We have expertise in data issues throughout the research life-cycle and can provide assistance on an array of topics, including project planning, data collection and measurement, data management, documentation, and analysis.
WE SUPPORT RESEARCH IN A VARIETY OF WAYS:
We provide expertise in data management and analysis by partnering with research projects, either on an ongoing basis or to support a specific, short-term task. We can help you create data plans, assess your data management needs, and identify data management solutions. We are also available to recommend best practices for keeping your data usable, now and into the future.
We provide consultation and assistance with data management including, but not limited to, merging data sets, performing de-identification and deductive disclosure risk assessment, creating data documentation, and creating scales and subscales.
We also consult and develop codebooks, data dictionaries, metadata, and other documentation necessary to understand data variables and constructed variables as well as the research study process.
Need help conducting data analysis or want to discuss writing data analysis plans?
Our staff can help operationalize research questions into testable models. We have knowledge in model development, statistical procedures such as frequencies, crosstabs, and regressions as well as social network analysis (SNA), structural equation modeling (SEM), and multilevel modeling (MLM). We can also help with problems using statistical software, including SPSS, SAS, and R.
If you have received funding for a new project, but are not sure where to start, our staff can help get you up and running.
We can help with thinking about what data is to be collected and issues to consider when preparing to collect, manage, and analyze data, including determining roles and responsibilities within the project. This includes components of data management plans, but also broader concerns of data sharing and reuse.
We also support data collection through measurement development and refinement. We have expertise in survey design and statistical analysis of data to determine measurement effectiveness.
Join us for our workshops on different aspects of the data management process. We present ongoing workshops at SSRI offering overviews of different aspects of the data management process. Slides from our data workshops are available in the workshop descriptions.
This three-session workshop offers an overview of the SAS programming language, focusing on data management activities.
Session 1 is a general overview of SAS with a focus on major SAS components (Program Editor, Log and Output), core concepts of SAS programming (DATA and PROC), and issues of importing/exporting, reading, and writing SAS data sets.
Session 2 focuses on data modification, including variable creation and variable recoding, as well as adding to and subsettting from particular data sets. Key SAS statements described include ATTRIB, SET, WHERE, IF-THEN, MERGE.
Session 3 focuses on data analysis, specifically descriptive analyses typically used in data management. Key SAS statements described include PROC CONTENTS, PROC MEANS, and PROC FREQ.
We have worked with the American Association of Gifted Children (AAGC) and the North Carolina Department of Instruction on archiving the data collected for the Bright Idea II project (2004–2009). Funded by the U.S. Department of Education, its goals were to help nurture and develop the interest and abilities of young children in underrepresented groups and to help these groups to become eligible for gifted programs in the school. This project includes data at several different levels — student, teacher, school, and district. The student data includes approximately 8,000 students from 28 schools in 11 districts. The completed project includes two data sets that can be linked or used independently, one on teacher data and one on student data.
A long-standing partnership between the Vice Provost for Academic Affairs, the Center for Instructional Technology, and SSRI focuses on data management and analysis of the Coursera data at Duke. Since 2012, Duke has conducted 74 session-based courses, representing 29 unique course areas. Across these 74 classes, over 3 million people have enrolled in courses and 76,913 have completed these classes. In 2015, Coursera shifted from a session-based fixed term course schedule to an on-demand platform, with 25 courses offered. Across this set of courses, over 500,000 have registered and 17,000 have completed. SSRI has played an integral role in the Coursera leadership at Duke, sitting on the Faculty Online Assessment Committee and presenting Coursera results to key constituents throughout the University.
During the summer of 2015, we partnered with Data+ to have an undergraduate student team focusing on Coursera data and the feasibility of making the data widely available for a DataFest. We supported the project through defining the project domains, handling the day-to-day mentoring and coaching of the undergraduate students, and assisting with products for the program including a final poster presentation and several other formal presentations. \
Avshalom Caspi & Terrie Moffitt’s research team deposited the Environmental-Risk (E-Risk) Longitudinal Twin Study research data with SSRI in 2015 in order to allow Duke researchers to use the data securely. These two Duke researchers, along with a researcher in the United Kingdom, Louise Arseneault, developed the E-Risk study in the 1990s to build knowledge about children’s behavioral development and mental health. Our staff supported the sharing of this data and spearheaded the creation of a request for projects that will conduct secondary analyses using the E-Risk Longitudinal Study data.
We have worked with Duke’s Global Education Office for Undergraduates (GEO) on analyzing their current student survey, with an eye toward improving future assessments. GEO collected student exit surveys from Duke undergraduates who attended any global education program between Fall 2011 and Spring 2014 and from undergraduates at other colleges and universities who had attended a Duke study-abroad program during the same time frame. Almost 2,000 students completed the survey. SSRI provided assistance in managing and analyzing the data that GEO had collected in a Qualtrics survey. This included importing the data into SAS, performing data cleaning tasks, coding program elements, and creating de-identified data sets for future use. We also created codebooks for each survey. Data analysis included basic descriptive analyses, focusing on elements that could be changed or improved in the current assessment. We provided GEO staff with a report of our findings and gave a presentation to the staff on our work with action points for future assessments.
Voices Together completed a two-year music therapy program with young adults living with autism in 2015, funded by the Ireland Family Foundation. An evaluation was conducted to determine if and how the program was effective in this population and included demographic information, behavioral coding of program sessions, behavioral assessments using standardized instruments, and behavior ratings by job supervisors. We supported this project by constructing an analysis plan, implementing the analysis, interpreting the results, and writing the evaluation analysis with the input and support of program leaders.
The Resilience Project is a collaboration between four colleges and universities in North and South Carolina that seeks to understand the sources of stress and anxiety for incoming college students as well as factors that help students to thrive despite adversity. Students at Duke University, Davidson College, Furman University, and Johnson C. Smith University are asked to share information on a variety of topics relevant to college life, including social interactions, physical and mental health, well-being and academics. The information gained through this process will be used to design interventions at each campus aimed at enhancing students’ resiliency. SSRI provides assistance in managing sensitive data that is collected by the Resilience Project, including tasks like importing data, performing initial data cleaning tasks, and creating de-identified data sets. We provide expertise on creating data documentation and setting up file structures and permissions, and we facilitate communication between the project team and SSRI IT regarding technical aspects of the Protected Research Data Network (PRDN).
Science and Society initiated a faculty fellowship in 2014 focused on science communication. During the semester-long program, faculty learn how to communicate about their research and influence decision-making on key issues. In support of this project, we constructed a coding framework for the pre- and post- video interviews of faculty participants with consultation from program leaders. The coding framework included micro codes which were assessed for every question and macro codes that were assessed across all of the questions overall. Once the framework was agreed upon, we implemented video coding of the taped interviews, analysis, and interpretation of the codes. Upon completion of the analysis, we created a report and presented it to our partners.