In my development as a researcher, one of the great lessons I learned from my advisor, Dr. Heitor Silvério Lopes, was the importance of clearly defining the central question of what we are studying beforehand.
I have brought this consideration not only to my scientific studies but also to the problems faced daily with data analysis (especially those involving link analysis) and in the consulting work I do for intelligence teams.
In general, when we come across a group of files and large volumes of data that need to be analyzed, our initial impulse is to dive into the data and try to understand what is available there and seek the first mental correlations. A search begins for something that stands out and might answer a question.
Especially in the world of link analysis, the initial action is usually: put everything into a diagram and let’s see what we can find. An inspection is then made to check what links are formed, what groupings are created, what bridges exist between these groups, what paths exist between the targets, etc. Everything visually.
In general, the hope is that visually we will be able to find some answer to a question.
But what is the question? What are we really looking for? What is the central problem to be solved?
It is well known that, in general, during the investigative process, we end up finding answers even to questions that were not posed beforehand, and by observing the data, a world of new questions ends up being formulated.
But what is the real question? Is it clear to me what I am really looking for? Who are the main actors? What are the initial events I can use as a guide? Did I take some time to organize my ideas, the requests I received, and formulate the questions that need to be asked of the data before starting the exploration?
In other words, what am I trying to find? What data do I have available? How can they help me reach the answers I intend to answer? Is the data sufficient? Is it organized in a way that can lead me to the answers? What diagrams will I create first? How should I structure the visual elements of link analysis based on the questions I have and the data I have at hand?
There is a great truth in this process: when we have a well-formulated question beforehand, and the parameters are clear, the ability to reach the answers expands, and navigation through the analytical process becomes better guided.
Therefore, having the questions in mind before starting the work, even if not all of them are clear yet, will help guide the analysis more effectively, as well as save time by not conducting experiments that may lead nowhere and only take up precious time from the analysts.
And for your challenge today, what is the question?
