Fourth year. It means different things to different people. Sometimes it means the end is in sight, you have a paper out and next year you will graduate. Sometimes it means that the end was in sight, but it has relocated somewhere on the map, you can’t find it, and you’ve stumbled into a very dark cavern where the walls are getting closer and closer as you move forward. And sometimes it just means you are chugging forward, sometimes running, sometimes sprinting, sometimes taking a water break. There is no universal experience for fourth year of graduate school. Heck, there’s no universal experience for graduate school, period.
My fourth year has been full. Our lab changed buildings. I am more involved with the Graduate Student Organization this year. Vikings killed off my favorite character (RIP Aethelstan). Carey Elwes (aka Dread Pirate Roberts for you Princess Bride aficionados) told me my name is beautiful (EEK!!!!!!!!). And the real kicker – I have spent most of the year with a gluttony of null data. What is null data, you ask? Null data is data that isn’t. It is data that tells you what your answer is not. It is the exact opposite of telling you what your answer is. Is it valuable? Is it garbage? The debate is fierce, and off-hand remarks about it can leave a graduate student scorched, like a thatched roof house after a viking raid.
Here’s a fun fact. If you look null up on Urban Dictionary, you get the following definition: NULL – n., adj. – signifying the absence of data. Things such as character strings and various types of numeric values can be described as “NULL” when they contain no data. (shadow:light::NULL:data). And, of course, this is really in the context of programming and database mining. But think on the first part, “signifying the absence of data.”
Imagine spending 5 months collecting data. You graph it, run statistics and the results you get show that it does not prove your hypothesis – maybe you treated animals with a drug over time and expected to see an effect, but see no effect whatsoever. You then panic, question yourself, question your technical skill, stop going to seminars because you feel you have to focus on fixing what is wrong, go through a few existential life crises, google clown college and patent law and take another 5 months to repeat the experiment. Same results. By the time this cycle is finished, you have lost, most likely, at least a year of time and know exactly what the drug is not doing. You HAVE data. By this time, you have a lot of data, actually. But it has no whistles. It has no glitter. It has no BOOM! POW! WHAM! In effect, you don’t actually have usable data, in terms of publications. You have negative data.
That, dear friends, is the curse of null data and the part that gets most publicity when talking about data with peers. If you are one of these unfortunate people, your conversations with peers might look like this:
Peer: How are things going?
You: Fine, just working on the usual.
Peer: Anything exciting?
You: Well, not exactly…
You: I have a lot of null data right now.
Peer: Oh no, are you going to change your project? [cue rise in heart beat] Is it even worth still doing? [cue existential life crisis] What did your PI say? [cue flashback to feeling stupid] What did you do wrong? [cue spiral towards depression]
And that’s how it goes. But here’s the blessing of null data that hardly ever gets publicized in conversations: Null data verifies. It verifies what it’s not and what it is. You don’t see an effect of the drug, but you start realizing, through other observations, the reason why there is no effect. And suddenly that worthless piece of null data is essential to your story. I posit, I assay that null data is essential to understanding the full story of your project. So, if you are wrestling with null data, or data that isn’t working out or data that seemingly gives you no answers, keep the faith! You can’t build bricks without clay, you can’t get answers without data. Keep moving forward and eventually you will find the light!