NSF Awards: 1642413
2019 (see original presentation & discussion)
Grades K-6, Grades 6-8, Grades 9-12
In response to numerous requests for information regarding the status of K–12 STEM education in the United States, Horizon Research, Inc. conducted the 2018 National Survey of Science and Mathematics Education (NSSME+). The plus symbol reflects the study’s added emphasis on computer science and engineering, two disciplines that are increasingly prominent in discussions about K–12 STEM education and college and career readiness. This study is the sixth in a series of surveys dating back to a 1977 study commissioned by the National Science Foundation. The 2018 NSSME+ assesses changes over time and provides current data on essential elements of the STEM education system—data that will inform future education policy and practice.
Eric Banilower
Vice President
Thank you for visiting our video. The 2018 NSSME+ provides a rich set of data about science, mathematics, and computer science education across the nation. Both a comprehensive technical report and a highlights report are currently available on our website (http://horizon-research.com/NSSME/), with more reports coming soon. Thinking about the results and the status of K–12 STEM education, what findings are you surprised by? What findings are you pleased by? What are you dismayed by?
Noah Feinstein
Associate Professor
I'm so glad that this survey now includes data on computer science education! I’m really curious about the challenges in recruiting and retaining highly qualified computer science teachers, and about any school- and district-level promising practices that are emerging on this topic.
I’m also curious about how computer science teachers have been located within high schools, especially - are they typically located within science departments? How do they relate to tech ed programs or math?
Eric Banilower
Vice President
Although we don't have data directly related to your questions, the NSSME+ data show that many high school computer science teachers started their teaching careers in other subject areas, particularly mathematics and business. We also found that high school computer science teachers are more likely than their science and mathematics counterparts to not have a teaching credential, which may reflect a tendency among schools to offer computer science classes taught by adjunct or part-time teachers with no credential, or college-level faculty providing dual credit opportunities. Hopefully some school and district participants in this event can share their experiences.
Noah Feinstein
Rabiah Mayas
Associate Director
Thank you for sharing this project - the data set is indeed very rich; I'm just starting to look through the technical report now! In your experience with this and previous iterations of the survey, how are K-12 educators and administrators using the data? Are the findings being used to effect shifts at the classroom and student engagement levels? And in what ways have issues of equity been present in the data sets you've collected?
Eric Banilower
Vice President
Good questions Rabiah. We see the primary audience for the study as researchers and policy makers at various levels of the education system. Although we don't know how teachers and administrators use the results, I can envision a number of ways. For example, using the survey items and national data to reflect on local policies and practices. What are your ideas about how K-12 educators and administrators might use the results?
In regards to equity, although the study wasn't designed specifically to be an equity study, we are able to look at the data by several factors, both at the school level (e.g., community type, SES) and the classroom level e.g., race/ethnicity, prior achievement level of students) that provide insight into equity issues. We included some of these analyses in the technical report, and shared equity-related findings in presentations at NARST and the NCTM Research Conference this year (slides are on our website: http://horizon-research.com/NSSME/2018-nssme/research-products/presentations). We are also in the process of writing a series of reports examining equity issues more deeply and hope to have those completed this summer.
Rabiah Mayas
Associate Director
Thank you for the links to access your NARST and other presentations! I was thinking that the data might be used to justify the need (and budget) for science-specific professional learning for elementary teachers, for example. Particularly as science-focused PD can be more supported by districts due to specialization in high school vs. elementary. Local school councils (which in some cities have a lot of power) are another potential audience - here in Chicago at least, the more progressive LSCs are paying attention to national standards and data toward ensuring that their school is pushing toward stronger teacher and principal performance.
Eric Banilower
Vice President
You're welcome Rabiah. Addressing the quality (and quantity) of science instruction in the elementary grades is a complex issue, especially when one considers the scale of the K-12 education system--there are close to a million elementary teachers responsible for teaching science, the large majority of whom are responsible for teaching all core subjects in self-contained classroom settings. I touched on some of the opportunities and challenges in a recent guest editorial in JSTE, and would be thrilled to hear how others are thinking about this issue. What are the promising approaches for improving elementary science instruction? Which are likely to be scalable?
Anne Kern
Associate Professor
I am curious about what kind of student achievement data is included or correlated with the data sets. Can the data be used to give information about the effectiveness of a group of policies, teacher activities, etc?
Eric Banilower
Vice President
Anne, the short answer is no. It is already extremely challenging to collect the data we do; asking schools and districts to provide student achievement data (which do not exist for all subjects/grades and vary from state to state) would likely result in schools refusing to participate at all. That said, there a number of practices that are associated with better student outcomes and this study provides an opportunity to see the prevalence of and variation in a number of those practices. The NSSME+ data also provide opportunities for the generation of hypotheses that could first be tested in smaller scale studies.
Mia Dubosarsky
Thank you for this important work!
I wonder if you have looked at integrated STEM education (in addition to the individual disciplines) or whether it could be assessed from your data?
Eric Banilower
Vice President
Hi Mia, good question. The surveys have a few questions about the integration of STEM subjects, but it wasn't a major focus of the study. One of the challenges, as pointed out in the 2014 NRC report on STEM Integration, is that a common vocabulary does not yet exist for the various approaches to integrating STEM and that additional research is needed to understand what is effective, for whom, and under what conditions. The lack of a common vocabulary, in particular, makes developing valid and reliable survey items extremely challenging. The surveys for the 2018 NSSME+ were developed in 2017. Has there been convergence toward common understanding of the different approaches, etc. since?
Jennifer Rosato
I'm looking forward to the additional reports that will be coming out! Do you know if the data will be available so that it can be broken down by states? There are a number of states that are looking for data specific to their area as part of computer science education landscape reports. For example, the Google/Gallup reports provided national data and then a page summarizing state data.
Eric Banilower
Vice President
Thanks Jennifer. We will be posting several additional reports later this week that focus on different grade bands and subjects (e.g., the status of elementary science, the status of middle school mathematics).
And we will be making de-identified data available for secondary analysis--we're trying to have these versions of the datasets ready by Spring 2020. However, the NSSME+ wasn't designed to provide state-specific results (doing so would have required a different sample design and a much larger, more expensive study). Although it is feasible to analyze data for some of the larger states (California, New York, Texas, etc.), there are not enough cases from smaller states to do so.
Further posting is closed as the event has ended.