Docstoc

Learning by Problem-Posing and Agent-Assessment

Document Sample
Learning by Problem-Posing and Agent-Assessment Powered By Docstoc
					Learning by Problem-Posing and Agent-Assessment
Tsukasa Hirashima Hiroshima University

1

Learning by Problem-Posing and Agent-Assessment
Learning method
•Students are required to make problems that the students are usually required to solve. •Promising method to promote deeper comprehension for problems

Feedback and guidance are strongly required Interactive and Intelligent Learning Environment
Automatic Assessment of Posed Problems
2

Learning by Problem-Posing and Agent-Assessment
Arithmetical Word Problems

How to pose problems

Problem Posing as Sentence Integration
•Several sentence cards are provided •Problems are composed by selecting and ordering sentence cards

•Simplification problem posing •Computer Readability •Keeping Essential Activity

3

Contents


Categorization of Learning by Problem-Posing
• Solution Based Problem Posing



Interactive Learning Environment for ProblemPosing
• How to pose problems
• Sentence Integration

• Explanation of Monsakun


Experimental Evaluation by Two Months Use 4

Categorization of Problem-Posing
Problem = “Given Information” + “Required Information”
Method to derive R-Info from G-Info is Solution Method.

A set of given information is Episode.
  

Solution-Based Problem-Posing Problem-Based Problem-Posing Episode-Based Problem-Posing
5

How to pose problems
(1) Natural Language (2) Sentence templates
• Sentence template has several blanks • Fill in the blanks by prepared concepts • Make a problem with the sentences

(3) Problem template
• Problem template has several blanks • Fill the blanks by prepared concepts

(4) Sentence cards
• Sentence cards are provided • Make a problem by selecting and combining them • Problem-posing with sentence integration
6

Process Model of Problem-Solving of Arithmetical Word Problem
Tom had five pencils. Ken gave three pencils to Tom. How many pencils does Tom have?
Natural Language

sentence

sentence sentence sentence

○+△= ?
Plan Execution

?=□

sentence

Transformation

A process model of problem-solving of arithmetical word problem

…
Integration

7

Problem-Posing by Sentence-Integration with Sentence Cards
Natural Language
sentence sentence

sentence
sentence

○+△= ?
Plan Execution

?=□

sentence

Transformation

Most important process in the problem solving is “Integration”. (1) Provide with a set of sentence cards. Each card has one sentence. (2) Select necessary sentence cards and arrange them in proper order. Integration process is remained, but transformation process is simplified
Problem-posing activity becomes simple, but keep the essential process
8

…
Integration

Learning Environment for Solution-Based Problem Posing as Sentence-Integration
•Simple arithmetical word problems
•Solved by one addition or subtraction

•Solution-based problem posing
•To make problems that can be solved by specified solution method

•Problem-posing as sentence integration
•Problem posing by Combination of sentence cards

Learning environment for leaning by problem-posing: MONSAKUN

9

Return

Change task

Quit

Pose a problem that can be solved by " 5 - 3 ".

Put a card in this blank

Eject

Put a card in this blank
Put a card in this blank

Eject

Tom has five pencils. Tom received three pencils from Ken. How many pencils does Tom have? How many pencils does Ken have?
Ken gave three pencils to Tom.

Eject

Check the problem! Put a card in the same shape blank

Ken received three pencils from Tom.

10

Return

Change task

Quit

Pose a problem that can be solved by " 5 - 3 ".

Tom has five pencils.
Ken gave threethis blank Put a card in pencils to Tom. How a card pencilsblank Put many in this does Tom have?

Eject

Wrong!! Tom received three pencils from Ken. Let’s think about the second Kensentence. three pencils received Does this problem solve by from Tom. 5-3? Try again. How many pencils does

Eject

Eject

Ken have?

Check the problem!

Put a card in the same shape blank

Ken received three pencils from Tom.

11

Demonstration

12

Experimental Uses in Elementary School
774 students
•Two classes in the third grade: 44 students(two class times) •Three classes in the second grade: 91 students (two class times) •Six classes in the second grade: 132 students (two class times) •Three classes in the second grade: 99 students (two class times and two months use in free time) •Two classes in the second grade: 46 students (two class times) •Three classes in the second grade: 76 students (two class times) •Three classes in the first grade: 104 students (one class time) •Four classes in the second grade: 143 students (two class times) •Two classes in the fourth grade: 39 students [eight class time]
13

Long Term Use
•Subjects: •99 elementary students •3 classes in the second grade •Period: •2 months (46 school days) •Situation

•Two systems in a class (6 systems in the school) •Out-of-class time •Free use

14

Experimental Evaluations
● Whether students use the system on their own will (1)Problem Posing Logs (2)Questionnaires ● Effect of long-term use (1) Extraneous Problem Test

15

Results of the use
•Total Posed Problems: 8,386 problems

•30.4 problems were posed in a day
•A student posed 84 problems in average.

•A student used the system 8.5 days in average.
Total days were 46 days There were only 6 systems for 99 students.

16

Results of Questionnaires

Answer Question Do you think MONSAKUN make arithmetic enjoyable? Do you think MONSAKUN is a game? Do you think MONSAKUN is useful for arithmetic learning? Do you like to use MONSAKUN more often? Do you think you could make problems easier than before?

Yes 84 52 82 80 75

No 1 21 3 2 3

No idea 6 18 6 9 13

17

Extraneous Problem Test
There are two apples. There are three oranges. There are seven bananas. How many apples and oranges are there in total?

•Extraneous problem includes extraneous sentence that is not necessary to solve the problem.
•It is required to judge the relevant sentences and extraneous sentence.

•It is useful to assess the ability to integrate sentences.

18

Extraneous Problem Test
Pre-test score High-score Low-score (>8.32) (=<8.32) 32 20 12 21

Number of posed problems

High-posed (> 77) Low-posed (=< 77)

Average score of pre-test = 8.32 Median of the number of posed problems = 77

19

Extraneous Problem Test
Condition High-score/High-posed (n =32) High-score/Low-posed (n =20) Low-score/High-posed (n =12) Low-score/Low-posed (n =21) Pre-test 11.06 (SD=1.16) 10.65 (1.18) 4.41 (2.78) 4.14 (3.04) Post-test 10.96 (1.28) 10.50 (2.25) 7.25 (2.52) 4.71 (3.81)

• Full marks are 12. • In the comparison between pre-test and post-test, there is significant difference only at low-score/high-posed group .

20

Results of experimental use
(1) Some of students posed problems with the system eagerly even in free use situation. (2) Students and teachers accepted the problemposing as useful learning activity. (3) This problem-posing was useful to improve problem-solving performance for lower score students who posed problems eagerly.
Our approach to realize “learning by problemposing” is promising.

21

Related Future works


Related and future works
• Solution-based and sentence-integration
• Lower grade students: first grade students • Comparing with other learning

• Other type of problem-posing
• Problem-Based Problem Posing: Problem-Transformation • Story-Based Problem Posing

• Other domains
• Mechanical problems • Multi-digit problems • Equation problems

• Model of problem-posing • As a promotion method for metacognition
22


				
DOCUMENT INFO