Conference paper
Constructing Linear Measures from Rating Scale Data with Rasch Modeling (Presidential Invited Session)
NARST Annual International Conference 2009 "Grand Challenges and Great Opportunities in Science Education" April 17-21 2009 Garden Grove CA, pp.245-245
2009
Abstract
While many science education instruments use rating scale data to assess an attitude, belief, or behavior, the Rasch model can be used to determine the extent to which meaningful linear measures can be constructed from these ordinal data. Construction of linear measures results in a better-targeted instrument, a more refined rating scale, and a better understanding of how precisely different samples of students and teachers can be evaluated. Furthermore, linear measures are rigorously tested against a set of strict specifications, and hence are scientifically defensible. Using the Science Teaching Efficacy Belief Instrument–B (STEBI-B) for Preservice Teachers as an example, we will demonstrate how a variety of Rasch diagnostics can aid in using current rating scales to construct science education measures. Specific focus will be on rating scale categories, targeting and separation, exploration of dimensionality, and construct interpretation.
Metrics
7 Record Views
Details
- Title
- Constructing Linear Measures from Rating Scale Data with Rasch Modeling (Presidential Invited Session)
- Creators
- Christine M Fox - University of ToledoToni Sondergeld - University of Toledo
- Publication Details
- NARST Annual International Conference 2009 "Grand Challenges and Great Opportunities in Science Education" April 17-21 2009 Garden Grove CA, pp.245-245
- Conference
- NARST Annual International Conference 2009, Grand Challenges and Great Opportunities in Science Education (Garden Grove, California, United States, 17 Apr 2009–21 Apr 2009)
- Publisher
- NARST
- Number of pages
- 1
- Resource Type
- Conference paper
- Language
- English
- Academic Unit
- School of Education
- Identifiers
- 991019335603604721