From the very small to the remarkably large, statisticians have dealt with data sets in varying sizes throughout history, most of which were collected manually. In the past, large data sets tended to be atypical, given the resources required to create them. Strong statistician skills were necessary for appropriately managing data sets of all sizes.
Over the past several years, however, the acquisition of extensive, often highly complex data sets (generally referred to as big data) is done faster and less expensively than ever before. Suddenly, statisticians who were previously working with data sets just large enough to power a statistical procedure have found themselves veritably drowning in data.
Data Rich, Information Poor
Many have taken a positive view of this rapid influx of data, as we have become a data-centric world. As some say, “the more the merrier.” Data are produced all around us—every time we use social media, touch our smartphones or go grocery shopping, we are providing businesses with stores of data regarding our behaviors, our attitudes, and ourselves. However, sometimes, rather than being “merrier,” "more" can be confusing. Several industries find themselves in a situation where they have an abundance of data but no clear understanding of the information and insights stored within it. Those trained in computer science are likely to have the capacity to organize the data and build algorithms around it, but the meaning contained within those models can still be elusive. This is where a great statistician comes in to make sense of the frequently insensible.
What makes a great statistician? The following are some of the traits of those who are good at statistics.
- Analytical Skills—a background in mathematics is essential to becoming good at statistics, as it is the very foundation of the profession. Skills in math inform the design of algorithms and procedures.
- Technical Skills—a great statistician will have developed agility with computers. They approach applied statistics and quantitative analysis with technology and software tools in mind.
- Flexible—statisticians should be able to evolve as needs within an organization change.
- Collaboration and communication—statisticians should be able to work with a variety of individuals and impart results to leadership accurately and clearly.
The Role of the Statistician in Big Data
There is a good deal of interplay within definitions of “data scientist” and “statistician,” and in some industries, they overlap. The most significant difference, however, is the type of training a statistician receives, which is grounded in theory. This is key among a statistician’s skills—the ability to apply theory to real-life problem-solving for the results to be the most accurate and contain the most useful and meaningful information.
An example of this is one of the more pressing issues with big data—data sets were not necessarily explicitly collected to answer the question the decision-makers or stakeholders may have. That means a lot of the context in the data is lost. A statistician with an understanding of sampling theory will know how to best utilize the data set to obtain a sample that minimizes selection bias and increases variability.
Where it may be too easy to be blinded by the volume of the data, a statistician has the skills to synthesize statistical theory with actual analytical practice and find the meaning in the models the algorithms produce. Among statistician skills, is the ability to use a model based on the big data and create inferences from it to solve the business problems at hand. As well, a great statistician uses their collaborative capability to work with those who have contextual knowledge of the data, such as health professionals, lawyers, marketers, etc.
Also central to becoming a great statistician in a data-centric world is being a leader. The statistician guides the way in:
- Framing the business question.
- Working in teams (e.g., subject matter experts, developers, stakeholders).
- Using the complex data sets to the most significant advantage in terms of providing the needed insights for growth and quality improvement.
Statisticians collaborate with others in the world of big data to produce remarkable innovations. Consider, for example, the human genome project, more accurate weather forecasting, and predictions, and market research that targets the right consumer at the right time.
How to Become a Statistician
If you have an interest in working with big data effectively, developing these statistician skills would be highly beneficial. Fortunately, gaining these skills no longer means having to relocate to attend a particular institution or devote several years. Michigan Tech offers an online Masters in Applied Statistics that covers all of the above areas.
The role of a statistician in a data-centric world is critical and crucial more now than ever before. The statistician's ability to create meaning from the morass will always be the profession's legacy and will only grow with time.