During the recent Kalido webinar on data science, I was asked a number of questions about data science, which have since been published as a Kalido Expert View. Here's my take on the first question:
Q: In your opinion, what is a data scientist?
I find it most useful to explain what a data scientist is in the context of data scientist vs. statistician. I was trained as a statistician and used to introduce myself as one, but the image most people got was
that I counted runs in baseball or cricket. That’s obviously not what a data scientist does. Data science is more like the next evolution of statistics. Rather than being a reactive type of process, it’s
very consultative and you work with people all across the organization — IT to prepare the data, the business units to figure out what the problem is, operations to actually deliver the results to the organization. A data scientist will start off with a problem the business has and actually go out and look for the data. We don’t start off with a nice, clean data file. Instead, we’re tasked with figuring out what insight exists in a messy data source and finding the results.
It’s a forward-looking process. We’re actually using statistical techniques and data science applications to predict the future and to help business users answer specific questions. So we’re shifting from questions around, “what happened in the data?” to “what will happen based on this data analysis? If we change that variable, what will the outcome be?”
You can find my take on the other questions,
Q: What are the top two or three characteristics you would look for when trying to hire a data scientist?
Q: How do you see the data science “method” evolving?
at the link below.
Kalido Expert View: Three Questions for David Smith VP Marketing, Revolution Analytics
Comments