In the dynamic landscape of information technology, solving client challenges isn't always a straightforward task. It requires a systematic and logical approach that goes beyond the surface of the issue. In this first chapter of a two-part blog series, we will explore the application of the scientific method in addressing complex problems, be it in scientific research or the ever-evolving realm of business challenges.
When a client approaches you with a specific question, it's tempting to jump straight into proposing a solution. However, this simple approach often overlooks the intricate process of problem-solving. For straightforward problems, a non-systematic approach might suffice, drawing from a pool of known truths and assumed knowledge. But when faced with highly complex issues and a scarcity of proven tactics, a systematic approach becomes paramount. Enter the scientific method.
‘Science is more than a body of knowledge. It’s a way of thinking. A way of skeptically interrogating the universe with a fine understanding of human fallibility’
— Carl Sagan , / at /
The scientific method offers a systematic and logical framework applicable to a myriad of situations, whether they be scientific research questions or intricate business problems. This method involves several key steps:
When faced with an observation, a client's question, or any problem, it's crucial to recognize that these may not represent the core issue. The initial challenge must be transformed into a focused research question. This involves determining what needs solving, the applicable framework, business and technological constraints, measurable properties for solution assessment, and the necessary data for analysis. Asking fundamental questions like what, who, where, when, and how aids in translating the problem into a researchable form.
Once the researchable problem is identified, the next step is to formulate a hypothesis. This is an educated proposition or solution to the problem at hand. “When we do A, then B will happen.”, “If we obtain C, we can calculate D”, ...
The hypothesis should be measurable, allowing for the possibility of proving or disproving it in the subsequent steps.
Designing experiments is a crucial step in the scientific method. It's essential to be wary of biases and sampling errors that might skew the results. The experiment should be objective, equally likely to prove or disprove the formulated hypothesis, ensuring a robust path towards a solution.
Upon completing the experiment, a thorough analysis of the results is necessary. Does the data confirm or reject the hypothesis? Are there any lingering issues or unintended consequences? Even a disproven hypothesis can yield valuable insights for formulating a new one.
Regardless of the experiment's outcome, documenting and communicating the results within the team is crucial for learning and growth. Assessing the potential for further research and proposing new hypotheses based on the conclusions is a vital part of this step. Never let a failed attempt discourage you from continuing.
Now, let's illustrate how the scientific method can be applied to a fictional problem within an IT context. Consider a navigation application based on GPS, receiving user feedback that the arrow indicator on the map drifts and jumps away from the route.
The problem is identified as an issue in location detection causing fluctuations in the UI. The research question is framed: "Can we address the location detection issues to improve user experience?"
You find that the GPS data received in the application often has bad accuracy in an urban environment.
The hypothesis is that filtering out GPS data with bad accuracy will eliminate the fluctuations in the arrow indicator.
A beta-test is conducted by filtering out inaccurate GPS data, and user feedback is collected to evaluate the improvement.
The results indicate a stable arrow indicator but with a new issue – static location. This means it often stands still while moving.
An iteration is required as the filtering solution introduces a new problem: the location becomes stale. The question now becomes: "Can we improve bad location updates without filtering them out?"
As the conclusions from the original proposed solution identified a new problem, you will need to reiterate through the steps to find a proper solution.
The new problem is identified as the risk of the location becoming stale when filtering bad location updates.
The hypothesis is formulated: projecting the location on the predicted route will eliminate fluctuations and prevent the location from becoming stale.
The solution is implemented, and the beta-group provides feedback for evaluation.
The results show a slightly less stable arrow indicator but without static location issues.
The new results indicate that the proposed solution is valid, and the team can proceed with its implementation.
The scientific method provides a structured framework for tackling complex problems. It inherently guides the problem-solving process, facilitating the discovery of new information, interrogation of proposed solutions, and surfacing of true problems during investigation.
This first part of the blog post has delved into how the scientific method works, highlighting key points to consider. It has been applied to a fictional problem to showcase its practicality in an IT scenario.
In Part 2, we will explore the application of the scientific method to solve a real-world problem: "Is it possible to detect the arrival and departure of a train in a station with a time resolution of a couple of seconds?"
Read Part 2 now as we unveil the challenges and triumphs of applying this method to a seemingly straightforward question with deceiving complexities.