Remembering that Research Ain't Easy
There's a project at work where we needed to get some data out of our ILS - essentially a very large, complex database - and doing so had a bunch of hiccups. Some were user error from not understanding or remembering the terms for the query tool, some were because the ILS is not very good at doing its job (none of them are), and then some were because I realized my colleague who I was working with wasn't familiar with Boolean logic.
They were not understanding how I approached the queries to pull the data because they don't think of search in those terms. I was struggling to explain my approach because I assumed they knew Boolean. Once we figured out that problem, we could go back to getting the data we needed to analyze.
That somebody working in a database intensive environment could go years without learning Boolean logic reflects how far our tools have evolved. Searching databases used to be precise yet crude. You had to understand the structure of the database to construct your search string based off of different elements and the interfaces were sparse. But now, people can free associate into a simple box and get a plausible answer back without needing to think much of what the results actually are. It's smooth and easy which has lead to many of us to forget what these tools are actually doing.
I am old enough to have taken a class in grad school on Dialog - the old school database platform that still lives on in-spite of many mergers and acquisitions. (Though not old enough to have really used it in my work.) The exercises required us to reflect on our research questions – who would be publishing on the topic, which databases would have the best results, what are the controlled vocabularies for those databases, and so on. And then we had to construct the most precise search string possible - using Boolean! - for each database file. This was because you had to pay by the minute to use Dialog, so the culture of librarians using it was to be as economical as possible before logging on. Then you'd export the results and log off. This method really forces you to think about the question before you start your search. It seems antiquated now, but the principles are still relevant.
I don't totally long for the days of Dialog (though I am prone to nostalgia), I do think it's important for all of us to slow down a little bit with our research methods. The allure of frictionless searching obscures the functions of these tools. I love Google Scholar for a lot of things, but I also warn people that it can be very problematic, especially when Google changes the algorithm without warning. This is also true of AI (LLM) search tools that scrape the web and package the results with convenient synthesis. (That's probably a whole other post.) The ease and speed with which we can access research material leads us to think we need more and more literature, but we still have to read it. One way to make it easier is to be more precise with your searching, but that probably means thinking more about the tools and using more appropriate ones. That's effort though.
To close this out - I'm working with my colleague to use this messy report from our ILS to learn Boolean enough to to help extract the actual data we need. And curse these expensive and fancy databases that over promise and under deliver on their core functions of search and retrieval.
