6

Click here to load reader

EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

  • Upload
    doandat

  • View
    212

  • Download
    0

Embed Size (px)

Citation preview

Page 1: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

EEG Analysis for Computational Thinking based Education Effect on the Learners’ Cognitive Load

SE-YOUNG PARK, KI-SANG SONG1, and SOON-HWA KIM

Computer Education Korea National University of Education

ChungBuk, 363-791 REPUBLIC OF KOREA

[email protected], 1Corresponding author:[email protected], soona6570@ gmail.com Abstract: - This paper shows the impact of Computational Thinking (CT) education on problem solving capacity and its affect to reduce students’ cognitive load in problem solving environment. CT education was delivered to 16 undergraduate students with 10 sessions and total 30 hrs courses, specially designed for applying CT process on problem solving situation. To measure the impact of CT education, participants’ CT capacity was tested before and after the teaching of whole courses for two groups; CT group and Not-CT (NCT) group. Also EEG measurement was pursued for observing the learners’ cognitive load level changes during the problem solving processes with Fp1 and Fp2 channels. The analysis results show that students learned with CT based curriculum showed higher result in CT based problem solving test. In terms of EEG result, SEF- 95% was lower with CT group than NCT group. Key-Words: - Computational Thinking, Problem solving, Cognitive load, EEG, SEF-95% 1 Introduction Modern society requires the cognitive ability to resolve the complexity of the real world. However, this ability is difficult to be acquired through a separate enhancement training of each discipline, and it should be addressed in a practical and comprehensive perspective. Recently, researchers from computer science disciplines assert that computer science is suitable for learning cognitive skills.

Computing capacity has become a major skill to carry out a logical and procedural thinking to solve complex problems in various academic and real life situations. Along with the emphasis on the computing capacity, computational thinking (CT) is also emphasized as a core ability for 21st century learners. Wing (2006) noted that CT can be acknowledged as basic literacy as 3R (Read, wRite, aRithmetic).

The purpose of this research is to examine the impact of CT based curriculum on students’ problem solving capacity and cognitive load with comparative analysis. For this purpose, two students groups were selected from the students taking same subject. Both two students group were taking Scratch programming course, and one group was taught emphasizing the CT characteristics and the other group was taught

traditional way of Scratch programming teaching methodology. To measure the effect of different approaches of learning Scratch programming, two different tests were given for the participants; paper and pencil based test for students’ CT capacity measurement, and EEG measurement for obtaining of objective data of the effectiveness of CT emphasizing curriculum.

2 Theoretical Background 2.1 Computational Thinking (CT) Introducing CT in education is beneficial to learners in terms of developing learners' thinking capability and promoting the strategies to put their knowledge to practical use (Barr & Stephenson, 2011). CT helps learners to choose and utilize appropriate tools and strategies for problems solving (Yadav, et al., 2011).

CT can be used to establish a suitable strategy for solving problems. This strategy can involve the use of suitable automated algorithm to solve the problem with computing system. Accordingly, CT is expected to improve the efficiency in solving problems in real-life or a variety of disciplines. In general, CT assist learners to adapt to information

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 38

Page 2: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

oriented digital society. CT education is effective to systematically improve the following capabilities (see Table 1): The ability to gather the necessary information effectively (Data Collection), The ability to analyze the pattern of the collected data (Data Analysis), The ability to filter out the essential elements needed to resolve the issue (Data Representation, Problem Decomposition, Abstraction), The ability to design a problem-solving process with procedure (Algorithms & Procedures), The ability to take advantage of the computing system to solve the problem (Automation, Simulation, Parallelization). Ultimately, with aforementioned systemically approach, CT is expected to improve the problem solving skills of learners. Table 1. Concepts of CT

Concept Definition

Data Collection The process of gathering appropriate information

Data Analysis Making sense of data, finding patterns, and drawing conclusions

Data Representation Depicting and organizing data in appropriate graphs, charts, words, or images

Problem Decomposition

Breaking down tasks into smaller, manageable parts

Abstraction Reducing complexity to define main idea

Algorithms & Procedures

Series of ordered steps taken to solve a problem or achieve some end

Automation Having computers or machines do repetitive or tedious tasks.

Simulation Representation or model of a process.

Simulation also involves running experiments using models.

Parallelization Organize resources to simultaneously carry out tasks to reach a common goal.

Source : ISTE, CSTA, NSF (2011) 2.2 Computational Thinking (CT) and Cognitive load According to cognitive load theory (CLT), cognitive overload occurs when the exceed of required cognitive resources happens to solve the problems. The cognitive overload is considered as a major cause of learning failure. Thus, there have been many attempts to identify

the factors that causing unnecessary cognitive load and the teaching strategies to manage the learners’ cognitive load during study (Corno & Mandinach, 1983).

Sweller (1988) researched to promote learners’ problem solving process by reducing the cognitive effort of learners and obtaining a cognitive structure that is needed to resolve the problem. It was about how learners can acquire the problem solving principle and store it to long-term memory storage device. Since CT is mainly deals with the problem solving strategies and how to apply it in real complex problem. It can be assumed that the sustainable application of CT based curriculum can positively influence on learners’ cognitive load during problem solving tasks.

2.3 Cognitive load and Brain Activation 2.3.1 Brain activation measurement Brain waves measurement is processed in the following manner. First, deriving the summation of brain electrical activity that occurs in the nerve cell population of cerebral cortex in vitro. Then the summation is amplified and recorded from the scalp phase (intact scalp) as the horizontal axis with the time and the vertical axis with electric potential. EEG is an objective, non-invasive, continuous test to assess cerebral function and it is widely used today for cerebral function evaluation. 3.3.2 Cognitive load and SEF-95 In general, SEF-95 is widely used as an indicator to assess the cognitive load, the mental stress levels and excessive mental arousal level when performing the task. The high value of SEF-95 indicates the EEG arousal or cognitive load is high (Kim, 2010; Choi, 2011). Through the activeness’ of the cerebral cortex, the extend of the problem solving process can be assumed In the case of the creative problem solving, the brain active of frontal and parietal lobes are tend to be lower (Matindale, 1999). In this study, researchers used the SEF-95 as an indicator for viewing the

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 39

Page 3: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

cognitive load level of participants during the problem solving tasks. 3 Method 3.1 Experimental Process The purpose of this study is to investigate the effect in learners’ cognitive load during the problem-solving process, after participating CT skills applied curriculum. In the experiment, after the 10 sessions of learning (30hrs in total), the problem solving ability test (paper-based) and EEG measurement of frontal lob were conducted between CT-group and NCT-group (who did not participate CT course). CT ability test for college students (Lee, 2009) was applied to measure general problem solving ability and problem solving questions of PISA was used for the EEG measurement.

Brain activation was measured with 2 channels (Fp1 and Fp2). To accurately measure brain activeness, international 10/20 EEG measurement electrode arrangement system was used as in Fig. 1.(Klem, et al., 1999) EMOTIV’s wireless device, EPOC Neuro Headset was used to measure the cognitive load of the participants and the EEG data was processed with EMOTIV’s Test Bench. The measured analog signal data is transformed to the digital signal with the Complexity 2.0 of Laxtha and EEG was analyzed with a Fast Fourier Transform (FFT).

Source : TCT (2012) Fig. 1. 10/20 system position

EEG measurement result can be affected by

location, the condition of participants, illumination and noise of the place. Thus, to prevent possibility of outside environment factors interfere the validity of the experiment, the measurement was conducted in the shield room. 3.2 Questionnaire Items to measure the problem solving ability should be able to minimize the effects of specific subject prior knowledge and able to measure the actual ability during the problem solving process. Thus, current research selected CT-based problem solving test (Lee, 2009) which is derived from the OECD / PISA problem solving questions and developed for the undergraduate students. The questionnaire was composed with 3 main categories and 8 sub-element of CT ability. Total 18 questions were designed and some examples are shown in the appendix.

Table 3. Questionnaire construction of assessing CT capability (Lee, 2009)

Category Sub- element of CT capacity

Question number

Number of Questions.

Problem Discovery

Logical Thinking 5, 14

3

Analytical Thinking 6

Problem Analysis and

Representation

Analytical Thinking 1, 7

3

Abstract Thinking 2

Problem Solution Strategy

Simultaneous Thinking 4

10

Precedence Thinking 16

Strategic Thinking 3,8,12, 13

Procedural Thinking 10, 11

Recursive Thinking 9, 15

Total 16

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 40

Page 4: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

3.3 Participants 3.3.1 CT based problem solving test Total 34 students (16 male, 18 female), 20 to 24 years of age participated in the experiment 16 students from the group of participated CT education course with Scratch program (CT-group) and 18 students from the group of participated basic computer skill course with Scratch program (NCT-group). 3.3.2 Brain activation measurement (EEG) Right handed 13 students (6 male, 6 female), 20 to 24 years of age participated in the experiment. 6 students from the group of participated CT education course with Scratch program (CT-group) and 6 students from the group of participated basic computer skill course with Scratch program (NCT-group). 4 Result 4.1 Analysis of CT based problem solving test result Pre - CT based problem solving based test was conducted with both CT and NCT group. According to the t-test result, both groups shows the almost similar value from most areas, thus it is possible to assume that CT and NCT groups are comprised as almost identical groups. Table 4. t-test result : Pre CT based problem solving result

Group Mean SD t Sig.

Analysis and Representation –

Analytic Thinking

CT .8750 .22361 -2.376 .024*

NCT 1.0000 .00000

Analysis and Representation –

Abstract Thinking

CT .8125 .40311 .243 .810

NCT .7778 .42779

Solution Strategy - Strategic

Thinking

CT .7344 .28090 .140 .890

NCT .7222 .22506

Solution Strategy - Simultaneous

Thinking

CT .8125 .40311 -1.181 .246.

NCT .9444 .23570

Discovery - Logical Thinking

CT .6250 .28868 1.547 .132

NCT .4444 .32970

Discovery - Analysis

Thinking

CT .7500 .44721 1.173 .249

NCT .5556 .51131

Solution Strategy - Recursive

Thinking

CT .7500 .36515 1.176 .248

NCT .6111 .32338

Solution Strategy - Procedural

Thinking

CT .5000 .36515 .434 .667

NCT .4444 .37920

Solution Strategy – Precedence Thinking

CT .6250 .50000 .636 .529

NCT .3889 .50163

After the 10 sessions, the post- CT based

problem solving test with problem isomorphs. Both groups’ participants show better records than pre-test. Especially, CT-group shows better result in the questions to measure solution strategy. CT group’s mean result of strategic thinking was 0.8438 while NCT group got 0.6240 (p<.05). Also, in simultaneous thinking CT group’s mean result was higher as 1.00 and NCT group was 0.7778 (p<.05). Especially, the t-test value was found highly significant (p<.01) between CT group and NCT group in terms of recursive thinking.

Table 5. T-test result: Post CT based problem solving result

Group Mean SD t Sig.

Analysis and Representation –

Analytic Thinking

CT 1.0000 .00000

NCT 1.0000 .00000

Analysis and Representation –

Abstract Thinking

CT 1.0000 .00000

NCT 1.0000 .00000

Solution Strategy - Strategic

Thinking

CT .8438 .22127 2.329 .026*

NCT .6250 .31213

Solution Strategy - Simultaneous

Thinking

CT 1.0000 .00000 2.074 .046*

NCT .7778 .42779

Discovery - Logical Thinking

CT 1.0000 .00000 1.618 .115

NCT .8889 .27416

Discovery - Analysis

Thinking

CT 1.0000 .00000 1.372 .180.

NCT .8889 .32338

Solution Strategy - Recursive CT .6875 .35940 2.768 .009**

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 41

Page 5: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

Thinking NCT .3333 .38348

Solution Strategy - Procedural

Thinking

CT .9375 .17078 1.430 .162

NCT .8333 .24254

Solution Strategy – Precedence Thinking

CT .6250 .50000 .716 .479

NCT .5000 .51450

4.2 Analysis of Brain Activation Test Table 6. T-test result: SEF-95%

Group Mean SD t Sig.

SEF- 95%

Test 1 CT 23.9895 3.2961

-.019 .985 NCT 24.043 5.9336

Rest 1 CT 20.3175 4.8853

-2.047 .068

NCT 26.2695 5.1846

Test 2 CT 27.6642 4.9233

0.806 .439 NCT 25.831 2.6148

Rest 2 CT 17.3268 2.5298

-.375 .004 NCT 23.7372 3.336

Test 3 CT 23.1513 3.376

-.689 0.506 NCT 24.4168 2.9734

SEF-95% analysis was conducted to assess participants’ cognitive load level during problem solving tasks. T-teat reveals that CT group participants performed more strategically on the task than NCT group participants. Both group started similarly at the first task (SEF Mean: CT -23.9895, NCT-24.043).

Fig. 2. SEF 95 means result per each task As the task continued, CT group students show more stable brain activity on the task. CT group shows strong attention on the problem solving task and they easily got back to more relaxed mode. Meanwhile, NCT group EEG result revealed rather irregular brain performance than CT group and higher cognitive load than average. It can imply the possibility that CT education can helps students to approach the problem solving task more strategically. The result can also be linked with paper based test result, which is the CT group performed better at solution strategy test. 5 Conclusion In this study, 10 weeks program for improving CT capability was applied, and impact of CT based curriculum on learners' problem solving ability and cognitive load. Following summary can be derived from these results.

First, proper study of the CT course, which presenting a systematic strategy for problem solving, can help to improve students' problem-solving skills especially on the ability to strategically thinking to solve problems. Second, by learning problem-solving strategies repeatedly, when learners met the complex problem situations the cognitive system can be stabled.

This study can be significant by conducting the measurement with both a traditional paper-based approach and biometric assessment to derive a more objective result. In the future research it can be expected to obtain more generalizable result with larger sample and measuring more EEG channels to elaborately analyze learners' cognitive load.

6 Acknowledgements This work was supported by the Ministry of Education and National Research Foundation of Korea Grant funded by the Korean Government (No. 2014S1A5A2A01010630). References:

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 42

Page 6: EEG Analysis for Computational Thinking based Education ... · EEG Analysis for Computational Thinking based Education Effect ... basic literacy as ... The ability to filter out the

[1] Barr, V., & Stephenson, C., Bringing computational thinking to K-12: what is involved and what is the role of the computer science education community?, ACT Inroads, Vol, 2., No. 1, 2011, pp.48-54

[2] Choi, Y.H., Analysis of Electroencephalogram (EEG) Activities of Middle School Students in Technological Problem Solving Thinking Process as Levels of Structured Problem, Journal of Korean Practical Arts Education, Vol. 17, No. 4, 2011, pp. 129-152

[3] Corno, L., & Mandinach, E.B., The role of cognitive engagement in classroom learning and motivation, Educational Psychologist, Vol. 18, No. 2, 1983, pp.88-108

[4] ISTE, CSTA, NSF, Computational Thinking Teacher Resources, 2011, Retrieved from : http://www.csta.acm.org/Curriculum/sub/CompThinking.html

[5] Kim, S.H., A Neuro-scientific Approach to the Relationship between Creativity and Knowledge, M.Ed. Thesis of Korea National University of Education, 2010

[6] Klem, G. H., et al. The ten-twenty electrode system of the International Federation. Electroencephalography, Clinic Neurophysiology, 1999, 52.suppl.: 3.

[7] Lee, E.K., A Robot Programming Teaching and Learning Model to Enhance Computational Thinking Ability, Ph.D Thesis of Korea National University of Education, 2009

[8] Martindale, C., Biological based of creativity. In Stenrberg, R. J. (Ed.), The handbook of creativity, Cambridge University Press, 1999

[9] Sweller, J., Cognitive load during problem solving: Effects on learning, Cognitive Science, Vol. 12, 1988, pp.257-285

[10] TCT, 10/20 System Positioning manual, Hong Kong, Trans Cranial Technologies ltd., 2012

[11] Wing, J.M., Computational Thinking, Communication of the ACT, Vol. 49, No. 3, 2006, pp.33-35

[12] Yadav, A., et al. introducing computational thinking in education courses. In: Proceedings of the 42nd ACM technical symposium on Computer science education. ACM, 2011, pp. 465-470.

Appendix. Examples of questions used for CT capacity test (Lee, 2009) Q-A. Following is a description of instructions for generating a figure below. Fill the blanks with appropriate instruction.

Q-B. Write appropriates commands for generating the next figure below. (Use the repeating structure, and a single variable (A))

Recent Advances in Computer Science

ISBN: 978-1-61804-297-2 43