Measuring progress. Inspiring action.

Ask A Researcher

October 2013

New STEM section: A research-based framework for action

Caryn Mohr

Minnesota Compass is pleased to introduce a new STEM (science, technology, engineering, math) section that was developed in partnership with Boston Scientific and with the guidance of an advisory committee, led by Paul Mattessich, Wilder Research Executive Director and Minnesota Compass Project Director; Marilee Grant, Boston Scientific Community Relations Director; Rose Chu, Interim Dean, School of Urban Education, Metropolitan State University; Margaret Anderson Kelliher, Minnesota High Tech Association President and CEO; and Doug Paulson, STEM Specialist, Minnesota Department of Education.

Caryn Mohr, lead researcher for the project, answers FAQs about the project, its implementation, and examples of what the information can tell us.

Why did you focus on STEM?
Skills in science, technology, engineering, and math (STEM) enable individuals to be productive workers and engaged citizens. They drive innovation and are required in some of our fastest-growing occupations. Even those who pursue jobs outside of STEM fields benefit from the problem-solving skills, technological literacy, scientific reasoning, and mathematical skills developed through STEM education. Competence and confidence in STEM skills is important in many aspects of 21st century life.

How were STEM key measures chosen?
This work reflects cross-sector leadership and input. A large advisory committee of STEM stakeholders helped us develop the cradle-to-career framework and identify key measures. We also reviewed research literature to gather background information on important benchmarks to consider. Ultimately, potential key measures were vetted by criteria designed to ensure their scientific integrity and relevance to project goals.

Advisors wanted the framework to reflect school-based as well as out-of-school factors that contribute to developing STEM skills and interest. They also valued understanding system “inputs” which may provide levers for action. In addition to measures of competency, you see measures reflecting interest, classroom time spent on science, teachers’ licensure for the assignment, and participation in science activities not for schoolwork. This combination provides meaningful context for interpretation.

Consider our key measure from ACT, for example. This measure tells us that only 39 percent of Minnesota high school graduates met both the math and science college-readiness benchmarks in 2012. There’s more to it than addressing skill gaps, however. Of those who were able in STEM, only 11 percent were interested in pursuing STEM based on confidence in their choice of a STEM major. At the same time, almost a third of students interested in pursuing STEM majors lacked the ability – meeting neither the math nor science college-readiness benchmark. We have gaps in developing STEM skills among those with interest, as well as gaps in interest among those with skills in these areas. What’s particularly troubling is that a much higher proportion of white students than students of color met math and science college-readiness benchmarks, but a higher proportion of students of color indicated interest in pursuing a STEM major. Research literature presented on the site tells us that both interest in a STEM career and proficiency in STEM subjects are necessary prerequisites for students to select and succeed in a STEM major. The Disparities section sheds light on opportunity gaps that underrepresented students face.

Were there any suggested measures that could not be included?
Yes, we considered a large number of potential measures, and in some cases an adequate data source could not be identified at this time. The About the Project page provides background documents prepared for the committee which looked into the possibilities raised. We selected key measures from existing data sources which met our criteria and reflect important markers along the continuum.

Through this process, we found that the project also helps highlight data needs which may inform future research. For example, advisors raised the importance of reflecting integrated STEM education and the impact of out-of-school time activities. These concepts are reflected in the project logic model and research summaries provided on the site, but there was not an adequate statewide data source for key measures in these areas at this time. In a couple of years, we will reconvene an advisory committee to revisit the key measures and consider any changes that should be made based on data available at that time.

How did you define STEM?
Definitions of STEM vary, even among federal agencies. In preparing our postsecondary and employment key measures, we needed to determine which fields would be categorized as STEM for purposes of these measures.

We started with a recommendation for defining STEM occupations developed by a federal committee to enhance comparability of data across statistical agencies studying the STEM workforce. The Standard Occupational Classification (SOC) Policy Committee recommendation identifies two major STEM domains: (1) science, engineering, mathematics, and information technology domain, and (2) science- and engineering-related domain, which includes architecture and health occupations. We included both domains in our STEM groupings, and grouped data in such a way that stakeholders can look at distinctive STEM fields.

This gave us an inclusive definition, but it still omitted some production and trade occupations that were important to our stakeholders. Our advisory committee iterated the importance of reflecting multiple pathways in STEM in our framework, and a recent Brookings report shed light on the extent to which blue collar or technical jobs require substantial STEM knowledge. Based on this, we added a production and trade occupational grouping. “Data & notes” pages for our employment key measures link to a detailed list of the jobs included in each occupational grouping.

Is the data user-friendly?
Yes! Compass provides reliable data in easily accessible graphs. Those interested in more detail can view the numbers behind each graph by clicking on “Data & notes” under the “View” tab.

We also worked hard to provide STEM data in clear, comparable categories. Guided by SOC groupings, we identified 10 distinctive STEM occupational groupings among the fields included in our definition.

We also felt it was important to be able to compare the employment measures with the postsecondary key measure, given stakeholders’ interest in understanding the extent to which degrees and certificates awarded align with workforce needs. To help with these comparisons, we used the same 10 groupings for our postsecondary and employment key measures. Because degrees and awards are categorized based on the Classification of Instructional Programs (CIP), we used a crosswalk developed for federal statistical agencies to align CIP with SOC codes to the extent possible. As with the employment measures, “Data & notes” pages for our postsecondary key measure link to a detailed list of STEM fields of study included in each grouping.

Are there opportunity and achievement gaps in STEM?
Yes, and providing data and research to explore these issues is an important goal of the site. In addition to providing data on gaps in proficiency, the site provides resources for exploring disparities that contribute to those gaps. Key measures can be viewed for different demographic subgroups, including by income status, gender, and race/ethnicity. We also provide information on disparities in STEM from research literature. We look forward to diving into these issues in greater depth in upcoming white papers.
I noticed that some postsecondary and employment graphs provide raw numbers instead of percentages. Why is that?
In some cases the key measure is the number of STEM degrees awarded or the number of jobs, rather than the proportion of all degrees or jobs that are STEM. The point is to support educational and workforce needs in STEM, and not to compare the slice of the pie which is STEM vs. not-STEM. It’s also important to recognize the extent to which STEM skills are used in fields that may not be categorized as STEM.

What’s coming up next for the STEM site?
We will continually update the site as new resources and data for each key measure become available. Over the next several months, we look forward to developing white papers which dive into several topics in depth, including disparities. Check the What’s New section of the site for updates.

We also look forward to supporting ongoing dialogue and use of the data in a variety of ways. We encourage you to follow our Twitter feed and Facebook page to stay involved, and use the #CompassSTEM hashtag to continue the conversation. We would love to hear how you use the resources available on the site.

Caryn Mohr can be reached at Follow her on Twitter @carmohr

STEM section

STEM (science, technology, engineering, math) section in Education topic: Find data, benchmarks, and best practices.

Featured trend

Primary refugee arrivals in Minnesota

Minnesota sees smallest number of refugee arrivals in more than a decade

Last year marked the smallest number of primary refugee arrivals in Minnesota over the last 17 years. About 670 refugees resettled in Minnesota in 2018, nearly half originally from Burma. Primary refugees are individuals who arrive directly in Minnesota from a country of asylum or refugee camp, while secondary refugees (not included in calculations) are those who migrate to Minnesota after arriving in a different state of resettlement.


Learn more about Minnesota’s immigrant population.

Data Update

Our state continues to see improvements in on-time high school graduation. Eighty-three percent of high school students graduated within four years in 2018, up from 78 percent in 2012.

Our Mexican-born population remains the largest immigrant community in Minnesota. Hmong, Somali, and Indian immigrants are tied for our second largest immigrant communities.

Hennepin County remains the most populous county in Minnesota, home to more than double the number of residents of any other county in the state.

Every county in the state has seen a 3 to 7 percentage point decline in the share of residents lacking health insurance since 2013, the first year of full implementation of the Affordable Care Act.