A Career in Survey Statistics?

Randy ZuWallack is a senior statistician with Abt SRBI, where he specializes in sampling design and data collection methodologies spanning topics such as public health, housing, and transportation. His road to survey statistics started with a BA in mathematics from SUNY-Geneseo and an MS in statistics from UMass-Amherst.

How likely are you to pursue a career in survey statistics? Would you say …

a. Very likely

b. Somewhat likely

c. Not very likely

d. Not at all likely

Wait, before you answer this question, let me tell you a little bit about survey statistics.

In the context of survey research, a “survey” is the process of collecting information about members of a population. Usually, a survey is conducted with a sample of the population, rather than with a census (all members of the population.) Survey administration refers to the methods and operations used to collect the information from the sampled members. We collect information about each individual member (usually a person) in the sample using a questionnaire, or the survey instrument. The collective data from the questionnaires are used to calculate survey statistics. A survey statistician may have a role in selecting the sample, developing the questionnaire, designing the methods and operations for administering the survey, calculating statistics, or any combination of these.

Now that we got that out of the way, let’s find out if survey statistics is the right career for you. Please check the number that best represents how you feel about these statements about a career …

OK, that’s it for the survey questions! But seriously, these are the things I most enjoy about working in survey statistics, so let me tell you what I mean.

Solving ‘Puzzles’

I like to describe survey research simply—asking the right questions to the right people. It sounds easy, but there are many pieces that need to come together to execute a survey.

• Right questions—what measurements (e.g., attitudes, behaviors, knowledge) do we need? How do we ask the questions in an unbiased manner? How do we structure these questions into a cohesive survey?
• Right people—who’s the target population? How can we select a sample that represents this population? What sample size? How do we contact those selected, and how do we get them to respond to our survey?

The decisions we make above are usually constrained by budget and schedule. A survey statistician helps put these pieces together to solve the “puzzle”—finding the most efficient way to ask the right questions to the right people (within budget and on time). It’s not necessarily easy, but it can be fun!

Multidisciplinary Teams

Survey research draws on many fields of study—statistics, cognitive psychology, sociology, computer science, and project management. A survey statistician will work on a team with experts in these fields. As the statistician, the team will look to you for technical expertise (e.g., sample size, power analysis) and also as the authority on sampling validity—is the sampling approach right for the target population?

Learning New Things

Survey research is applied in any number of domains, such as public health, transportation, labor studies, finance, social program evaluation, environmental studies, marketing and product development, wildlife management—the list goes on. As a survey statistician, I am exposed to these areas of research so not only do I learn more about survey statistics, I also get to learn about many other interesting areas of study.

Innovating New Ideas

Survey statisticians explore new methods and technologies to improve data quality and/or reduce cost. We are constantly looking for ways to do things better, faster, cheaper—and we always champion data quality. Our innovations often are spurred by world changes that force us to adapt. New technologies mean new opportunities for alternative ways to reach our target population. Constant change means our profession never goes stale and we are allowed to explore our creative side.

Exciting Real-World Applications

One of my favorite things about being a survey statistician is that my work makes a tangible difference. Smart decisions are supported by good data. We help get that good data. Consider these examples:

• Labor: An employee and employer perspective of the Family and Medical Leave Act (FMLA), including awareness of benefits, experience taking leave and eligibility, sponsored by Department of Labor
• Traffic Safety: A survey to assess the behaviors, attitudes, and perceptions about distracted driving; knowledge of laws; and effectiveness of prevention programs,sponsored by National Highway Traffic and Safety Agency
• Environment: A combination of surveys to measure the number of salt-water anglers, number of fishing trips, and details on their catch, www.st.nmfs.noaa.gov/recreational-fisheries/in-depth/how-anglers-sampled/index, sponsored by National Oceanic and Atmospheric Administration
• Current events: Polls that measure public opinion on national politics, the economy, popular trends, current events, and more—there are many, such as the ABC News/Washington Post poll.

There is any number of examples in which survey data are used to make a difference. To me, there is nothing more exciting than seeing your work used for good!
So now that I’ve told you what I like about being a survey statistician, let me ask that first question again—how likely are you to pursue a career in survey statistics? Would you say …

a. Very likely

b. Somewhat likely

c. Not very likely

d. Not at all likely

That’s great! I’m so glad you are all interested. Now, you might ask (as survey statisticians are prone to do) how you should prepare for a career in survey statistics. Here’s my recommendation:

• Coursework: Of course, math and stats will help prepare you for a career as a survey statistician. But don’t forget, survey research is multidisciplinary. Consider taking courses in human behavior (check psychology, sociology, anthropology courses). One of the major challenges to survey research is getting people to respond to surveys. Understanding the cognitive and social constructs that influence behavior allows us to develop methods to maximize response to our questions.
• Sampling Theory: If your college offers sampling theory, take it! It’s the foundation of everything you’ll need to know about selecting efficient and valid samples. If your current college doesn’t offer it, plan to take it postgrad.
• Statistical Programming: Proficiency in one or more statistical packages will most likely be a required skill listed in job postings. Popular software packages include SAS, SPSS, STATA, and R. Learning skills in one or more of these packages will allow you to contribute on your first day.
• Communication Skills: You’ll be a very popular survey statistician if you can convey complex concepts in simple and clear language. If your college offers one, take a technical writing course. And practice in your everyday communications (e.g., emails, homework assignments). Think about whom you are communicating with and what information is important.

Some universities offer degree programs in survey methodology and will cover most or all of these areas. See the ASA sections website for a list.

I hope I’ve provided you with enough information for you to want to find out more about a career as a survey statistician. There are a number of employment opportunities available in government and the private sector. Visit any survey research company and you are likely to find survey opportunities. Also, don’t forget to check the job opportunities in Amstat News. Finally, if you are interested in learning more about survey research, visit these websites:

• American Statistical Agency, Survey Research Methods Section
• American Association of Public Opinion Research
• Council of American Survey Research Organizations