So it’s been a full month since my last update. This is due to holidays, business traveling at Manchester, and general hurry that this Wednesday resulted in my first presentation on my research. The audience was economic and social history PhD students and the angle was mostly methodological and data driven. Delightful session and I’m happy I was pushed by the deadline to get some writing done!
In addition to presenting my big data material as a sample research data set, my main point was that paradoxically, social scientists analysing big data must still conduct qualitative analyses. In order to make these big amounts of data understandable to humans, even data scientists are constantly making choices and selections, similarly to categorizations made by qualitative researchers.
In other words, more data doesn’t only mean more maths. Once there is large amounts of data to select from (instead of scarce data that needs attention to match the population) it needs to be reduced in order to “start making sense”, and this reduction can’t happen automatically or statistically alone. Otherwise, we risk both overfitting the data, and losing some pieces of information that computers can’t yet connect to others. This is especially true when studying complex social scientific phenomena.
I’ve been so excited lately about these baby steps I’m taking in research – and overwhelmed by all the things I see must be done. So today, I tried to prioritise my back log, to see which of the top three things to do first, I actually need to start with first. I think I’m pretty clear and have a plan… but I’m also quite sure that the plan will soon be ruined and something else comes up… well, but at least I’m striving to take the time to e.g. learn to program better, to learn about my domain, to understand the methodological options, and to find internationally the people who are interested in the same topic as I. Phew! I did ditch a couple to-do’s for later spring, but despite I still wouldn’t call it a short to-do list.