Coen Jonker – Data Scientist at Mobiquity Europe
Quite often, I find myself answering the following questions. I am asked what a data scientist does, what my job is at Mobiquity. Here is my answer.
Humans need reliable information to make decisions. As a data scientist, I apply scientific methods to datasets to provide usable information for human decision makers. I truly believe that businesses and societies improve tremendously when two things change:
My job is to make these things happen.
Anyone can calculate an average. If you ask someone to summarize a set of numbers they oftentimes will calculate an average. But if you make decisions based on that average, you are doing so based on an assumed model: a bell-curve. My job would then be to explain to you what the implications are of modeling data this way and to teach you how to create reliable and valid information from your data.
Decisions should never be taken on averages alone. They simply do not tell the whole story. Averages disregard the highs and lows, the outliers, and these are oftentimes the most interesting data points.
During my secondment at Schiphol I was asked to calculate the average time it takes to attach a passenger boarding bridge to a plane. Based on sensor information from the bridges it is possible to calculate movement time for each individual bridge attachment.
However, attaching a bridge never takes less than zero seconds. As a data scientist, I know that when a value has an upper or lower time limit, the average often is not a good estimate for the bulk of the data. Most likely, using an average makes the process appear to take longer than it does in reality. I did three things.
First, I started asking questions about the question, because without context it is difficult to judge which indicators can provide valuable information. Second, I showed that the average time to connect a passenger boarding bridge does indeed not provide reliable information about typical process duration. Third, I created a model that does accurately describe the process of attaching a boarding bridge.
Based on this information, Schiphol was able to identify specific bottlenecks and opportunities and launched an innovation project to improve the process. Once this innovation is implemented, I get to do another study to show whether the process improved or not.
Every organisation stores more and more data. In order to summarize these data and create valuable information, proper methods need to be applied. The improper use of averages is just an example, albeit a frequently occurring one. Unfortunately, it is really hard to distinguish nonsensical averages in pie charts from valuable information based on correct scientific data processing.
Data scientists like me help clients to cast a discerning look at their data and how their data relates to decisions and impact in the real world. At Mobiquity, we aim to humanize the digital world. Making great information accessible and intelligible for humans is a major part of that.
And that is what we data scientists do.
See the original blog, click here.
We believe that addressing customer challenges gives you opportunities to delight. Using our proprietary Friction Reports and strong industry expertise, we dig deep into customer sentiment and create action plans that remove engagement roadblocks. The end result is seamless, relevant experiences that your customers will love.