Hi Todd,
I have a master's degree and work in business. My title is Statistical Software Engineer. The development group at my company is comprised of about 15 people, however I'm the only one with statistical experience.
While I don't have a "dedicated mentor" per se, the statistician working for our company previously (now retired) does allow me to contact him to bounce ideas off of and to point me in the right direction. I should mention that all of the statistical measures that I add to our application are things that have been well researched and published in respectable journals. I always look for publications that have datasets that I can use to check my code. Most of my statistical work involves new statistical measures that they want added to our consumer software. I research quite a bit before beginning to write the actual functions. I would say that most of my time is spent coding, unless a new measure is being added to our package (for example, more normality testing). For those, I would say that I spend up to two weeks researching the ideas thoroughly before starting to write the code (although usually not that long). Not only do I want to make sure that I fully understand the measurement itself, but I also read variations and spend a good deal of time thinking about the best way to write an algorithm to use in our code.
I prefer to code in C#.NET, but our software is written in VB.NET. I also know Python and R. I also have several more senior developers who help me hone my code (but they aren't familiar with statistics enough to check my algorithms). I think having another developer or two is critical. You can use them as a sounding board, and sometimes they'll have a better way to do what you want to accomplish.
I'm usually assigned one or two projects at a time. My managers treat me like an adult, assuming that I will be responsible for my own work. Many times, I write a part of a larger project, but it will be the entire encapsulated portion. If I have any issues, again, I have other developers to help me come up with solutions. My manager is the one who decides what project we should work on next. He'll share the thought process behind it. Sometimes he'll ask for my input, but it's mostly, "go do this," or "go research that", or "we're going to be working on ____, so start thinking about it." It's hard to say how long it takes to write/complete each project. It took me three months to write a DLL, from start to finish, but the next one I wrote took me less than a month.
As I mentioned, I write algorithms and new functionality, along with DLLs to be used as needed. Many times, I work on the UI to add accessibility to our new functionality, but the primary UI is handled by someone else. And, of course, I research. If there is nothing that I need to code at any point, I'll start researching something that I think might be useful to add to our software package.
I'm lucky to be allowed to work from home whenever I desire, but I have all my research and books in my office, so I usually work there. Unfortunately many companies do not allow you to work from home, even if you only write code all day. I don't know why.
Best advice.... learn, learn, learn, read, read, read. I subscribe to a bunch of journals. Most of the stuff in them I don't understand nor need for my work, but SOMETIMES I run across something that sounds either interesting or applicable to one of my new projects. I'm not saying read all the journals, but read through the titles or the articles and see if anything jumps out. I really think people in these types of jobs need to constantly be learning, or at least keeping up with the time.
All that aside, I started working as an "analyst" right out of school. Luckily I was soon allowed to start writing software, and all the companies that I worked for wrote software to do some sort of analysis or reporting for others to consume. This really helped me work more with algorithms and large amounts of data.
I hope this helps.
------------------------------
Jessica Walls
Northwest Analytics, Inc.
------------------------------
Original Message:
Sent: 07-10-2019 21:13
From: K Todd
Subject: What Does a Recent Graduate do in their First Few Years?
Hi All!
I'm curious to know what new graduates of Statistics or Biostistics are experiencing in their first year out of school; let's say those graduated since 2015, just to put some parameters up. I'm the youngest statistician in my department by ~30 years, so I'm reaching out to the ASA community to explore what it is to be new in the field of statistics.
Maybe this'll be a little educational to those long-since graduated and to those who have yet to leave school!
- What is your title, and are you in academia, government, or business? Do you have a BS, MS, or PhD?
- How large is your group in total, or in terms of statisticians/analysts/data mangers? Do you have a dedicated mentor or someone that you can ask questions of?
- How many and which coding languages do you use? What is your primary? How much of your time is spent programming? Does someone look over your work, or are you expected to be competent enough to be independent?
- How independent are you on each project; are there none, one, two or more statisticians on the same project that you get to work with? Are you observing their decision making process, actively participating, or expected to be a lone wolf on your project(s)?
- What are common roles you are asked to fill?
- Assuming you are working on more than one project, how many are you assigned? On average (I know each project can differ largely), how fast is your turn over rate?
- Do you work remotely, or on site?
- What's the best advice you've been given in your first few years?
If you have other things you'd like to add in, please do! This is not a comprehensive list.
Thank you so much!
K Todd