Surveys
Surveys are usually the data source students are most familiar with — because we've all been surveyed. The classics include the U.S. Census, the National Health and Nutrition Examination Survey (NHANES), the Pew Research Center surveys, the European Social Survey (ESS), and the Consumer Expenditure Survey. These are foundational for an enormous amount of social science and applied research.
This survey uses four question formats common in real research instruments: Likert scales, a slider, a ranking task, and multi-select checkboxes. Answer honestly — the result reflects your responses, not a right answer.
I enjoy finding patterns in data.
Learning statistics and math feels rewarding to me.
How many hours per week do you currently spend working with data (spreadsheets, code, charts, etc.)?
Which of these data science applications interest you most? (Select all that apply)
I feel confident writing code to analyze data.
Rank these data science skills from most to least important to learn first (use arrows to reorder).
Use the arrows to reorder.
I would consider a career in data science.
A live survey using four common question formats — Likert scales, a slider, ranking, and multi-select. Notice how each format captures a structurally different kind of data.
Advantages
- Customizable. You control the questions and can tailor them to your specific research question.
- Dual capture. Can gather both qualitative and quantitative data, making them flexible.
Limitations
- Response bias. Responses may be biased, skewed, or simply false. People misremember. People want to seem better than they are.
- Limited reach. You may never hear from the population you're most interested in.
- Subjectivity in interpretation. The same question can be read differently by different respondents.
The delivery method shapes who responds — and therefore what you learn. Online platforms (Qualtrics, Google Forms, SurveyMonkey) are free or low-cost and easy to share, but they introduce selection bias toward internet-accessible, motivated respondents. In-person methods (focus groups, interviews, paper surveys) yield richer detail but bring interviewer bias and higher logistical cost. Mail surveys carry very high non-response bias and aren't generally recommended. Phone surveys can work for semantic analysis but face spam-filtering challenges — most people don't answer unknown numbers anymore.
Why Survey Methodology Shapes What You Think You Know
Every public-health intervention you've ever heard of was informed by survey data at some point. NHANES alone is the foundation under thousands of papers and most U.S. nutritional guidelines. When you read "X% of Americans believe Y," you are reading a survey result — and the choices made by whoever ran that survey (sample, delivery, question wording) are shaping what you think you know.
You want to survey internet usage patterns across a wide demographic range, including elderly and low-income populations. Which survey delivery method is most likely to introduce selection bias against your target population?