Communicating with Data Analysis

Data visualization can be a powerful tool for actuaries R. Dale Hall December 2016/January 2017

A “data avalanche” is one way to describe the vast amounts of data we have available to do our actuarial work every day. With so much data out there, how do we make it stand out to our clients and employers? The key is for us to refine our data visualization.

I have to admit that I’m a visual learner. I find that images help me understand concepts and can be quite powerful when used in conjunction with technical data. Actuaries were among the first to graph, chart and visually display results. Today, with high-powered computers and growing analytics software, actuaries have an opportunity to ensure the work is more understandable and easy to communicate. With this in mind, we created our recent Society of Actuaries (SOA) essay collection to provide examples of how helpful the data visualization tools are for actuaries.

Good Research Reads

MORTALITY IMPROVEMENT SCALE MP-2016

The Society of Actuaries (SOA) released its annually updated mortality improvement scale for pension plans, MP-2016. It incorporates three additional years of Social Security Administration data on U.S. population mortality.

PREDICTIVE ANALYTICS PAPER COLLECTION

The SOA released a new collection of papers providing examples of how actuaries are using predictive modeling in practice, such as with the random forest algorithm and machine learning.

MODELING LONG-TERM HEALTH CARE COST TRENDS

The SOA continues to make annual updates to the model originally developed by Thomas E. Getzen of Temple University. The SOA has updated this Getzen Model of long-run medical cost trends and also has developed a document describing practical issues for actuaries using the model.

As part of the essay collection, John Hegstrom, FSA, MAAA, developed an article on the power of effective communication, where he looks at the intersection of design and data science in telling a story. He notes how “data needs to be analyzed and communicated in an efficient and effective manner.” I find his examples of data representation with violin plots, strip charts, density plots and histograms to be quite useful.

The article from Jen Houng Lie, FSA, CERA, covers how actuaries can apply data visualization with predictive underwriting, asset allocation and stochastic modeling through interactive adjustments of charts. The author notes: “Communicating with data visualizations is not just about being more effective by replacing text and numbers. A thoughtfully crafted visualization increases our understanding of the data.”

An article from Kailan Shang, FSA, CFA, PRM, SCJP, identifies how social network data can be used with more traditional measures of statistics. The article focuses on the different forms of visualizing social network data, such as through word clouds and geolocation. Making the data understandable in terms of a word cloud can help communicate the importance of particular phrases and concepts that otherwise may go unnoticed.

As actuaries, we spend a lot of time gathering, testing and analyzing data, and we can use data visualization to tell a powerful story.

I encourage you to visit SOA.org for updates on our ongoing research projects and new proposals. Additionally, visit SOA Engage. It is our online community where you can comment on the latest ideas and research, share your perspectives and more.

R. Dale Hall, FSA, CERA, MAAA, is managing director of Research at the Society of Actuaries.