Home / Faculty Club / Best Lessons / Make Big Data Relatable by Analyzing U.S. Unemployment

Make Big Data Relatable by Analyzing U.S. Unemployment

Employed or unemployed? It is not that simple. David Powell, MS, uses data from the Bureau of Labor Statistics to demystify statistical analysis.

Educator

David Powell, MS, MBA

Adjunct Instructor of Mathematics/Statistics, Berkeley City College in California

MBA, MS in Industrial Engineering and Operations Research, BA in Applied Mathematics

The nature of employment has changed dramatically in the past few decades, particularly with the growth of the “gig economy,” but the way the government measures unemployment has not quite kept pace. This leads to substantial confusion, says David Powell, MS, an adjunct professor at Berkeley City College in California. It also provides the perfect backdrop for discussion in his Introduction to Statistics course.

“In this age of Big Data and news overload, students must be comfortable with basic statistical analysis and reasoning,” Powell says. “And at many schools, such as BCC, statistics courses are a requirement for more majors than ever before, including sociology, psychology, education, and business.”

A substantial portion of Powell’s class is devoted to projects and activities that facilitate the understanding of unemployment statistics. The topic is a good one, he says, because public reports from the Bureau of Labor Statistics are issued monthly. The topic is also relatable, particularly for students who will soon graduate and transition to a four-year college and/or reenter the workforce. “Until this class, students think they understand the numbers they read or hear in the news and may take them at face value,” Powell explains. “In my class, they learn that statistics is about insights and analysis, and the end goal of the course is to create educated analysts and consumers.”

Challenge

Statistics are math-heavy—but not absolute

It is no secret: Many students dread statistics courses. “It’s probably due to the nature of the classwork,” says Powell. “Many students struggle with math, and statistics seems to compound the problem.” Whereas in math there is an answer to each problem, statistics is a discipline that requires reasoning about uncertainty, incorporating variables, and understanding context, he explains. “And that can quickly turn coursework into a fast-moving jumble of formulas and tables,” Powell says.

Innovation

Use real projects to make statistics relatable

Including a project-based component to the class—particularly on a hot-button topic with plenty of news coverage, such as U.S. unemployment—engages students much more than rote memorization of formulas, says Powell. “The topic of unemployment also raises some interesting questions for analysis, such as ‘What is the effect of the gig economy on unemployment numbers?’ and ‘How should we change the system for tracking unemployment?’”

While Powell does give traditional lectures along with group activities, he has found that basing the overall course on important, real-world statistics generates real payoff. “Students can then form their own opinions and draw their own conclusions,” he says. “But the opinions are now based on actual data, knowledge, and insights.”

See and share lecture notes, practice tests, and teaching materials.

Get access now

Context

“Until this class, students think they understand the numbers in the news and may take them at face value. In my class they learn that statistics is about insights and analysis, and the end goal of the course is to create educated analysts and consumers of data.”

— David Powell, MS

Course: MATH 13 Introduction to Statistics

Course description: Introduction to theory and practice of statistics: Collecting data: Sampling, observational and experimental studies. Organizing data: Univariate and bivariate tables and graphs, histograms. Describing data: Measures of location, spread, and correlation. Theory: Probability, random variables; binomial and normal distributions. Drawing conclusions from data: Confidence intervals, hypothesis testing, z-tests, t-tests, and chi-square tests; one-way analysis of variance. Regression and nonparametric methods.

Lesson: The statistics of unemployment in the U.S.

Here is a basic outline of how Powell uses unemployment reporting to hit all the key points students need to learn basic statistics and its gathering, analysis, and usage. Powell suggests breaking the coursework into the following steps. (Note: This plan could be easily adapted to suit another set of statistics for a newsworthy topic in another course, such as sociology, for example.)

Establish consistent terminology

“In statistics, there are numerous conceptual and language barriers,” says Powell. “For example, you never ‘accept’ a hypothesis, but you ‘do not reject’ it. And ‘answers’ are often given as ‘confidence intervals,’ which are ranges. For example, if you do polling, you might get an interval that stretches from 49.3% to 51%, which could be a win or a loss!”

For these reasons, Powell leverages explanation and discussion before assigning any readings. In addition to covering statistical terms, he also covers definitions and measurements used by the Bureau of Labor Statistics (BLS). For example, for this assignment he will discuss the language in the BLS’s Current Population Survey, a monthly survey that goes to a sample of households.

Assign readings and then expand to project group discussion

“Now we turn to the gist of this project,” says Powell, who assigns several articles for independent reading and project team discussion: a piece from the BLS website titled “The Current Population Survey: Tracking Unemployment in the United States for over 75 Years,” the Brookings piece “Measuring American Gig Workers Is Difficult, but Essential,” and the Seeking Alpha piece “Tracking Peak Boomers and Associated Peak Employment.” In order to drive and orient the discussion, Powell presents students with a series of “driving questions,” which they answer wrestle with both alone and in their project groups (see sidebar).

Powell’s Driving Questions on the Current Population Survey

These are the “driving questions” that Powell provides to students for consideration as they work through the assigned readings:

  • What was the original purpose of the Current Population Survey (CPS) and why has it been a uniquely useful instrument for more than 75 years?
  • What are some key extensions of the CPS (describe at least two)?
  • What are some of the key statistical concepts that have been further developed by the CPS? For example: Why did the CPS prove more reliable than the Census? How does the CPS handle employment seasonality? Why has there been so much so much controversy over the years about the definition of “unemployment”?
  • How is the CPS related to the ideas of polling?
  • When, how, and why has the CPS come under “political” attack?
  • What were the major “reviews” of the CPS and what changes have resulted?
  • Why were so many “alternate” measures of unemployment introduced and how are they correlated? What (if any) “slant” do you see in the CPS article?
  • What kinds of complications in employment reporting are caused by part-time workers, and is the problem growing?
  • And finally: What is the “gig economy” and how is this related?
Introduce students to a variety of data sets

The next step is for students to find and examine highly relevant economic data. For this, Powell first has them consider two sets of big-picture, macroeconomic data: trends in the gross domestic product (GDP) found on TradingEconomics.com (or other sources), as well as trends in unemployment found on that site’s page The United States Unemployment Rate. Lastly, Powell has them seek alternative data sets—for example, the labor force participation rate (also found on TradingEconomics.com). “The picture that emerges is an apparent conundrum—a decline in labor participation combined with economic growth and declining unemployment rate. The student teams then need to grapple with questions such as ‘What does it all mean? How could these trends coexist?’”

“Overall, the students are taught to combine considerations of the short-term, medium-term and long-term data series and to think about possible hypotheses and conclusions,” says Powell.

Guide them through data analysis

Next Powell encourages students to analyze the data sets and draw insights from them. “This is especially timely because it makes students examine and wrestle with many current issues,” says Powell. For example, politicians have recently claimed that this is the strongest economy ever. “But is that really true?” he asks. The numbers may well not align with students’ own experiences, for example, in Berkeley. “Students can leave the class and see people living on the streets,” Powell says. “It is a lesson in looking at the big picture.”

There are also a variety of theories that the BLS data are somehow manipulated. Powell addresses this by helping students review how the data is gathered, presented, and used.

Weave the theme into other projects

Ongoing group assignments throughout the semester also consist of writing papers and presenting slides about unemployment—as well as tackling several smaller projects. For example, in addition to the larger study of unemployment, Powell might assign a couple of other topics that also lend themselves to statistical analysis, such as election polling.

“Much of the … class can consist of discussion of what might really be going on with the statistical data, and how to use it to generate hypotheses and draw conclusions,” Powell says. In weaving the various topics together, he says, students learn about “taking knowledge that [they] have gained from one project and reusing and expanding [their] knowledge base as [they] work through the other projects.”

Leave a reply

Have you tried this in any classes you teach? Would you like to try it? Share your questions, critiques, comments, and insights below.

What Is Course Hero?

Course Hero is an online learning platform where you can access course-specific study resources contributed by a community of students and educators.

What Is the Faculty Club?

The Faculty Club is a multi-disciplinary community of educators sharing ideas to advance innovation and celebrate excellence in higher education.