Reflections on The Year of Statistics

​Like many graduate students, I endured statistics. While studying statistics was a great bonding experience with my classmates, I did not recognize at age 22 how important it would be to my career and to life in general. Unfortunately, I saw it only as a means to an end—the end being passing the required courses and moving on. Statistics was hard and seemed foreign to me, so I reasoned that there were “stats people whose brains worked differently and enjoyed this sort of thing,” and then there were “normal people” (I kid—sort of!). I also saw professors who had graduate students do all their statistical analyses for them and further reasoned that there was a division of labor that would excuse me from being a statistical expert. While it turned out to be true that no one expected me to be an expert, I wish someone would have imparted on me that I would regret not learning more while I had the opportunity to take classes and devote time to practicing. Without the continual impetus of coursework, it is a challenge to maintain skills, let alone build new ones.

So, why was I asked to contribute my thoughts to this Year of Statistics feature?

Luckily, NORC recognizes that even employees are who not trained statisticians can appreciate statistics and might have a desire to learn more about this discipline. I was asked to represent those of us in this category. I hope others can identify with my situation and relate to my responses to the questions  below. I also hope that those who are statistically savvy understand that we non-statisticians try our best to make sense of their world!

How many years or what percentage of your career have you utilized statistics as a core element in your work? 

Despite having only basic skills, statistics has been central to my work at NORC. Having been the data delivery / data quality task leader for the previous and current rounds of the Survey of Consumer Finance (SCF), and the sole soul responsible for data quality evaluation on the Youth Villages project, I routinely apply simple but critical statistical methods to understand and search for patterns in the data. Such procedures are paramount in identifying interviewer-level and project-level issues that can be addressed to improve the quality of the data. The SCF’s mantra of “continuous improvement” is greatly aided by the creation of data-driven learning modules that hone in on specific and re-occurring issues in the data. It also goes without saying that the examination of paradata is crucial to this effort. So as you can see, statistics are an integral part of my work.

Did you ever think that statistics would become as popular and as large a part of public discourse as it has become (e.g., Nate Silver as a widely quoted pundit, Moneyball and advanced metrics for baseball and other sports)? Are you surprised by the “moment” statistics is having in the public eye right now?

Overall, I am happy that statistics has become a popular and large part of public discourse. Like many others, I was obsessed with the FiveThirtyEight blog during the 2012 election! I am also a big fan of Moneyball, although that may have as much to do with Brad Pitt as with Billy Beane’s ingenious application of statistics to baseball! 
I do hope, however, that data-driven actions are always combined with critical thinking and common sense. I believe it is a mistake to blindly accept and act on statistical findings without consideration of real people and their real situations. There is no doubt that statistical findings are influential and convincing, but they should always be considered in conjunction with qualitative assessments, an understanding of history, and our knowledge of personal, political, and institutional realities, especially in terms of policy decisions. As a political scientist, I believe that too often the public and policymakers are swayed by opinion polls with faulty methodologies and unbiased “stats.” It is maddening to me when people misrepresent or selectively use findings to further an agenda or generate support for their predetermined conclusion. Like with any powerful tool, statistics needs to be used appropriately and with proper caution. 

Based on your experience and career, what advice would you offer to young people thinking about going into statistics?

Coming from the point of view of a social scientist, and not a person who was inherently interested in “going into statistics,” it is highly possible that you will not like statistics when you begin your coursework, because it is intimidating and hard. My advice is: EMBRACE STATISTICS. Cleaning and preparing data, which are tasks undertaken by nearly all projects at NORC and other survey research institutions, is important and required. Beyond obligatory tasks though, having the ability to explore and make sense (at least to some degree) of data, regardless of whether for work or personal interest, is fun and rewarding. Not to mention that being able to execute your own analyses provides independence and expands the contributions you can make to your project and the literature. 

What are some statistical goals you are still trying to accomplish?

As previously mentioned, I am working to refresh and expand my statistical proficiencies. I have learned a lot in recent years, but I still have a lot of work to do. While I am interested in learning to execute new statistical techniques, I am equally interested in learning how to interpret more complicated statistical procedures. I am motivated to understand how others executed their analyses so that I can fully understand their results and be a more critical examiner of their findings.
- Shannon Nelson, 10.7.13
For more information about Shannon, her work, and her publications, e-mail her.


Web Part Error: A Web Part or Web Form Control on this Page cannot be displayed or imported. The type NORC.Website.WebParts.SubscriptionConfirmation, NORC.Website, Version=, Culture=neutral, PublicKeyToken=22a40bce948dece1 could not be found or it is not registered as safe. Correlation ID: f6d3a29e-22e5-c045-0cfe-277bdebdce2b.