Understanding the Greenhouse Effect Using a Computer Model

1
Understanding the Greenhouse Effect Using a
Computer Model

• Lisa Schultz
• July 30, 2009

2
Agenda

• Introduction
• Research Questions
• Literature Review
• Settings Participants
• Research Methods
• Results Discussion
• Conclusions Suggestions

3
Introduction

• Motivating factors for investigating computer
  models in the classroom
• IDEAS (Inquiry-based Dynamic Earth Applications
  of Supercomputing)
• NSF funded ITEST program
• Purpose is to help teachers to use models in the
  classroom to teach about climate systems and
  unifying themes.
• NetLogo Climate Change model and Greenhouse
  Effect model

4
NetLogo Greenhouse Effect Model
5
Research Questions

• Did students gain a better understanding of
  system relationships pertaining to the greenhouse
  effect after using the model?
• To what extent were students' misconceptions and
  difficulties congruent with previous research?
• Are concept maps an effective assessment tool for
  measuring student understanding and detecting
  misconceptions?

6
Literature Review

• What do students understand about the greenhouse
  effect?
• To what extent have computer models been shown to
  be effective learning tools?
• How have concept maps been used for educational
  research?

7
Greenhouse Effect

• What do students understand about the greenhouse
  effect?
• Do not distinguish between sunlight and infrared
  energy
• Deterioration of ozone hole causes global warming
• Unclear about the difference between the natural
  greenhouse effect and the enhanced greenhouse
  effect.

8
Computer Models

• To what extent have computer models been shown to
  be effective learning tools?
• Improve students abilities to explain abstract
  phenomena
• NetLogo Agent vs. aggregate behaviors
• Open-source modeling software
• Students can have control over the model
• Studies included models within other learning
  activities (Ex. Predict Observe Explain )

9
Concept Maps

• How have concept maps been used for educational
  research?
• Show conceptual understanding of system
  relationships with propositional phrases
• Shown to reveal changes in student understanding
  over time
• Difficulties with using as an assessment tool
• Students are unfamiliar with concept maps
• Scoring inter-reliability

Linking term
10

• Example of an Expert GHE Concept Map

11

• Example of an Expert GHE Concept Map

12
Setting Participants

• Study conducted in two Maine middle-school
  classrooms, Spring 2009 (n 147)
• 7th grade n 136, 8th grade n 11
• Indicates advanced class
• Indicates 8th grade class

13
Research Methods

• Pre-Test
• 15 minutes
• (a) Concept Maps, (b) Free-Response question, and
  (c) Multiple-Choice questions
• Exploring the model
• 20 minutes
• Post-Test
• Identical to Pre-Test

14
Research Methods

• 1. Pre-Test
• Concept Maps
• Introduction to concept maps for School A
  students
• Concept Map Instructions
• Create a concept map that shows how the Earth's
  greenhouse effect works and how the greenhouse
  effect influences the Earth's temperature.
• Mandatory word bank Sunlight, CO2, Infrared
  Light, Earth, Earths temperature.
• Optional word bank Sun, Atmosphere, Greenhouse
  gases, Clouds, Heat Energy, Outer Space, Earths
  surface

15
Research Methods

• 1. Pre-Test
• Free-Response Question
• Explain in words how the Earth's greenhouse
  effect works.

16
Research Methods

• 1. Pre-Test
• Multiple-Choice Questions
• Four questions from Greenhouse Concept Inventory
  (Keller, 2006) Schools A B
• One question added to detect misconception about
  behavior of sunlight with CO2 School B only

17
Example Multiple-Choice Question

• If human civilization had never developed on
  Earth, would there be a greenhouse effect?
• Yes, the greenhouse effect is caused by naturally
  occurring gases in the atmosphere.
• Yes, the greenhouse effect is caused by plants
  giving off gases during photosynthesis.
• No, the greenhouse effect is caused by humans
  burning fossil fuels and releasing pollutants.
• No, the greenhouse effect is caused by humans
  depleting the ozone layer in the atmosphere.
• No, there is no conclusive evidence that a
  greenhouse effect exists.

18
Research Methods

• Exploring the Model
• Approximately 20 minutes
• Guiding Instructions
• Use the model to determine how each of the
  variables affect the Earth's temperature.
• Determine how and why these variables affect the
  Earth's temperature.
• Post-Test
• (4 Interviews with students from School A)

19
Analysis

• Scoring concept maps free-response question
• Presence of propositional phrases
• Based on expert concept map, the model, and
  student responses
• Phrases weighted to focus on quality of the
  response
• Recorded misconceptions

20
Examples of scored propositional phrases
21
Misconceptions
Misconceptions recorded from concept maps and
free-response questions
22
Example of scored student concept map
7th grade students post concept map (above
average)

• S1 Sunlight can be absorbed into Earths surface
  (1 pt)
• C6 CO2 is a gas in the atmosphere (1 pt).
• I1 IR energy is radiated from Earths surface (5
  pts)
• I4 IR energy affects the Earths temperature (1
  pt)
• R1 Surface reflection affects the amount of
  sunlight and/or IR energy absorbed into Earths
  surface (5 pts)
• Total 13 pts

23
Example of a scored student free-response question
7th grade students post free-response answer
(above average)

• S1 Sunlight can be absorbed into Earths surface
  (1 pt)
• C3 CO2 reflects and/or absorbs IR energy (5
  pts)
• I1 IR energy is radiated from Earths surface (5
  pts)
• R1 Surface reflection affects the amount of
  sunlight and/or IR energy absorbed into Earths
  surface (5 pts)
• Total 16 pts

24
Analysis

• Scoring inter-reliability
• Sub-sample of pre- and post-concept maps and
  free-response questions were scored by two MST
  graduate students.
• No statistically significant differences were
  found between mean scores for each test
• paired two-tailed t-test with a confidence level
  of 95

25
Analysis

• Scoring Multiple-Choice questions
• Point for each correct answer
• Sum of points gave total score
• Only analyzed gains in Questions 2 5 since
  School A did not answer Question 6
• Recorded frequency of responses to analyze
  changes in students understanding

26
Results Discussion

• Pre- and post-test analyses (Research question I)
  
• Equivalence between groups
• Pre- to post-test mean scores
• Normalized gains

27
Results Discussion

• Pre- and post-test analyses (Research question 1)
  
• Equivalence between groups
• Pre- to post-test mean scores
• Normalized gains
• Frequency of responses
• Frequency of Misconceptions (Research question 2)
• Correlation between concept map and free-response
  scores (Research question 3)

28
Results Discussion

• Pre- and Post-Test Analyses
• Equivalence between groups
• Analysis of variance (a 0.05)

29
Results Discussion

• How did students understanding change after
  exploring the model?
• Paired one-tailed t-tests to compare pre- to
  post-test mean scores (a 0.05) indicates
  significant result

Max concept map and free-response score 48
Max multiple choice score 4
30
Results Discussion

• How did students understanding change after
  exploring the model?
• Paired one-tailed t-tests to compare pre- to
  post-test mean scores (a 0.05) indicates
  significant result

Max concept map and free-response score 48
Max multiple choice score 4
31
(No Transcript)
32
(No Transcript)
33
Results Discussion

• Normalized gain ltggt
• Independent one-tailed t-test to determine if
  mean ltggt was significantly greater than zero (a
  0.05)

34
Results Discussion

• Students test scores increased, but what did
  they learn?
• Frequency of students responses for each
  assessment

35
Results Discussion

• Greatest increases in propositional phrases from
  pre- to post-test (n 147)

36
Results Discussion

• Change in multiple-choice responses from pre- to
  post-test

37
Results Discussion

• To what extent were students' misconceptions and
  difficulties congruent with previous research?
• Frequency of misconceptions detected in
  free-response questions and concept maps
• Frequency of misconception distracters in
  multiple-choice questions

38
2 (a) Concept Map Free-Response Misconceptions
39
2 (b) Multiple-Choice Questions Misconceptions
40
2 (b) Multiple-Choice Questions Misconceptions
41
2 (b) Multiple-Choice Questions Misconceptions
42
2 (b) Multiple-Choice Questions Misconceptions
43
2 (b) Multiple-Choice Questions Misconceptions
44
Results Discussion

• Are concept maps an effective assessment tool for
  measuring student understanding and detecting
  misconceptions?
• Pearsons coefficient
• Test scores and frequency of responses

45
3(a) Correlation between Pre-Concept Maps and
Free-Response Questions
46
3(a) Correlation between Post-Concept Maps and
Free-Response Questions
47
Results Discussion

• 3(b) Test Scores and Frequency of Responses
• No significant difference between mean pre-test
  scores between CMs and FRs.
• Significant difference between mean post-test
  scores between CMs and FRs.
• Free-response question detected 74 of
  misconceptions. Concept maps detected 26.

48
Conclusions Suggestions

• Middle-school students had gains in understanding
  about how the GHE influences the Earths
  temperature after exploring the model for 20
  minutes.
• IR energy is radiated from the Earth.
• Earths temperature increases when sunlight is
  absorbed into the surface.
• Increasing CO2 in the atmosphere increases the
  Earths surface temperature.
• Students held on to their misconceptions from the
  pre- to post-test.

49
Conclusions Suggestions

• The model may be more effective within a guided
  inquiry activity.
• Predict, observe, explain
• Forum for students to ask questions about
  behaviors Why does the IR energy get reflected
  by CO2 and not sunlight?
• Confront misconceptions
• Critically analyze model
• Benefits and limitations
• Opportunities to modify the model

50
Conclusions Suggestions

• Concept maps should be investigated further to
  determine if they are effective educational
  research assessment tools.
• Different grade levels
• More practice at making concept maps
• Better as a learning tool?

51
Thank you