Maximizing the Impact of Data Science Using the Scientific Method

• Read original article here

We live in a Big Data world where everything is being quantified. As a result, businesses are trying to make sense of their ever-expanding, diverse, streaming data sources to drive their business forward. If your competitors have access to the same type of data (CRM, ERP, weather, etc.) that you do, how can you keep ahead of them? One way is to get better insights from your data. They can accomplish this task through the use of data science.

Gil Press offers an excellent summary of the field of data science. According to Press, the term, data science, first appears in use in 1974. He concludes that data science is way of extracting insights from data using the powers of computer science and statistics applied to data from a specific field of study. Additionally, in our research, we found that, data science skills can be grouped into three skills areas:

To improve your chances of finding insights from your Big Data projects, you’ll need to apply all three of these skills. But how do these skills contribute to our understanding of the phenomenon we are studying? We turn to the scientific method.

The scientific method is body of techniques for objectively investigating phenomena, acquiring new knowledge, or correcting and integrating previous knowledge. This method includes the collection of empirical evidence, subject to specific principles of reasoning. The scientific method follows these five general steps:

The application of the scientific method helps us be honest with ourselves and minimizes the chances of us arriving at the wrong conclusion. By following where the data lead us, this method helps us understand how the world really works. Through trial and error, the scientific method helps us uncover the reasons why variables are related to each other and the underlying processes that drive the observed relationships.

When we cross the three data science skills with the five steps of the scientific method (see Figure 1), we see how businesses can harness the power of the three data science skills in the context of the scientific method. I posit that two of the skill areas, business (subject matter expertise) and statistics/math, play a bigger role throughout the scientific process. The subject matter expert (in this case, an expert in business) uses her skills to formulate the right questions or problem statements and generates the right hypotheses to test. The data scientists who is an expert in statistics/math uses her skills to gather or generate the data as well as to analyze that data to draw the right conclusions. Both the business-savvy and quantitative-minded data professionals work together to communicate the results of the study or help the organization take the right actions based on the study results. The data professional who is an expert in technology/programming uses these skills to help get access to the data.

Because not any single data professional will possess proficiency all of the data science skills, businesses’ best bet to successfully apply data science to solving their problems is to organize a team of data professionals who have complementary skills. We know that data professionals work better with other data professionals who have complementary skills. Business-savvy data professionals are happier with their work outcomes when they are paired with researchers who are proficient in statistics. Likewise, these researchers are also happier with their work outcomes when they are paired with business-savvy professionals. Different types of data scientists, with their unique and complementary skills, help address the specific steps of the scientific method.

To stay ahead of their competitors, companies can optimize the value of their data by leveraging different data science skills to interrogate a phenomenon of interest. Companies need different types of data professionals to adopt a data science approach to ask and answer their important business questions. To do this, businesses can hire a team of data scientists, identify current employees who have the necessary data science skills, or educate/train employees on different data science.

While current employees might not be formally trained in the practice of data science, I believe that many employees can still bring their skills to bear on data heavy projects with proper training. Over the next several blog posts, I will explore each of the three data science skills more fully. Additionally, I will provide various online resources that businesses can use to help improve the skills of their employees to help them start the journey toward becoming a more data-savvy company, giving them the insights to use their data more effectively.

An earlier version of this article originally appeared on CustomerThink.