Architects and Scientists

Is the “Architect” in “Analytics Architect”, “Infrastructure Architect” or “Data Architect” analogues to the “Scientist” in “Data Scientist”? Clearly the answer is no. 

“Architect” in a technical job title generally implies that responsibilities include structural design of the information system, like a traditional Architect’s responsibilities include structural design of a building or other physical structure. In these technical job titles the word “Architect” is not intended to be interpreted as a traditional Architect. Traditional Architects’ opinions on this seem to range from insult to flattery. See this article titled I'm an Architect by Amanda Kolson Hurley at architectmagazine.com.

 The “Scientist” in “Data Scientist” is intended to say that the person is an actual scientist. As I discussed in my first post to this blog, a data scientist generally does not publish their work, an important part of the scientific method; therefore, I am not convinced that the work they do is actually science and so I do not believe that they are actual scientists.

In conclusion, an Analytics Architect is not an actual architect and a Data Scientist is not an actual scientist. The title “Analytics Architect” is not intended to mean that the person is an actual architect. The title “Data Scientist” is intended to mean that the person is an actual scientist. Stopping there is not entirely fair because while an Analytics Architect’s work barely resembles the work of an actual architect, a Data Scientist’s work comes pretty close to being actual science. 

For my next post I will cover a technical topic.

What is a Data Scientist?

According to the Harvard Business Review’s October 2012 article titled, Data Scientist: The Sexiest Job of the 21st Century, a data Scientist is:

A high-ranking professional with the training and curiosity to make discoveries in the world of big data.

From my understanding, these are the skills a data scientist must have: expertise in data technology, statistics and probability, and business. A data scientist must also be a relentless problem solver driven by an entrepreneur-like passion to improve the world by understanding it.

I listened to some podcasts recently that seemed to give somewhat varying emphasis within their definitions.

Randy Zwitch, Sr. Data Scientist at Keystone Solution described the continuum between an Analyst and a Data Scientist on the company’s podcast, Podcast #2 – R U Serious?. In his explanation, Mr. Zwitch emphasizes the Data Scientist’s knowledge of technology and problem solving abilities.

In the episode titled Rebuffing "Buyer Beware"Attitude on Data Science from IBM’s The Big Data Hub podcast, the data scientist David Smith, vice president of Marketing and Community at Revolution Analytics, emphasizes statistics.

David Smith does talk about the need to know the statistical platform and language R; however, Randy Zwitch expands the technical skill requirements.

What is a Scientist? Is a Data Scientist a true Scientist?

The past week I was thinking and reading about the definition of a scientist. The title “scientist” is not like “attorney" or “police officer” which have clear legal definitions. Therefore, anyone can really apply their own definition. For example: Christian Scientists. Another example might be someone who called themselves a homeopathic scientist. I, for one, would not consider either of these examples scientists but there is nothing stopping the word “scientist” from being used in these ways. That being said, I want to explore the working definition(s) of “scientist” that most scientifically literate people should be able to agree on.

The online dictionary entries that I found all essentially defined a scientist as a person who is an expert in or is studying one of the natural or physical sciences. This definition makes me uncomfortable because it doesn’t necessitate that the person be doing science on a regular basis. Here is a good discussion related to this subject http://www.physicsforums.com/showthread.php?t=515662.

Wikipedia basically breaks it down into two definitions:

Broad definition: person engaging in a systematic activity to acquire knowledge.

Narrow definition: person who uses the scientific method.

I have trouble with the broad definition simply because if a person is not using the scientific method, are they really doing science? At this point, I like the following definition:

Scientist: A person who does science on a daily basis.

The question then becomes: what does it mean to be doing science? The answer has to be: following the scientific method. Therefore, I have to conclude that Wikipedia’s narrower definition is the correct definition of a scientist.

I was thinking about the definition of a scientist because I was trying to answer the question: Is a data scientist a true scientist? My next post will be on the definition of a data scientist. For now, I’ll just say: no, a data scientist is not a real scientist. Please comment if you don't agree but a data scientist generally does not practice full disclosure, an important part of the scientific method. Full disclosure being the sharing of data and methodology. This is because a data scientist's work is usually corporate intellectual property.