measure is measuring), not a scale where there are required scores (for example, to reach clinical significance) based on previous research. This is good because I just make 18 the highest score instead of 20. Do I wish I still had the full 20 in there? Of course. But is it ruining my dissertation? Nope. Now, you’re probably thinking to yourself, how did this happen? Did she just not proof read her survey? This would never happen to me. And you’d be fair in thinking these things. I might have. But here’s the thing, I did proof read it. So many times, I can’t even give you a number, but it’s probably at least 10. I have no idea how this happened. But it did. And I’ll deal with it. Just another fun lesson learned!
As I’ve written these two incidents out they hardly seem as significant as they felt in the moment. Maybe that’s the real lesson learned: Take a step back when things feel completely overwhelming (or messed up or impossible or…) and let it sit for a bit. It’s not the end of the world and it will probably be okay. Most things are eventually. Oh, and breathe. And don’t run away to work with the elephants no matter how tempting it might sound in the moment.