So while browsing through job applications, I thought to myself, 'Nope, nope, I think I have to find something outside academia'. Maybe it's my own limitation of not being smart enough to make my own projects. Or, my shyness to do more networking hinders my chances to be more well connected. I don't know. But I also know that I like data mining, statistics, and analysis. This weekend I was doing some code learning, and oh boy oh boy, it was terribly fun! And then we have our group meeting/journal club this Monday, and although the subject is something I like (writing papers!) in the back of my mind, for whatever reason, I keep on thinking, I gotta get away from academia, gotta get away from academia!
Maybe it's the thing about a week. It's not a new news that struggling in this world is tough. Professorships are low in number in Germany. Being a postdoc, you have to be really really high qualified. I notice, personally, there are sooo many PhD graduates every year! (I mean, I already attended 6 PhDs defense just by the end of February this year!) Most, wanted to stay in academia, but there's not enough space! And so I know myself, and my quality, and realizing what I enjoy the most, I know I should do something else. The thing I want to do the most is either do work that will practically apply approaches to betterment of the degradation of environment, and I'm not joking here, because I am truly concern at the rate we're going, and we need solutions or else there won't be any world left to do science at anyway. But to do such work, I need experience. And my years of dealing with soil samples isn't really helping. But I tried. I applied for a position related to that.
The other thing I'm interested with is to do data analysis. Really. I LOVE IT. I could spend hours just struggling in front of R. I'm currently doing meta-analysis (hahaha how long have I been talking about this, 2 years??) and today I just got stuck, but it'll worth it when I figure out how to get around with it. I was reading a meta-analysis paper this afternoon, after running away from office because I felt so stuck, and I was so surprised that reading this was so much easier compare to reading conceptual papers!! See, I really shouldn't force myself on something I'm mediocre-ly good at, I should focus my energy on something I'm actually good at!
This afternoon I was also chatting in a group whatsapp of fellow Bremen Indonesian master and doctorates to be, and I got handy advice about yeah, graduating as a doctorate does not mean you're gonna work on the same subject. More likely, you're gonna have to learn other subjects, and do your best at it. And you know, data scientist isn't so bad. It's even - at least for me- thoroughly exciting. I LOVE building figures, graphs from scratch. I love interpreting statistical results after a rigorous careful 'cleaning' process.
Which reminds me of another thing. I thought about it on my way home. Especially after today's discussion. Okay, go back a little, so yesterday I cleaned up my raw dataset and re-run my analysis. I practiced and tried everything already last week, but I know I have to do a quality control. So, after finding some mistakes here and there, I was ready to go. But I know I still don't fully understand what bootstrapping confidence interval is about. How do I know it's already 'okay'?
So I talked to a friend who is absolutely excellent in meta-analysis. She said, did you check the normality of data distribution? Well of course NOT, because I found that there's a huge heterogeneity, so I practically assume it's not normal. She said, check with Saphiro (significantly not normal, damn it), and then you can transform your data. Aaaah I don't wanna transform my data! Make me so confused with interpretation!! Can I just do something else? And then she said, now that's the purpose of doing bootstrap. If the distribution after bootstrap is normal, or the Q-Q plot fits the fitting line, your data will be fine. And boy was I relieved. Although I still have to understand meaning of data spilled by the bootstrap function. Unfortunately this is gonna be a super busy week! But people are important too!!!
Anyways, I'm doing a rant, topicless blog post this time now. But, honestly, I really really reaalllyyy wanna do data scientist work. Relational model using SQL practice I have this weekend was sooo much fun. Doing R is SO much fun. Practicing Python is also really really fun!!! I hope I can have a job soon related to this! And oh yeah, I'm a job seeker now, so if you got any data analysis/research scientist/data management work, please let me know!! :P
No comments:
Post a Comment