In looking at the preliminary results of his salary survey, Brent Ozar noticed that female salaries seemed to be lower than males. The post focused on a simple analysis of data, the kind that many of us have done in our organizations. We will look at some data, notice some anomalies, and produce some report that allows others to look deeper or make a decision.
The examination of the salary data was taken a bit further by Eugene Meidinger, who wrote a post that started the title with “Practicing Statistics“. His analysis was interesting, not trying to determine the reasons or causes, but just decide if there were some significant patterns in the data. Whether you think there are or aren’t issues with salary, I thought the statistical analysis was done well, and really a basis for what I would call data science.
Data Science is a hot topic right now in many organizations. In fact, for the last year, quite a few organizations are trying to incorporate more data science into their applications, and the hiring of “data scientists” is rising, with higher salaries being paid. There are various definitions of the practice, with none being standardized. There are many curriculums out there from colleges, and even one from Microsoft. In fact, quite a few SQL Server people have completed that coursework.
Data Science seems to mean many things, which is both good and bad. Like being a DBA, there is a lot of room for interpretation and quite a bit of variance in what we may do as a job. I’ve often made a good living as a DBA, being slightly out of the normal reporting structure, having autonomy at work, and usually able to make a difference to a variety of groups. However, I’ve also found that many places don’t want to hire a DBA or don’t think they need one, preferring to let some Windows admin or developer perform those duties. Microsoft doesn’t have DBAs as a job title inside the company, usually using IT Ops staff for those duties. That means it can be hard to find a job at times if many companies don’t think they need that position, which might be the case for data scientists as well.
I prefer to think positively, that data science and data scientist positions are going to grow and be profitable for some of us. We’ll need to learn to have some statistical basis for our analysis, and certainly regularly improve our knowledge of the tools for things like machine learning, but we will find ways to perform data analysis that’s beyond what most business people would complete in Excel. I think we’ll find that those of us that work with data analysis will have lots of opportunities in the future, no matter what we’re called.