Modeling a Longitudinal Relational Research Data System

<p>A study was conducted to propose a research-based model for a longitudinal data research system that addressed recommendations from a synthesis of literature related to: (1) needs reported by the U.S. Department of Education, (2) the twelve mandatory elements that define federally approved state longitudinal data systems (SLDS), (3) the constraints experienced by seven Midwestern states toward providing access to essential educational and employment data, and (4) constraints reported by experts in data warehousing systems.</p><p>The review of literature investigated U.S. government legislation related to SLDS and protection of personally identifiable information, SLDS design and complexity, repurposing business data warehouse systems for educational outcomes research, and the use of longitudinal research systems for education and employment outcomes. The results were integrated with practitioner experience to derive design objectives and design elements for a model system optimized for longitudinal research. The resulting model incorporated a design-build engineering approach to achieve a cost effective, obsolescence-resistant, and scalable design. The software application has robust security features, is compatible with Macintosh and PC computers, and is capable of two-way live connections with industry standard database hardware and software. Design features included: (1) An inverted formal planning process to connect decision makers and data users to the sources of data through development of local interactive research planning tools, (2) a data processing module that replaced personally identifiable information with a system-generated code to support the use of de-identified disaggregate raw data across tables and agencies in all phases of data storage, retrieval, analysis, visualization, and reporting in compliance with restrictions on disclosure of personally identifiable information, (3) functionality to support complex statistical analysis across data tables using knowledge discovery in databases and data mining techniques, and (4) integrated training for users. The longitudinal research database model demonstrates the result of a top down-bottom up design process which starts with defining strategic and operational planning goals and the data that must be collected and analyzed to support them. The process continues with analyzing and reporting data in a mathematically programmed, fully functional system operated by multiple level users that could be more effective and less costly than repurposed business data warehouse systems.</p>

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Teaching & Learning

Year

2010

Author dc:creator

• Hunt Olsen, Michelle D.

• `

Repository record dc:identifier

https://commons.und.edu/theses/1022

OAI identifier oai:identifier

oai:commons.und.edu:theses-2024