Learning from Attempts to Improve School:
The Contribution of Methodological Diversity
By Stephen W. Raudenbush
This article provides an interesting read. It presents itself as a thought piece for education policy makers on the perils in the process of finding what works in improving learning outcomes amongst students.
It proposes two seemingly unconnected claims of policy of what ought to be done in education research. The association becomes obvious when the author cleverly rehashes the statement - ‘questions should drive methods’ several times in the article.
Raudenbush claims that instructional regimes and interventions, not resources, accountability and governance, should be conceived as the causal agents for instructional improvements, and they should be the central themes in education research.
He justifies these by highlighting the marginal impact of resources, accountability and governance on learning outcomes, and warrants that ‘effective instruction is not likely to flow from the exertion of policy levers’. He provides several reasons to permit this linkage. He observes that resources, governance and market forces do not remove the knowledge gaps of educators in conducting professional development for teachers, and in selecting, organising, co-ordinating and enacting effective interventions for improving school-wide learning.
This means that the return of investment on the policy levers may show up when the research questions are about the implementation of instructional regimes and interventions, and keeping others as complementary elements that enhance the intensity of the outcomes.
This points us to the primary argument by Raudenbush. He says that systematic experiment is a better way to evaluate claims about casual effects of the interventions deployed in improving teaching and learning in the classrooms. He suggests complementing randomised experiment with mixed or multiple research methods because randomised experiment, while is the best way to discern causal effects, is not enough to verify these effects alone. He indicates that given the cost and the finite amount of funds available for randomised research, there is a need to identify the instructional regimes and interventions that ‘hold the greatest promise in achieving those outcomes’.
This means that education researchers need to known how an instructional intervention could be implemented prior to a ‘randomize trial of effectiveness’. More importantly, they need to be informed if the selected intervention is the most promising one. Therefore, the author warrants that they need to use a variety of research methods, including small randomised research, to conduct descriptive and correlation studies that could suggest that the intervention will work well with the target students in the given learning settings.
In conclusion, I agree that education researchers need to be aware of two blind spots when conducting randomised experiments, and these are deeply ingrained in the statement - ‘questions should drive methods’. Researchers should establish the true causal agents and identify the right target group and settings for the investigation. All else will inform incorrectly, and Raudenbush has used a number of examples in the article to illustrate the use of methodological diversity to avoid this phenomenal.
The Contribution of Methodological Diversity
By Stephen W. Raudenbush
This article provides an interesting read. It presents itself as a thought piece for education policy makers on the perils in the process of finding what works in improving learning outcomes amongst students.
It proposes two seemingly unconnected claims of policy of what ought to be done in education research. The association becomes obvious when the author cleverly rehashes the statement - ‘questions should drive methods’ several times in the article.
Raudenbush claims that instructional regimes and interventions, not resources, accountability and governance, should be conceived as the causal agents for instructional improvements, and they should be the central themes in education research.
He justifies these by highlighting the marginal impact of resources, accountability and governance on learning outcomes, and warrants that ‘effective instruction is not likely to flow from the exertion of policy levers’. He provides several reasons to permit this linkage. He observes that resources, governance and market forces do not remove the knowledge gaps of educators in conducting professional development for teachers, and in selecting, organising, co-ordinating and enacting effective interventions for improving school-wide learning.
This means that the return of investment on the policy levers may show up when the research questions are about the implementation of instructional regimes and interventions, and keeping others as complementary elements that enhance the intensity of the outcomes.
This points us to the primary argument by Raudenbush. He says that systematic experiment is a better way to evaluate claims about casual effects of the interventions deployed in improving teaching and learning in the classrooms. He suggests complementing randomised experiment with mixed or multiple research methods because randomised experiment, while is the best way to discern causal effects, is not enough to verify these effects alone. He indicates that given the cost and the finite amount of funds available for randomised research, there is a need to identify the instructional regimes and interventions that ‘hold the greatest promise in achieving those outcomes’.
This means that education researchers need to known how an instructional intervention could be implemented prior to a ‘randomize trial of effectiveness’. More importantly, they need to be informed if the selected intervention is the most promising one. Therefore, the author warrants that they need to use a variety of research methods, including small randomised research, to conduct descriptive and correlation studies that could suggest that the intervention will work well with the target students in the given learning settings.
In conclusion, I agree that education researchers need to be aware of two blind spots when conducting randomised experiments, and these are deeply ingrained in the statement - ‘questions should drive methods’. Researchers should establish the true causal agents and identify the right target group and settings for the investigation. All else will inform incorrectly, and Raudenbush has used a number of examples in the article to illustrate the use of methodological diversity to avoid this phenomenal.
This report was 1st created on 24 June 2008.
Copyright 2008. Anthony Mok. All Rights Reserved.
0 comments:
Post a Comment