Metrics
9 Record Views
Journal article
The Theil-Sen Slope for High-Stakes Decisions from Progress Monitoring
Behavioral disorders, Vol.37(4), pp.271-280
08/2012
Web of Science ID: WOS:000309751900005
Abstract
More and more, schools are considering the use of progress monitoring data for high-stakes decisions such as special education eligibility, program changes to more restrictive environments, and major changes in educational goals. Those high-stakes types of data-based decisions will need methodological defensibility. Current practice for summarizing progress monitoring data is to use a hand-fit trend lines (for practitioner use) or linear regression (for research). This study critically examines both approaches and compares them to a new nonparametric slope called the Theil-Sen. A field test with 372 published data series compared hand-fit, linear regression, and Theil-Sen slopes against evaluative criteria of power and precision, meeting data assumptions, and agreement with visual judgments. Results indicate promise for Theil-Sen slope in defensible high-stakes decision making.
Related links
Details
- Title
- The Theil-Sen Slope for High-Stakes Decisions from Progress Monitoring
- Publication Details
- Behavioral disorders, Vol.37(4), pp.271-280
- Resource Type
- Journal article
- Publisher
- Council Children Behavioral Disorders
- Number of pages
- 10
- Copyright
- © 2012 Hammill Institute on Disabilities.
- Identifiers
- WOS:000309751900005; 99381340010406600
- Academic Unit
- Teaching, Leadership, and Research; School of Education
- Language
- English