Computational Literacy

A keynote address given to the educational track of the SuperComputing '07 conference. This was a discussion of some preliminary results of a 3-state study on computational science education.

Transcript

1. Chad Thompson The Krell Institute Computational Literacy Friday, January 11, 13
2. A non-profit organization focused on Computational Science in the classroom, workforce and academic community. • Graduate Fellowships • Conference/Workshop Planning • Web and Print Publications .... and research projects! (We’re also located in Ames, IA!) Friday, January 11, 13
3. What Is “Computational Science?” Computational Science is an interdisciplinary field that focuses on using mathematical, numerical and statistical methods to build models that allow us to understand scientific or engineering problems. Computational Science can also be understand as a ‘third mode’ of science in addition to theory and experiment. Computational Science is also focused on problem solving and problem-based Friday, January 11, 13
4. Crystallizing Molten Metals (Tantalum) Molten tantalum is similar in properties to molten plutonium. http://www2.krellinst.org/csgf/deixis/2006/stories/doe_research/llnl/llnl2.shtml Making Impossible Experiments Possible Friday, January 11, 13
5. Making Expensive Experiments Less Expensive A simulated crash test http://www.nytimes.com/2007/06/17/automobiles/17CRASH.html Friday, January 11, 13
6. Computational Literacy The capacity to understand the relationship between “domain knowledge” and the technical knowledge required to understand and apply computational models. ? Friday, January 11, 13
7. The Study • “Treatment”: Students taught subject (based on standards) through the use of a highly scaffolded computational model. • “Control”: Students taught subject using traditional methods. The Study: - The hypotheses Goal: Measure Differences In Student Performance 24 Schools, 3 States, 2400 Students Friday, January 11, 13
8. CCT Friday, January 11, 13
9. • Phase One: Develop The Computational Laboratory Simulations Using Existing Resources. • Phase Two: Field Test and Modify Simulations Based on Teacher/Student Feedback In Small Test Groups • Phase Three: Conduct Research Study With Full Groups. Friday, January 11, 13
10. Predator/Prey Spread of Disease Carbon Cycle Rock Cycle http://www.computationallaboratory.org Friday, January 11, 13
11. Assessing Student Performance • Step 1: The Pre-Test (State Standards-Based) • Step 2: • Treatment: Use Simulations Guided by Driving Questions • Control: Study Subject With Traditional Methods • Step 3: The Post-Test • Step 4: The “Far Transfer” Test Friday, January 11, 13
12. The Results • Data was collected via paper tests... so ‘scoring’ is an ongoing endeavor through this year. • Full Data Sets will be published later this winter or spring. Friday, January 11, 13
13. The Population “The Paper/Permission Slip Problem” Carbon Cycle: 605 Students Took Pre-Test 435 Students Took Post-Test ... but only 400 Students have Both 0 200 400 600 800 Carbon Cycle Pop. Dyn. Friday, January 11, 13
14. “The Random Population Problem” African-American Hispanic Free/Red. Lunch? Control 20% 26% 20% Treatment 35% 28% 25% Our random assignment of Control/Treatment in the smaller population becomes not so random. In the Population Dynamics Data, almost all treatment students come from one school. Friday, January 11, 13
15. Carbon Cycle - Unadjusted 3.50 3.75 4.00 4.25 4.50 Pre-Test Post-Test Control Treatment Gains By Treatment Are Not Statistically Significant Friday, January 11, 13
16. Carbon Cycle - Adjusted 3.50 3.75 4.00 4.25 4.50 Pre-Test Post-Test Control Treatment When Adjusted The Gains Are Statistically Significant Friday, January 11, 13
17. Population Dynamics - Unadjusted 5.00 5.75 6.50 7.25 8.00 Pre-Test Post-Test Control Treatment Gains By Treatment Are Not Statistically Significant Friday, January 11, 13
18. Population Dynamics - Adjusted 5.00 5.75 6.50 7.25 8.00 Pre-Test Post-Test Control Treatment Gains By Treatment Are Statistically Significant Friday, January 11, 13
19. Can We Demonstrate Gains in Learning (or “Computational Literacy” Using Models in the Science Classroom? Answer: From our data, we have evidence, but not proof. Friday, January 11, 13
20. Lessons Learned Friday, January 11, 13
21. • Teachers Matter! Lessons Learned Friday, January 11, 13
22. • Teachers Matter! • The Type of Simulation Matters! Lessons Learned Friday, January 11, 13
23. • Teachers Matter! • The Type of Simulation Matters! • The Story Behind the Simulation Matters! Lessons Learned Friday, January 11, 13
24. • Teachers Matter! • The Type of Simulation Matters! • The Story Behind the Simulation Matters! • The Technical Environment Matters! Lessons Learned Friday, January 11, 13
25. • Better assessments. (“10 yr.” vs. “3 yr.” cycle) • Cognitive research behind “how we use and interpret simulations”. • Cognitive research to explore “highly scaffolded” simulations vs. “project based” simulation activity. • Research on effectiveness of teacher training. Future Directions Friday, January 11, 13
26. • Simulation Library • Concept Maps • Research and Reading •“Expert Voices” http://nsdl.org Friday, January 11, 13
27. Thank You! http://www.computationallaboratory.org http://www.krellinst.org thompson@krellinst.org http://www.krellinst.org/csgf/lab_pics/ Friday, January 11, 13