Progress 03/01/17 to 07/31/18
Outputs
Target Audience:The target audiences reached by my training, networking, and research efforts included researchers, business professionals, rural community leaders, and students. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?This sabbtical project greatlyenhanced my professional development by helping me acquire new data science expertise and skills.By completing the online course in Machine Learning taught by Professor Andrew Ng, Stanford University, I developed new optimization, programming, and modeling expertise. This excellent course helped me learn about supervised and unsupervised machine learning approaches and gave me experience with these algorithms using MatLab software. By becoming familiar with diverse applications (e.g., image recognition, email text interpretation/filtering, cluster analysis/customized marketing) of machine learning algorithms through course lectures and assignments, I gained valuable knowledge about the use of these approaches by the public and private sectors. The materials from this course provide a strong roadmap for designing a high-quality course. In short, the first-rate instruction and delivery of the online course provide a strong example for online training in data science. By completing three on-line courses in the Johns Hopkins University Data Science Certificate, I developed several key data science skills. By completing the Data Scientist Toolbox Course, I increased my knowledge of key data science resources, including various R tools and the networking/collaborative Git & Github tools for developing and distributing data science products. In addition, I learned about diverse types of experimental designs and the distinct ways in which data can and cannot inform responses to research questions. Learning about approaches for making causal inference (and the required weight of evidence) from the perspective of multiple disciplines was extremely valuable. The required course in R Programming provided a review of introductory R programming, gave me the skills to prepare my own R packages, and strengthened my knowledge of different R resources and user-communities. The Getting & Cleaning Data course provided a solid introduction to different types of data resources and data science standards for documenting, cleaning, downloading, and manipulating data. Together, these courses gave me a valuable introduction to data science. Further, experience with Git & Github tools prompted me to think about the potential of these collaborative coding and distributed coding tools, many of which are not widely used by applied economists. However, the instruction and materials on these courses were of relatively lower quality and did not provide a strong roadmap for future data science course design or online training. By completing the online Big Data & Social Analytics course offered by MIT and earning a certificate from the MIT School of Architecture and Planning, I gained valuable knowledge of novel big data streams and business and government applications developed based on these data resources. I acquired expertise in Python programming, and joined a global network of professionals interested in acquiring new data science expertise. This challenging course gave me firsthand experience working with big data streams, including multiple types of data collected via smart-phones, and introduced to me both an international network of data scientists and academic research on social dynamics, social processes, and big data. In addition, I learned about numerous open-source tools for collecting and manipulating data and writing and storing programs (e.g., Jupyter Notebook) - resources that will help me and others with future research and instruction. In the course of this project, I encountered much hype around big data and AI,and developed numerous questions about privacy issues and data science. Moving forward, I am encouraged by the scientific advancesand economic & community development that could be supported from greater integration of data science andapplied economics. Newanalyses, new rural data resources, and strategic connections across fields have great potention support educational and business innovations. Nonetheless, many of these do not require big data. Instead, they involve thoughtful and creative strategies for assembling information, datasets, and learning communities. How have the results been disseminated to communities of interest?I disseminated ideas about future rural applications of big data and data science with researchers and business professionals enrolled in the courses I completed as part of this project and when networking at the multiple data science conferences. All of the online courses I completed involved discussions with peers and exchange of ideas. Such exchanges were most productive in the MIT Course on Big Data and Analytics, where the audience for the course were business professionals. I received initial feedback from peers and instructors on how new streams of data or technologies for sharing and visualizing data could support improved decision-making in rural areas. In addition, I shared insights about data science approaches and new data streams with applied economists through mutiple conference presentations (AAEA, W4133) and publications (two in Landscape and Urban Planning). Smart phone-based data streams have potential to change our understanding of individual behaviors and the behaviors and importance of networks of individuals. Broader application of open data science principles and sharing of data and code by applied economists have great potential to accelerate scientific research of rural areas and support decision-making in these areas. I have initiated plans to improve data science curricula aimed at social scientists and rural areas at University of Maine. I have modified my existing applied economics courses to embrace more data science content and share my recent discoveries and training with students. I have also changed the way I train undergraduate and graduate research assistants, placing more emphasis on agile learning and computing skills. Lastly, I have initiated draft papers and other communication materials documenting these insights and advances for other applied economists and plan to disseminate these through conventional professional circles and my university website. These include sharing my course syllabus, a summative white paper on applied economics and data science, focused work integrating scholarship of rural community development and data science, and more informal website content on lessons learned and useful resources for applied economists and other social scientists. By engaging in conversations with community leaders in rural forested communities around workforce, education, and entrepreneurship issues, I shared insights about data science's role in future rural economic development strategies and more broadly as a tool to engage students and leaders in thoughtful discussions of their histories and potential futures. Additional plans for dissemination to these groups include a presentation at a conference of educational and business leaders and a rural education conference. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
Big data and associated analytics and technology present opportunities to advance understanding of rural economies, rural communities, and rural entrepreneurs. Until recently, the phrases big data and rural rarely appeared together; the majority of research insteadfocused on big data and urban areas. Yet, big data and data science more generally have potential to advance understanding of rural areas and support economic and community development in these areas. I completed data science training, developed and formalized new professional collaborations, and shared insights with researchers, professionals, community leaders, and students to explore this potential, advance new research, and initiate new education and training plans. (1) Acquiring data-science expertise by completing formal and informal training I acquired key data science expertise by expanding my computer-programming skills, becoming familiar with new data resources and analytical tools, and interacting with data science researchers and professionals from diverse fields and industries. I gained skills, joined new professional networks, and got feedback from peers about economic applications of data science approaches. I incorporated these insights and skills into active research projects, successfully designed and funded a new research project, and assembled materials to help with future research as well as educational and training initiatives. I successfully completed formal data science training in machine learning (Stanford University), data science (Johns Hopkins University), and big data & social analytics (MIT). I also initiated self-driven training using resources and texts. I purposively assembled a library of useful texts on data science themes and methods to support future research, course and training development, and students. I gained valuable firsthand knowledge of how computing software, languages, and tools (MatLab, R, Python, Github, Jupyter Notebook) and new (big and small) data resources can enhance conventional approaches and resources used by applied economists to study rural areas. My training experiences provide a strong foundation for future course development and helped me better assess the viability of some proposed online training solutions for students as well as rural workers. In summary, my successful completion of formal and informal training accelerated my knowledge of relevant data, software, and methods for studying innovation in rural areas; gave me multiple opportunties to get initial feedback on ideas for applying data science approaches to study rural areas and foster innovation by rural entrepreneurs; and provided a solid foundation for initial research applications and future research and course development plans. (2) Developing and formalizing new professional collaborations to stimulate research innovations by networking strategically and seeking resources to support such collaborations I strengthened my connections with researchers and professionals interested in data science by attending multiple conferences (2017 Open Vis Conference, 2017 Poptech, 2018 Open Data Science East and Accelerate AI) and engaging with contacts at research centers (SESYNC) and then strategically built on such connections to enhance active collaborations focused on the prosperity of rural areas. These events helped me continue my education about data science and data visualization and broadened my research network to include a diverse set of data scientists, entrepreneurs that included individuals from the tech industry, non-government organizations, private businesses, and universities. At all of these events I strategically networked with attendees interested in rural areas & rural entrepreneurs (notably scarce in some instances) and came away with multiple new contacts and ideas for future research. I interacted with key staff at the NSF-funded SESYNC Center about training and networking opportunities and networked with economists and ecologists interested in big data resources and their implications for science, program evaluation, and teaching. Building on these insights, I collaborated with two colleagues at University of Maine on a successful internal grant proposal to support an interdisciplinary research project focused on strengthening the connections between rural schools and rural innovation and prosperity. This project which is supporting three undergraduate students has allowed me to test initial training ideas and to test how improved data vizualization and analytics can help support education, research, and stakeholder engagement. In summary, I learned much from attending conferences and networking with professionals about the growth of data analytics and visualization, the changing modes of communicating data insights, and the value of numerous open-source data science tools. By gaining knowledge of these advances and tools, I developed insights about potential future research, teaching, and business applications; launched a new line of training and research via an undergraduate interdisciplinary project; and initiated strategic collaborations to support future research and training activities. (3) Sharing insights with colleagues and stakeholders to advance research, decision-support, and educational outcomes. I shared insights with researchers, extension staff, students, and community leaders to advance research, decision-support, and education focused on rural areas. I integrated new ideas, tools, and data into the ongoing work of two active USDA NIFA funded research projects. I used new data visualization skills to help promote discussions about estate planning and conservation decisions of family-forest landowners with forest extension staff and advanced a research publication contrasting such decisions in multiple northeast states. I applied new data science skills gained in my online course training to extend ongoing research being conducted on another USDA-funded project focused on the economic trajectories of rural communities. By testing different types of classification approaches, I extended our team's knowledge of the groupings of rural, forested communities in the United States in terms of similar economic and demographic transitions. I initiated research of rural prosperity and economic trajectories at multiple scales (county and sub-county) to enhance understanding of the differences and similarities amongrural economies and assemble useful databases and visualizations related to these themes. Having gained new knowledge of data communication and tracking tools, I started to develop and share ideas about how to improve the evaluation of extension and other program interventions. Improved monitoring and analysis of the impacts of different extension and outreach interventions has the potential to improve decision-support tools aimed at rural actors and outcomes in rural communities. I contributed to multiple manuscripts and conference presentations focused on rural communities that integrated my new data science insights with active project research on rural communities and rural, forestlandowners. I have started to summarize my key takeaways for other applied economists interested in data science and big data to help colleagues find training that is right for them and encourage the development of additional trainings and resources. My work on this project is helping me develop new data science training for social scientists at the University of Maine and has already changed the content and instructional approach I take in my economics courses. By fulfilling these these three project goals, I am advancing new knowledge and understanding of rural communities and entrepreneurs, identifying opportunities for improved decision support, and delineating needs for changing curricula and workforce development.
Publications
- Type:
Journal Articles
Status:
Awaiting Publication
Year Published:
2018
Citation:
Bell, K.P., Markowski-Lindsay, M., Catanzaro, P., and J.E. Leahy. (2018). Family-forest owner decisions, landscape context, and landscape change. Landscape and Urban Planning, DOI: 10.1016/j.landurbplan.2018.08.023.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2018
Citation:
Bell, K.P. (2018). Big data, economics, and rural economies: facilitating innovation and economic opportunities in rural communities. Presentation, accepted for presentation at the 2018 NIFA Agricultural Economics and Rural Communities (AERC) Project Directors Workshop, Washington, DC, August, 08.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2018
Citation:
Bell, K.P., Crandall, M., Munroe, D., Colocousis, C., and A. Morzillo. (2018). Rural Forest-Based Communities, Economic Shocks, & Economic Trajectories. Conference presentation, accepted for presentation at Rural Forest-Based Communities, Economic Shocks, & Economic Trajectories, 2018 Annual AAEA Meeting, Washington, DC, 07 August.
|
Progress 03/01/17 to 02/28/18
Outputs
Target Audience:The target audiences reached by my training, networking, and research efforts included researchers, business professionals, rural community leaders, and students. I engaged directly with these audiences during my formal coursework, at professional conferences, and while conducting field research in rural communities. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?During the reporting period, this sabbtical project offered numerous opportunities for training and professional development. By completingmultiple datascience courses and participatingin informal professional development opportunities at data and tech conferences, I updated my data science skills. These investments in my professional development will support future training of peers, students, and interested professionals. How have the results been disseminated to communities of interest?During the reporting period, I acquired feedback on my initial research ideas from peers and instructors in my online courses and from colleagues on multiple research projects focused on rural communities. At conferences and regional stakeholder events I also engaged with business professionals about workforce needs and the use of data analytical and communication tools by industries and small businesses. In addition, I shared informally my experiences with online training with numerous colleagues at conferences and workshops.Engagement with thesetarget audiences helped improve multipledraft outputs of this project. What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, I will focus on finishing several project outputs documenting my training experiences and research advances and attendadditional data science conferences to strengthen further my research networks and collaborations; participate in "boot-camp", in-person training activities; and get feedback on research on rural communities.
Impacts
What was accomplished under these goals?
(1) Acquiring data-science expertise by completing formal and informal training.Over the reporting period, I acquired key data science expertise by expanding my computer-programming skills, becoming familiar with new data resources and analytical tools, and interacting with data science researchers and professionals from diverse fields and industries. By taking several online courses from three different universities, I gained skills, joined new professional networks, and got feedback from peers about economic applications of data science approaches. By completing the online course in Machine Learning taught by Professor Andrew Ng, Stanford University, I developed new optimization, programming, and modeling expertise.Challenging computer exercises gave me first-hand experience implementing machine learning algorithms using Matlab software. By writing and accessing numerous sample programs using Matlabdsoftware, I established a strong basis for future instruction and research activities. By completing three on-line courses in the Johns Hopkins University Data Science Certificate, I developed several key data science skills. By completing the Data Scientist Toolbox Course, I increased my knowledge of key data science resources, including various R tools and the networking/collaborative Git & Github tools for developing and distributing data science products. The R Programmingcoursegave me the skills to prepare my own R packages, and strengthened my knowledge of different R resources and user-communities. The Getting & Cleaning Data course provided a solid introduction to different types of data resources and data science standards for documenting, cleaning, downloading, and manipulating data. However, the courses were not particularly well taught so I opted not to pursue the fullcertificate. By completing MIT'sBig Data & Social Analytics course and earning a certificate from the MIT School of Architecture and Planning, I gained valuable knowledge of novel big data streams and business and government applications developed based on these data resources. I acquired expertise in Python programming, and joined a global network of professionals interested in acquiring new data science expertise. This course gave me firsthand experience working with big data streams, including multiple types of data collected via smart-phones.I learned about numerous open-source tools for collecting and manipulating data and writing and storing programs - resources that will help with future research and instruction. I also got feedback from peers on ideas about how to apply big data analytics to better support private and public decisions in rural areas. In summary, my successful completion of these online courses accelerated my knowledge of relevant data, software, and methods for studying innovation in rural areas and gave me multiple opportunties to get feedback on ideas for applying data science approaches to study rural areas and foster innovation by rural entrepreneurs. (2) Developing and formalizing new professional collaborations to stimulate research innovations by networking strategically and seeking resources to support such collaborations. During the reporting period, I strengthened my connections with researchers and professionals interested in data science by attending multiple conferences and engaging with contacts at research centers. By attending the 2017 Open Vis Conference in Boston, Massachusetts, I continued my education about data science and data visualization and initiated connections with academic and industry data science contacts. I learned much from the diverse speakers and attendees about the growth of data analytics and visualization, the changing modes of communicating data insights, and the value of numerous open-source data science tools. By gaining knowledge of these open-source tools, I developed insights about potential future research, teaching, and business applications of these tools. By attending, the 2017 Poptech Conference in Camden, Maine, I engaged with a diverse set of innovators that included individuals from the tech industry, non-government organizations, and universities. While this conference attracted a national audience, I strategically networked with other attendees interested in rural areas & rural entrepreneurs and came away with multiple new contacts and ideas for future research. I interacted with key staff at the NSF-funded SESYNC Center about training and networking opportunities and networked with economists and ecologists interested in big data resources and their implications for science, program evaluation, and teaching. Working with two colleagues at University of Maine, we developed and submitted a grant proposal to an internal call aimed at strengthening the connections between rural schools and rural innovation and prosperity. (3) Sharing insights with colleagues and stakeholders to advance research, decision-support, and educational outcomes.During the reporting period, I began drafting documentation of my training experience, current data science training offered by land-grant universitities, possible and key resources for applied economists interested in gaining new data science skills. Reflecting on my own mixed experience with online courses and reviewing data science curricula at other universities, I gained valuable knowledge about data science education and also developed many questions about the use of online tools to train/re-train individuals. These insights are guiding my ideas about needs for changing curricula and workforce development programs. Over the reporting period, I focused on applyingmy new training in data science tocomplement the ongoing work of two active USDA NIFA funded research projects. I applied new data science skills gained in my online course training to extend ongoing research being conducted on another USDA-funded project. By learning about different types of classification approaches and network analyses, I am giving our research team a broader perspective as we study the grouping and interactions of rural, forested communities in the United States. How these communities navigate multiple shocks will significantly influence innovation and entrepreneurship in rural areas. Having gained new knowledge of data communication and tracking tools, I also started to develop ideas about how to improve the evaluation of extension and other program interventions. These insights have the potential to strengthen future activities of a second active USDA-funded research project focused on rural forest landowners. Improved monitoring and analysis of the impacts of different extension and outreach interventions has the potential to improve decision-support tools aimed at rural actors and outcomes in rural communities. During the reporting period, I contributed to multiple manuscripts focused on rural communities that integrated my new data science insights with active project research on rural communities and rural landowners. By fulfilling these these three project objectives, I am advancing new knowledge and understanding of rural communities and entrepreneurs, identifying opportunities for improved decision support, and delineating needs for changing curricula and workforce development.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2017
Citation:
Van Berkel, D.B., Rayfield, B., Martinuzzi, S., Lechowicz, M.J., White, E., Bell, K.P., Colocousis, C.R., Kovacs, K.F., Morzillo, A.T., Munroe, D.K., Parmentier, B., Radeloff, V.C., and B.J. McGill. 2017. Recognizing the sparsely settled forest: Multi-decade socioecological change dynamics and community exemplars. Landscape and Urban Planning, DOI: 10.1016/j.landurbplan.2017.10.009.
|