A Study on How Narrating Improve Students’ Fluency in Computer Code Development

A Study on How Narrating Improve Students’

Fluency in Computer Code Development

Hoiming Lee

ED690
Pace University

April 15th, 2024

Page 1
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Table of Contents
I. Problem Statement.................................................................................3
II. Context Background.............................................................................5
III. Literature Review................................................................................7
IV. Research Questions and Methods.....................................................13
IV.1 Research Questions...................................................................13
IV.2 Research Method......................................................................14
V. Data Collection..................................................................................19
VI. Date Analysis....................................................................................20
VI.1.a Proficiency in Narrating (foreign language high performer).20
VI.1.b Proficiency in Narrating (foreign language low performer)..20
VI.2.a Accuracy in Narrating (foreign language high performer)....21
VI.2.b Accuracy in Narrating (foreign language low performer).....21
VI.3.a Usage of CS Vocabulary (foreign language high performer)22
VI.3.b Usage of CS Vocabulary (foreign language low performer).22
VI.4.a Code Tracing Ability (foreign language high performer).....23
VI.4.b Code Tracing Ability (foreign language low performer)......23
VI.5.a Use of Cause & Effect (foreign language high performer). . .24
VI.5.b Use of Cause & Effect (foreign language low performer)....24
VII. Conclusion.......................................................................................25
VIII. Next Steps & Reflections...............................................................28
References...............................................................................................30
Appendix A - Survey on Motivation to Learn Computer Science.........32
Appendix B - Assessment Data..............................................................33

Page 2
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

I. Problem Statement

The Digital Use Divide is widening, as evidenced by high school students stuttering in

computer code narration [1].

3.5% of adult population makes up 5% of wealth investors in our society [2]. The years-

long government effort to promote equity in our society consumed billions of dollars. Back

35 years ago, $40B was invested to facilitate the use of technology to improve professional

development skills in the high school education system [3]. The US Government has invested

heavily in enhancing language literacy in schools, and much of the focus lies on the reading

and writing skills in ELA subjects. According to a Federal report, A Nation at Risk[4],

Computer Science is identified as one of the five essential proficiency for high school

graduation requirements. Although the US is ranked among the top 10 countries in high

school computer literacy [5], another report also uncovers that 10% of NYS 8th graders need

more literal skills [6]. Every limited data is available to understand the spread of this 10% of

8th graders; it will be a big concern if the spread is focused in New York City. The NYCDOE

emphasizes the critical factor of Digital Equity in the school system [7]. As active on-duty

teachers in the NYCDOE system, we may be able to discover some hints about students’

computer literacy skills based on observations from students in NYC schools.

In the past three years, I taught 600 students in computer science, and less than 5% of

students were proficient in computer coding. Although the low percentage is not the complete

picture of the NYC school learners, it causes my curiosity about the linkage between the data

from reference [6] above and my observation at work. Interestingly, these 5% of students are

not the top students; they are just in love with computing. Their interests are video games,

Page 3
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

online social media, etc. Only some are highly academic and want to leverage computer

literacy to assist their studies in other academic subjects. The remaining 95% of observed

students are brilliant in foreign languages and ELA subjects. The data is based on my three

years of teaching observation [Appendix A]. Most students find it challenging to learn

computer coding in high school, but they want to learn it because it is trendy among their

peers. Many students speak fluently in Economics, Social Science, ELA, and History. Their

learning behavior differs significantly between ELA and Math or Computer Science classes. I

once thought learning Computer Coding was like learning another foreign language, but

students’ learning disposition told me a different story.

Students naturally speak and participate in ELA class discussions but behave quietly and

respond hesitatingly and cautiously in Computer Science class. I always find out many

students are fluent in reading novels or ELA articles, but they stutter in reading computer

code. Students spend many years in school learning ELA literacy skills. Students can not

proficiently demonstrate the ability to leverage their ELA literacy skills in learning compute

coding. If we can understand what prevents students from applying their ELA literacy

reading skills in reading computer coding, we can improve computer science learning among

students. The study focuses on reading skills to enhance computer code writing, similar to

literacy reading skills. There is little research on coding narrating. An initial study recognizes

the code narrating problem and identifies reading factors in coding and their relationships to

ELA reading.

Page 4
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

II. Context Background

In BTHS, every sophomore must take the AP Computer Science Principles (CSP) class,

which is counted as a Maths elective for high school graduation requirements. The course

does not require any programming background. Students spend 60% of their time coding and

40% covering diverse computer science topics from internet security, software bias, and data

analysis. By the end of the course, every student must complete independent software

projects and take a test to answer four written response questions. The course provides

students with all the learning materials and access to school computing devices.

BTHS is a big school. The sophomore student population has 1,500 students. There are

45 CSP class sessions, each with 34 students. I am teaching six sessions of the CSP class, and

five of them are selected for the study. Six students are chosen from each of the five sessions

to participate in the study without affecting students' learning routines. A total of 30 students

participated in the survey [Appendix B].

Computer languages have their primitive constructs, syntax, semantics, and algorithms,

similar to a foreign language's vocabulary, grammar, sentence structure, and composition.

The study's idea is to understand the similarities and differences between learning computer

language and other foreign languages in the high school learning environment.

The study aims to experiment with the effect of coding narration on students who are low

performers with computer science grades lower than 80%. Among these 30 computer science

low-performance students, one-third are high performers in foreign languages with grades

above 95%. Another one-third are medium performers in foreign languages with grades

below 90%. The third is low performance in a foreign language with a grade below 80%.

Page 5
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

[Appendix ]. We use the naming convention Group L, Group M, and Group H to represent

the population sampling in the study.

Page 6
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

III. Literature Review

Students' access to technology is significantly improved today compared to twenty ago.

Every student in the Computer Science class has access to a computer, and the use of Google

Classroom also provides every student equal access to computing devices. In the computer

science class, the teacher also prepares coding materials in Google Docs, making it easy for

students to follow instructions to complete coding assignments. When I ask students to

narrate a segment of programming code, they stutter and use little computer science academic

language. I wonder if students also stutter in other ELA subjects. I observed other ELA

classes where students showed smooth narration. It seems there is a direct relationship

between students’ coding proficiency and the ‘Digital Use Divide.’

When every student has equal access to computing devices, it is unlike a digital divide issue.

It is evidence that students have different access to digital use experience. There are active

and passive users of technology, including programming language. There are distinct

differences in the coding performance between active and passive users [1]. Passive users

follow teacher-provided instructions to fill in the digital sheet, put the coding blocks together,

and follow the laboratory exercise guidelines to press software buttons as instructed. Passive

users receive digital information and only operate under the condition of detailed instructions.

When instruction is at a high order level, or prediction is required, passive users easily get

lost or slow down in coding. Active users are different. They engage in creative, critical

thinking, imagination, reflection, and participation. They explore, build, and collaborate in

coding and solving problems. When instructions are unavailable, active users will explore

with critical thinking, analysis with reasoning, and create with collaboration to complete

computer science coding tasks. Active users demonstrate the attitude of finishing a task; on

the contrary, passive users illustrate the attitude of completing assigned exercises.

Page 7
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

The class survey [Appendix A, Table 1.1] reflects the disparity of access to programming

experience. While 17% of the sample population are active users of programming languages,

the rest, 83%, are passive users of programming coding in the AP Computer Science class.

The disparity data is consistent with the Digital User Divide [1].

MONTAGUE & XUE [9] provide more research data to characterize active and passive

users. In MONTAGUE’s studies, passive users are operators who are likely to monitor tasks

such as recording and following instructions step by step. In the same research,

MONTAGUE names the active users as operators who are likely to take the roles of

planning, analyzing, and predicting tasks. MONTAGUE’s research results support the claims

of the Digital Use Divide[1].

In the following figure, in a high school classroom, the technology represents computer

science programming; active users are students who are proficient in coding, and passive

users are weak and skilled in coding. They are interacting among themselves.

Source: doi: 10.1016/j.apergo.2011.11.002 [9]

Passive users are prevalent in the real world [9]. From [Appendix A] statistics, the

passive users comprise 83% of the sampling population; we can shift the learning

demographics from 83% to a lower percentage, which is the focus of this active research.

Page 8
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Although passive users are necessary for specific job tasks, such as aviation, education, and

health industries[9], we should equip students with an active learning attitude for the

educational purpose of student-centered learning. And in this study, it means active users of

programming code. What we are going to investigate does not contradict MONTAGUE [9]

results; instead, MONTAGUE [9] characterization of activeness and passiveness provides

guidance when selecting students in this study. Both passive users and active users are

equally important[9]. The existence of low-proficient student coders provides teachers with

the grounds to reform instructional pedagogy to fit the learning needs of the weak-proficient

student coders. That is also the theme of this research: select narrating from a literature

device to improve coding proficiency. It also echoes MONTAGUE's [9] idea that considering

active and passive users requires a one-size-fits-all tactic. When applying the MONTAGUE

[9] notion in a computer science classroom environment, the active and passive students are

the sound evidence of the need for differentiated instruction to teach computer science to

high- and low-performance coders.

MONTAGUE [9] also argues that active and passive users exhibit different levels of trust

in human-to-human and human-to-technology. Extending the MONTAGUE [9] concept in a

computer science learning environment, active learners tend to build a trusting relationship

with programming technology. Digital Use Divide[1] describes a similar concept but more

elaborately describes the characteristics of active users participating in creative, critical,

exploratory, analytical, and reflective activities. MONTAGUE[9] also argues that passive

users are opposite to active users when establishing a trusting relationship with technology,

which could mean learning coding in the school environment. Because of the nature of the

trust relationship between active and passive users with programming language or

technology, some measures are needed to support the passive users so that they can reinforce

Page 9
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

their trust in computer programming. An early article, Making Connection, by Caine[8],

describes the traditional method of connecting teaching and the human brain.

Caine[8] described learning as making connections of information, and based on the

complexity of information connectivity, our brain was positioned to quest for patterns.

Identifying patterns is the key to extracting useful information to analyze and solve problems.

There is no doubt that high school students receive learning materials every day they come to

school. In computer programming, teachers provide coding exercises with detailed

instructions, and the integrated coding platform is very user-friendly today compared to 20

years ago. However, some students are still stuck in computer programming. In Caine's [8]

studies, a reform of how students learn or how teachers teach is based on understanding how

the human brain and teaching interact.

Many education theories have been done or studied. Teachers have learned this education

theory in teacher school or in-service training. Much has been taught and talked about.

Caine([8],p.viii) believed that a more complex teaching framework is imminent. The task is

to address how the framework can help learners make connections of their already known

knowledge. Caine[8] pointed out that rote memorization is not the same different from

information. Learners retrieve information through rote memorization, but this does not mean

the retrieved data is connected. They are pieces of unconnected details that could be more

useful for analyzing and solving problems. Even if learners can make the connections

between information, Caine([8], p.viii) pointed out that the Connection could be in an optimal

state or a depressed state. There are one hundred billion neurons in the human brain, and the

enormous ways of neuron connectivity contribute to the different levels of learning

experience. Only if the information is connected is the precedence before the human brain

Page 10
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

can carry on to do the following task: recognizing patterns. That also explains why some

learners need to improve in computer coding.

Caine described that a healthy human brain can recognize patterns and make

approximations, correct from self-reflection and experience, have the capacity for different

types of memorization, and create new knowledge ([8], p.3). Such an idea is also described in

the characteristics of active and passive users in the later published Digital Use Divide [1]

and MONTAGUE [9]. These brain features are not new discoveries; the question is how

learners can access these brain features, which are critical tools to improve a learner's coding

efficiency. Both Digital Use Divide[1] and MONTAGUE[9] provide an in-depth description

of active and passive users and indirectly confirm Caine's idea that accessing the brain feature

to make a Connection of information is the goal of a complex learning framework to narrow

the gap of digital use divide between the active and passive users.

Brain research supports learning needs to engage in reading, writing, viewing, listening,

talking, and valuing ([8], p.6). In other words, the goal is to apply the above six skills to

connect information. High school students spend years acquiring the above abilities in ELA

subjects. Their learning experience acquired through the six skills begins from K1 to K12.

Students have enough tools to learn new knowledge. However, when students start high

school, many struggle with STEM subjects, including computer science.

Every learning subject has its language to obtain feedback from the environment and acts

as a mediator between different conscious and unconscious systems internal or external to

human objects ([8], p.122). Computer Coding is a subject that has its own language,

computer programming language. Students are taught to write computer code to make their

Page 11
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

experience in learning computer science concepts understandable. Because the emphasis on

writing code in daily life is seldom emphasized in speaking code. Caine[8] clearly described

the ways to make a learner's experience understandable, speaking and writing together.

Speaking and writing are skills used in ELA subject learning. However, computer coding

seldom uses speaking skills. As described above in the problem statement, many school

learners need help attaining coding proficiency. The active research described in this

manuscript aims to study narrating to improve learners' coding proficiency. Even if students

practice code narrating, the quality and quantity of using relevant terms and appropriate

academic vocabulary and language will be studied.

Page 12
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

IV. Research Questions and Methods

IV.1 Research Questions

In the above LITERATURE REVIEW session, Caine[8, p.122) describes speaking and

writing as ways of making those experiences understandable. This manuscript study

investigates how narrating improves students' fluency in computer code development. The

data presented in the above PROBLEM STATEMENT states the disparity of students'

performance in learning computer coding. It is essential to understand the disparity's nature to

narrow the Digital Use Divide gap. We believe the human brain has the features to attain

experience and connect with previous experiences to learn new knowledge. In computer

coding, students write code in daily computer lab exercises, but speaking code is rare in class

or daily practice; it makes sense to wonder if speaking may be the fuse to support students in

reconnecting information in coding. Doyle & Carter [12] also describe the critical role of

narrating in making sense of experience. When students learn computer coding, new

computer language terms are imposed as extra abstract information in students' brains. Also,

the computer coding language is similar to the literacy language [13]; the programming

language primitives are still designed less naturally than the literacy language—the

disconnect between the experience attained from coding language and students' previous

experience must be bridged. Narrating appears to be neglected in the repository toolbox of

computer coding.

This study aims to conduct a series of experiments with the students around using

narrating in coding. The experiments are designed to generate results to answer the following

research questions:

Page 13
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

[RQ1] What are the quantitative measurement parameters?

[RQ2] What is the correlation between ELA and computer code narration?
[RQ3] How do we select narrating devices?
[RQ4] Are the selected narrating devices applicable in computer code narration?

More research is needed on code narration. The study is based on teachers’ experience in

subject contents and leverages literacy devices from other ELA subjects. The study collects

data by administering a survey, a pre-assessment to establish a baseline, and a post-

assessment after the narrating in coding to sampling students. The collected qualitative and

quantitative data are analyzed, and results are inferred to answer the five research questions.

IV.2 Research Method

Assessment Strategic Code Assessment

Baseline Narrating Benchmark

Speaking
Writing
Viewing
Listening
Valuing
Acting

Fig 1.1 ASA Experiment Framework

An experimental framework is proposed to study the impact of learning experience on

students’ coding development. The study in this manuscript focuses on narrating, which is

one of the ways to make sense of students’ previous experiences by connecting already

existing information in students’ brains. Studying what works well and what does not is a

continuous process. Dewey[10] and Quay[11] describe learning as a constant experience

Page 14
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

attaining; there is no best method in pedagogy. The way to do this is to continue the

experiments by trial and error to improve the learning and teaching experience. Dewey and

Quay’s advocacy is aligned with the Digital Use Divide[1], which centers around the issue of

inequitable access to active learning experiences. Sticking the principle of trial and error, a

study framework needs to be flexible to accommodate a variety of variables in the experiment

and fit into the high school class environment. The framework should also accommodate

students’ learning pace calendar without overloading students with extra assignments and not

disrupt a teacher's regular teaching hours. ASA stands for Assessment, Strategy, and

Assessment. Assessment can be in different forms and complexity; strategy is a particular

tactic applied in classroom pedagogy. There are two assessments. The second one is to

measure any impacts after implementing tactics.

Thirty sophomore students participated in the study. Six students from each of the five

sessions joined the study. Participants were characterized and grouped into low, medium, and

high performance in learning foreign languages, as in Table 2.1. We would like to see if there

is any correlation between students with different levels of foreign language and learning

coding language.

All participants Foreight Language Foreign Language Foreign Language Subtotal No. of
coding performance Low Performer Medium Performer High Performer participants
are <80% Group-L (<80%) Group-M (<90%) Group-H (>95%)
Session 1 2 2 2 6
Session 2 2 2 2 6
Session 3 2 2 2 6
Session 4 2 2 2 6
Session 5 2 2 2 6
Subtotal participants 10 10 10 30

Table 2.1 Students grouped into Low, Medium and High Performance in Foreign Language.

Page 15
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

IV.2.a Assessment Baseline

An assessment is designed to include 15 questions for students to answer. Google Form

is used to implement the 15 questions. Students self-evaluate their learning experience

addressed by the questions. There are two types of questions: the Coding Experience and the

ELA Literacy Experience. The Coding Experience questions help students to retrieve their

past learning experience in coding. As discussed in the above LITERATURE REVIEW

SESSION, a human brain can learn but needs an agent to interconnect the pre-deposited

information. Otherwise, students will find learning new materials, such as coding,

challenging. The agent is like the ELA literacy skills students have acquired in past years of

ELA subjects. Participating students self-evaluate their experience strengths using a scale of

1 to 5. The most substantial experience is rated a five. It is essential to note that the coding

experience differs from the ELA Literacy Experience. The intent is also to select students

with different backgrounds and foreign language experiences. Below is the list of assessment

questions. As introduced above, the research framework ASA is designed to be flexible and

fit into a high school teaching environment. Researchers can modify the set of questions to

experiment with different tactics.

Assessment
Assessment Code Assessment Questions Assessment Experience
Type
BA1 proficiency in narrating. Baseline Coding Experience
BA2 smoothness in narrating. Baseline Coding Experience
BA3 accuracy in narrating. Baseline Coding Experience
BA4 speed in narrating. Baseline Coding Experience
BA5 usage of CS vocabulary. Baseline Coding Experience
BA6 code tracing ability. Baseline Coding Experience
BA7 use of Iteration Baseline Coding Experience
BA8 use of Selection Baseline Coding Experience
BA9 use of Procedure Call Baseline Coding Experience
BA10 use of input parameters Baseline Coding Experience
BA11 use of arithmetic expression Baseline Coding Experience
BA12 use of boolean expression Baseline Coding Experience
BA13 use of Framing Story in Coding Baseline ELA Literary Experience
BA14 use of Cause and Effect in Coding Baseline ELA Literary Experience
use of Compare & Contrast in
BA15 Baseline ELA Literary Experience
Coding

Page 16
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Table 2.2 Assessment Baseline Questions and Types

IV.2.b Strategic Code Narrating

The experiment is launched during the month of the student’s independent project period.

Students can use this period to complete their project and also participate in using code

narrating to develop their program code. Students were briefed daily and participated in code

narration while developing their program code. The experiment lasted 15 days. Students were

provided a code segment to narrate every day in class. Participants started narrating with a

code segment length of 10, which increased gradually. On Day 15, the code segment length

was increased to 30 lines. Also, the complexity of the code segment increases gradually by

including computer science concepts of iteration, selection, and procedure. Students start to

use ELA Literacy devices such as ‘Framing Story,’ ‘Cause & Effect,’ and ‘Compare &

Contrast’ in their code narration. Students are expected to be immersed in the sea of ELA

Literacy and Code Literacy experiences, which make sense of their previous experience to

improve coding development proficiency. Table 2.3 demonstrates the progressive pacing of

infusing different elements in the daily code narrating activity.

Strategic Code Use Use Use Compare Include Include Include

Code Segment Framing Cause and & Contrast Iteration Selection Procedure
Narrating Length Story Effect
Day 1 10 No No No No No No
Day 2 10 No No No No No No
Day 3 10 No Yes No No No No
Day 4 15 No Yes No No No No
Day 5 15 No Yes No Yes Yes No
Day 6 15 No Yes No Yes Yes No
Day 7 20 No Yes Yes Yes Yes No
Day 8 20 No Yes Yes Yes Yes No
Day 9 20 No Yes Yes Yes Yes Yes
Day 10 25 No Yes Yes Yes Yes Yes
Day 11 25 Yes Yes Yes Yes Yes Yes
Day 12 25 Yes Yes Yes Yes Yes Yes
Day 13 30 Yes Yes Yes Yes Yes Yes
Day 14 30 Yes Yes Yes Yes Yes Yes
Day 15 30 Yes Yes Yes Yes Yes Yes

Page 17
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Table 2.3 ELA and Code Literacy Devices in Daily Code Narrating Activity

IV.2.c Assessment Benchmark

After 15 days of daily code narrating, students were asked to self-evaluate their

experience in the same scope in the Assessment Baseline. That same set of questions is used,

except the Assessment Code for each question is changed to reflect post-assessment

benchmark data. The goal is to investigate the difference in coding experience before and

after inhabiting code narrating. Because the same set of questions was used, a comparison can

be made with consistency between pre and post-data. Below is the set of assessment

questions with changes in the Assessment Code.

Assessment Code Assessment Questions Assessment Type Assessment Experience

PA1 proficiency in narrating. Benchmark Coding Experience
PA2 smoothness in narrating. Benchmark Coding Experience
PA3 accuracy in narrating. Benchmark Coding Experience
PA4 speed in narrating. Benchmark Coding Experience
PA5 usage of CS vocabulary. Benchmark Coding Experience
PA6 code tracing ability. Benchmark Coding Experience
PA7 use of Iteration Benchmark Coding Experience
PA8 use of Selection Benchmark Coding Experience
PA9 use of Procedure Call Benchmark Coding Experience
PA10 use of input parameters Benchmark Coding Experience
PA11 use of arithmetic expression Benchmark Coding Experience
PA12 use of boolean expression Benchmark Coding Experience
PA13 use of Framing Story in Coding Benchmark ELA Literary Experience
PA14 use of Cause and Effect in Coding Benchmark ELA Literary Experience
PA15 use of Compare & Contrast in Coding Benchmark ELA Literary Experience

Table 2.4 Assessment Benchmark Questions and Types

Page 18
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

V. Data Collection

Participating students were provided a Google Form to self-evaluate their coding and

ELA Literacy experience. Data from the Assessment Baseline and Assessment Benchmark

were collected and integrated into a single database for later data analysis. Students use a

scale of 1 to 5 to reflect their experience in coding and ELA Literacy. The numerals are

collected and counted to identify patterns and determine correlations between individual

coding experience and ELA literacy experience. Our goal is to discover which pairing of

experiences has a higher correlation. Although there is no best solution, we want to see if a

better choice is available based on research data and take further studies on it. [Appendix B]

is a sample of collected data in tablet format. Integrating both assessment data into one tablet

facilitates analysis later. When we want to analyze and find a correlation, an integrated

database facilitates more variety of correlation exploration between different questions.

The design of the database is also important. In [Appendix B], the table integrates

‘Assessment Group’ if classification exploration is needed. Database design is also part of the

ASA Framework. When experiments select other tactics [Fig 1.1], additional parameters can

be added to the database for analysis.

Page 19
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VI. Date Analysis

VI.1.a Proficiency in Narrating (foreign language high performer)

Proficiency in Narrating Foreign Language High Performer Effective

Total Total Effective

3.5 7
3 3 6
3 6
2.5 5
2
2 Total 4 Total

1.5 3
1 1 2
1 2
1 1
0.5 1
0 0
1 2 3 4 5 2 3 4 5
Chart6.1.a1 Baseline Chart6.1.a2 Benchmark

VI.1.b Proficiency in Narrating (foreign language low performer)

Proficiency in Narrating Foreign Language Low Performer Ineffective

Total Total Ineffective

4.5 3.5
4 3 3 3
4 3
3.5
2.5
3
2.5 Total 2 Total
2 2
2 1.5
1.5 1
1 1 1
1
0.5 0.5
0 0
1 2 3 4 5 2 3 4 5
Chart6.1.b1 Baseline Chart6.1.b2 Benchmark

Page 20
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VI.2.a Accuracy in Narrating (foreign language high performer)

Accuracy in Narrating Foreign Language High Performer Ineffective

Total Total Ineffective

4.5 4.5
4 4
4 4
3.5 3.5
3 3
2.5 Total 2.5 Total
2 2 2 2 2
2 2
1.5 1.5
1 1
1 1
0.5 0.5
0 0
1 2 3 4 5 2 3 4 5
Chart6.2.a1 Baseline Chart6.2.a2 Benchmark

VI.2.b Accuracy in Narrating (foreign language low performer)

Accuracy in Narrating Foreign Language Low Performer Effective

Total Total Effective

4.5 4.5
4 4
4 4
3.5 3.5
3
3 3
2.5 Total 2.5 Total
2 2 2
2 2
1.5 1.5
1 1 1
1 1
0.5 0.5
0 0
1 2 3 4 5 2 3 4 5
Chart6.2.b1 Baseline Chart6.2.b2 Benchmark

Page 21
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VI.3.a Usage of CS Vocabulary (foreign language high performer)

Usage of CS Vocabulary Foreign Language High Performer Effective

Total Total Effective

4.5 4.5
4 4 4 4
4 4
3.5 3.5
3 3
2.5 Total 2.5 Total

2 2
1.5 1.5
1 1 1 1
1 1
0.5 0.5
0 0
1 3 4 5 2 3 4 5
Chart6.3.a1 Baseline Chart6.3.a2 Benchmark

VI.3.b Usage of CS Vocabulary (foreign language low performer)

Usage of CS Vocabulary Foreign Language Low Performer Ineffective

Total Total Ineffective

3.5 4.5
3 4
3 4
3.5
2.5 3
3
2 2 2
2 Total 2.5 Total
2
1.5 2
1 1.5
1 1
1
0.5 0.5
0 0
1 2 3 4 5 2 3 4 5
Chart6.3.b1 Baseline Chart6.3.b2 Benchmark

Page 22
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VI.4.a Code Tracing Ability (foreign language high performer)

Code Tracing Ability Foreign Language High Performer Effective

Total Total Effective

4.5 4.5
4 4
4 4
3.5 3.5
3 3
3 3
2.5 Total 2.5 Total
2 2
2 2
1.5 1.5
1 1
1 1
0.5 0.5
0 0
1 2 4 5 2 3 4 5
Chart6.4.a1 Baseline Chart6.4.a2 Benchmark

VI.4.b Code Tracing Ability (foreign language low performer)

Code Tracing Ability Foreign Language Low Performer Ineffective

Total Total Ineffective

4.5 6
4
4 5
5
3.5
3
3 4
2.5 Total 3 Total
2 3
2
2
1.5 2
1
1
1
0.5
0 0
1 2 3 4 2 4 5
Chart6.14b1 Baseline Chart6.4.ab Benchmark

Page 23
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VI.5.a Use of Cause & Effect (foreign language high performer)

Code Tracing Ability Foreign Language High Performer Effective

Total Total Effective

4.5 4.5
4 4 4
4 4
3.5 3.5
3
3 3
2.5 Total 2.5 Total
2 2
2 2
1.5 1.5
1
1 1
0.5 0.5
0 0
1 2 3 4 2 3 5
Chart6.5.a1 Baseline Chart6.5.a2 Benchmark

VI.5.b Use of Cause & Effect (foreign language low performer)

Code Tracing Ability Foreign Language Low Performer Ineffective

Total Total Ineffective

4.5 4.5
4 4
4 4
3.5 3.5
3
3 3
2.5 Total 2.5 Total
2 2 2
2 2
1.5 1.5
1 1 1
1 1
0.5 0.5
0 0
1 2 3 4 5 2 3 4 5
Chart6.5.b1 Baseline Chart6.5.b2 Benchmark

Page 24
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VII. Conclusion

Experiment results show code narrating has improved students’ code development in

some areas but not every area. After practicing three weeks of code narrating, I noticed

different degrees of impact on students of high performers and low performers in foreign

languages. For students who are high performers in foreign languages, data shows that code

narrating improves their experience in proficiency in coding, usage of computer science

vocabulary, and code tracing ability. On the other side, code narrating is ineffective for

students who are low performers in foreign languages. Students who are high performers in

foreign languages show no improvement in accuracy; on the contrary, code narrating

improves code accuracy for students who are low performers in foreign languages.

Foreign Language High Foreign Language High

Performers Performers
Proficiency in Narrating Effective Ineffective
Accuracy in Narrating Ineffective Effective
Usage of CS Vocabulary Effective Ineffective
Code Tracing Ability Effective Ineffective
Use of Cause & Effect Device Effective Ineffective

Table 7.1 Effectiveness of Code Narrating

At the outset of this experiment, according to the theory and information collected from

the Literature Review, it is initially believed that Code Narrating should improve students’

code development. After a more detailed study, the claim is primarily accurate but not

confirmed. As we can see from the results of the Data Analysis session, only students with

access to a good learning experience of a foreign language can benefit from code narrating.

Code narrating seems ineffective for students who do not have access to an excellent foreign

language experience.

Page 25
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

The results from the data analysis answer the first research question, RQ1. Each

assessment question is a parameter to access students' experiences in a specific context. There

are patterns identified in VI.1a to VI.4.b to make sense of the effectiveness or ineffectiveness

of code narrating. The parameters or questions set can be modified to a particular context to

reflect another set of parameters for use in another experiment. The study provides evidence

of a correlation between coding and the cause & effect of devices from ELA. However, the

correlation is valid in the presence of high foreign language performers. It answers positively

to RQ2. Narrating devices are from the ELA literacy context. Because of the flexibility of the

ASA framework, any ELA literacy devices can be selected for use. Considering the learners'

background, collaboration with ELA subject teachers will bring invaluable insight into the

choices of devices. Because we are using students' time in the implementation of new ideas,

we should start devices that are simple and consume less time for students to execute. Three

selected devices were used to collect data, the result in VI1.5.a shows a strong relationship.

Although the data analysis of the other two devices is not available yet, as they are similar, it

is believed that the selected devices are applicable in code narrating to support answering

RQ4.

The results of the experiments demonstrate the importance of instructional

differentiation. We may not use instructional time wisely if we emphasize code narrating to

low-performing foreign language learners. However, if we differentiate the instruction for

high-performing foreign language learners, their code development skills can be improved

significantly.

As Dewey[10] and Quay[11] advocate, education is a continuous trial and error to access

a better learning experience, and there is never an optimal solution. The flexibility of the

Page 26
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

ASA experiment framework demonstrates how we can trade in other tactics quickly to find a

solution to serve the still under-performing students in the coding adventure.

Page 27
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

VIII. Next Steps & Reflections

The study results show that narrating is not a 100% remedy for students' improvement in

coding. In Caine's [8] vision, viewing is another alternative feature of the human brain to

make sense of information. The next step can phase in the viewing instead of code narrating

in the ASA Framework. A set of new questions should be developed for data collection and

verification. We can keep using the same participant selection methodology. Instead of based

on foreign language performance, it can be based on other subjects or spot performance for

grouping students in a new study.

If we select viewing as another tactic instead of code narrating, we may want to consider

students' background in performance art. Active research is a collaborative and reflective

practice; engaging other teachers in collecting feedback is critical. Another teacher may have

insight into the factors affecting students' learning experience, but those teachers may need a

feasible research framework to study the problem. When I started working on this project, I

learned that the research effort should be sustainable, which means the research process and

methodology can be reused. This will save an enormous amount of time when crafting the

outline.

Finding a correlation between two variables is very time-consuming and limits the scope

of exploration in terms of data analysis. Visual comparison of bar charts to determine the

effectiveness of code narrating could be more manual. I am not aware of any other way to do

it. Further research is needed to determine what other research will do to minimize these

tedious efforts.

Page 28
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Only a few statistics skills are used in this research, and only some bar charts are used to

compare and contrast. I appreciate the data collected, which shows me the effectiveness and

ineffectiveness through the bar charts. Statistics skills should be used to help uncover more

hidden details, such as inconsistencies found in the studies.

Reflecting on the project idea and its development, I should have paid too much attention

to the mid-term presentation. When I started to collect data and wrap up the writing of the

final paper, the mid-term presentation helped me a lot with the outline and guidelines.

Following the mid-term outline, I can smoothly swirl through the final paper writing. I also

undermine the importance of the Literature Review. When I started writing my paper, I found

my initial literature review needed to be more profound, and I needed to look for more

insightful articles and journals. Although it took me much more time and effort to go through

the Literature Review process again, it was invaluable. I could locate several good papers that

helped me develop my research idea. I would not know more about John Dewey if I had not

read about the critics of John Dewey by Quay.

I also thank you for my course professor's insight guidelines. Without her advice, I might

be running my project with 200 participants, which would definitely expand the scope, and I

would probably not be able to finish the study within the thin scheduled time frame. Let that

start a bit limited. The scope contributes to frustration and over expansion of ideas, ideas, and

effort.

Page 29
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

References

[1] The Digital Use Divide. Office of Educational Technology, Department of Education.
https://tech.ed.gov/netp/digital-use-divide/

[2] LADSON-BILLING, G., From the Achievement Gap to the Education Debt:
Understanding Achievement in U.S. Schools. Educational Researcher, Vol. 35, No. 7
(Oct., 2006), pp. 3-12.

[3] CULP, K., HONEY,M., & MANDINAC, E. (2003, October). A Retrospective on Twenty
Years of Education Technology Policy. U.S. Department of Education, Office of
Educational Technology.

[4] National Commission on Excellence in Education. (April, 1983). A Nation at Risk.

Available online at http://www.ed.gov/pubs/NatAtRisk/risk.html

[5] New Study on U.S. Eighth-Grade Students’ Computer Literacy, (2019, Nov.) . National
Center for Education Statistics.

[6] AINLEY, J. (2018). Students and Their Computer Literacy: Evidence and Curriculum
Implications. Springer International Handbooks of Education. Springer, Cham.
https://doi.org/10.1007/978-3-319-71054-9_4

[7] New York State Computer Science and Digital Fluency Learning Standards GRADES K-
12 December 2020. NYCDOE

[8] CAINE, R., CAINE, G., Making Connections: Teaching and the Human Brain.
Association for Supervision and Curriculum Development, Alexandria, Va.

[9] Montague E, Xu J. Understanding active and passive users: the effects of an active user
using normal, hard and unreliable technologies on user assessment of trust in
technology and co-user. Appl Ergon. 2012 Jul;43(4):702-12. doi:
10.1016/j.apergo.2011.11.002. Epub 2011 Dec 20. PMID: 22192788; PMCID:
PMC3422026.

[10] Dewey, J. (1965). Experience and Education. New York: Collier.

[11] Quay, J. John Dewey’s Conceptualization of Experience. Experiential Learning and

Outdoor Education (2020) Chapter 6. Routledge, Taylor & Francis Group. ISBN: 978-
0-429-29880-6 (ebk)

[12] Doyle,W., Carter, K., "Narrative and learning to teach: Implications for teacher-
education curriculum", Journal of Curriculum Studies, vol. 35, no. 2, pp. 129-137,
2003.

[13] Mani,I., Computational Narratology(2013), The Living Handbook of Narratology

(http://www.lhn.uni-hamburg.de).

Page 30
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

[14] WOO, K., FALLOON, G., The Search for Computer Science Concepts in Coding
Animated Narratives: Tensions and Opportunities. Volume 61, Issue 7,
https://doi.org/10.1177/07356331231174929

Page 31
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Appendix A - Survey on Motivation to Learn Computer Science

CS Learning Motivation 2021(%) 2022(%) 2023(%)

Pass the course 20% 24% 23% Passive User
S School Requirement 39% 34% 34% Passive User
Parents' advise 7% 8% 6% Passive User
Want to build video game 5% 4% 4% Active User
Want to build social media 4% 5% 4% Active User
Want to build personal web-page 6% 4% 5% Active User
Friends are taking CS courses 16% 16% 17% Passive User
Apply CS in other subjects. 3% 3% 4% Active User
Survey returned (%) 99% 98% 98%
Survey not returned (%) 1% 2% 2%
Number of survey returned 190 178 186
Number of survey not returned 2 3 4

Table1.1 Students from 2021 ~2023 attending AP Computer Science Principles class.

Page 32
 A Study on How Narrating Improve Students’ Fluency in Computer Code Development

Appendix B - Assessment Data

Table 5.1 Assessment Baseline Raw Data

Table 5.2 Assessment Benmark Raw Data

Page 33