BRIEF RESEARCH REPORT

published: 13 June 2022

doi: 10.3389/fpsyg.2022.868698

Integrating Multiculturalism Into

Artificial Intelligence-Assisted
Programming Lessons: Examining
Inter-Ethnicity Differences in
Learning Expectancy, Motivation,
and Effectiveness
Chia-Wei Tsai 1 , Yi-Wei Ma 2 , Yao-Chung Chang 1 and Ying-Hsun Lai 1*
1
Department of Computer Science and Information Engineering, National Taitung University, Taitung City, Taiwan,
2
Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei City, Taiwan

Given the current popularization of computer programming and the trends of

Edited by:
Yu-Sheng Su,
informatization and digitization, colleges have actively responded by making
National Taiwan Ocean University, programming lessons compulsory for students of all disciplines. However, students from
Taiwan
different ethnic groups often have different learning responses to such lessons due
Reviewed by:
to their respective cultural backgrounds, the environment in which they grew up, and
Jun Gu,
Communication University their consideration for future employment. In this study, an AI-assisted programming
of Zhejiang, China module was developed and used to compare the differences between multi-ethnic
Danping Lin,
Shanghai Maritime University, China
college students in terms of their theoretical and actual learning expectancy, motivation,
*Correspondence:
and effectiveness. The module conducted analysis through the deep learning network
Ying-Hsun Lai and examined the relevant processes that the students underwent during programming
[email protected]
lessons, as well as the types of errors they had committed. Their learning motivation
Specialty section: for and actual learning performance in programming were then examined based on the
This article was submitted to cognitive learning theory. The results of the experiment, which involved 96 multi-ethnic
Educational Psychology,
college students, indicated that the two groups had dissimilar theoretical performance
a section of the journal
Frontiers in Psychology in terms of their expectancy and motivation for learning programming. The indigenous
Received: 03 February 2022 students’ main concern was whether programming would affect their families or tribes,
Accepted: 25 May 2022 and this concern affected and was reflected in their learning outcomes. In contrast, the
Published: 13 June 2022
learning motivation and goals of Han Chinese students were driven by the cognition of
Citation:
Tsai C-W, Ma Y-W, Chang Y-C the value of programming to themselves. The research findings can contribute toward
and Lai Y-H (2022) Integrating the cognition and understanding of multi-ethnic students when learning computer
Multiculturalism Into Artificial
programming and development of the appropriate teaching methods, and serve as
Intelligence-Assisted Programming
Lessons: Examining Inter-Ethnicity a reference for subsequent research on integrating multiculturalism into computer
Differences in Learning Expectancy, programming lessons.
Motivation, and Effectiveness.
Front. Psychol. 13:868698. Keywords: multiculturalism, program education, expectancy-value theory, AI-assisted program, learning
doi: 10.3389/fpsyg.2022.868698 effectiveness

Frontiers in Psychology | www.frontiersin.org 1 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

INTRODUCTION skills, identify their problem-solving abilities, and stimulate their

curiosity through the problem-solving process. Other studies
With scientific and technological developments, the application showed that the process of learning programming enhances
of information technology (IT) has gradually become inseparable critical thinking by finding ways to think about problems. The
from the various aspects of human life like food, clothing, main goal of these IT or programming lessons is not only to
housing, transportation, education, and entertainment. This give trainees the basic knowledge on how to write programs,
situation is exacerbated by the latest trends of the Internet but also to train them to improve their own living conditions by
of Things (IoT) and artificial intelligence (AI), such that reflecting on their own unique experiences, as well as identify
programming languages have become a common language and the thinking and learning methods that best suit them through
basic capability of the 21st century (Essel et al., 2017; Nouri constantly learning from mistakes. This form of educational
et al., 2020). In response, countries around the world have spirit and proposition in non-professional fields helps strengthen
formulated the related education policies. Bill Gates (2008) from the spirit of innovation and competition (Juškevičienė et al.,
the United States advocated computer programming as “a basic 2021; Lai et al., 2021).
skill in the 21st century that every student should learn.” In Globally, there are many related and supporting educational
2016, then-President Obama emphasized that in the AI era, IT course designs intended to train and cultivate in students
capabilities have become a new set of general skills that all citizens the relevant abilities for future competitiveness and problem
must possess. He also signed the “National Artificial Intelligence solving through programming. For example, Li (2018) proposed
Research and Development Strategic Plan,” which designated combining programming with the flipped classroom so that
computer science as a key academic field and general capability students cultivate self-training, and then learn the basic
(U.S. National Science and Technology Council, 2016). Schools knowledge through self-directed learning. The students’ overall
were also encouraged to include IT science as a subject in their learning motivation would then be effectively enhanced.
foundational curriculum. Moreover, with problems in life becoming increasingly complex,
In the United Kingdom, programming was included in the computational thinking can be used to effectively analyze
curriculum in 2014. The aim was to teach children programming complex issues. Problems can be solved with computational
languages from an early age, with the requirement being the concepts, and humans’ problem-solving behaviors can be
mastery of at least two such languages (Sangwin and O’Toole, analyzed and understood. Computational thinking is not
2017). In South Korea, public junior high schools introduced about making humans think like computers. Rather, it is
programming lessons in their first-year curriculum since 2018 to about the full set of thinking tools needed to effectively
ensure that the students learn about IT from an early age. The solve complex human problems. Barr and Stephenson (2011)
purpose goes beyond teaching students how to use computers; subdivided the capability dimension of computational thinking
they are taught the way computers work and about structured and introduced it into other courses (such as mathematics,
processes, to develop their analytical and problem-solving skills science, social studies, language, and art) to examine the
(Hsu et al., 2018). Japan announced its “Future Investment concept of integrating the cultivation of computational thinking
Strategies” in 2017, which incorporated programming into the abilities, as well as the corresponding goals in terms of
curriculum of elementary and junior high school education, learning behaviors.
together with the development of teaching materials and There have also been research that applied several
systems for programming (Kanemune et al., 2017). The overall programming-based activities to support learners’ reading
educational goal was to cultivate thinking in programming and abilities, and for them to write using a programming language
solve problems through IT. and engage in computational thinking. In view of the emphasis
In 2019, programming lessons were officially added to the on programming lessons in the 12-year national education
school curriculum in Republic of China (Taiwan). Elementary curriculum, many studies have highlighted the critical issues
and junior high schools were required to train their students that must be overcome (Gushwa et al., 2019; Lee et al., 2019).
in programming skills and give them the foundation to acquire In terms of the software and hardware in the educational
the relevant IT skills through computer-based learning. This environment for programming, schools need to upgrade the
way, their systematic thinking abilities are cultivated and their relevant IT equipment and set up new computer classrooms so
logical computing skills are strengthened, thereby improving that the learning needs of the overall school for IT education
their problem-solving skills, teamwork, and innovative thinking are met. In terms of teaching resources, it is crucial to cultivate
abilities through their own training capabilities (Lin, 2019; more teachers in IT programming to meet the demand.
Saputro et al., 2019). Colleges have also adopted basic computer It is even more necessary to introduce new concepts to
programming as a general educational course for students in students and parents so that they stop being obsessed with
all faculties, not only those in science and engineering. One the expectations of traditional education, namely, prioritizing
targeted outcome is the search for ways of thinking through inter- pure academic performance while ignoring the cultivation
disciplinary integration; another is to develop students’ non- of practical and operational skills and character building.
programming abilities through training and classes in computer In higher education, there must be an understanding of the
programming (Gong et al., 2020). different ways that students from various faculties regard
The relevant studies have pointed out that learning programming lessons. These should be introduced based
programming-related skills help learners improve their analytical on their respective professional needs so that they take

Frontiers in Psychology | www.frontiersin.org 2 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

to such lessons. An overall multi-professional curriculum environment that enhances students’ interest and outcomes
design should be applied when cultivating and strengthening when learning programming.
IT literacy. Emphasis should also be placed on the use of
the computational thinking method for problem solving,
which can double up as potential training to ready them for LITERATURE REVIEW
entering the workplace (Lye and Koh, 2018; Gürer et al., 2019;
Scherer et al., 2020). Multiculturalism and Multi-Ethnic
Against the trend of popularization of informatization, many Education
indigenous families and tribes continue to live in remote areas Culture is defined in anthropology as the consensus of a
due to the influence of history and terrain. This often results specific group of people regarding the sum total of their values,
in their relative lack of economic resources and informatization cognitions, codes of conduct, beliefs, customs, and material
level. Indigenous peoples can be further divided into tribes that life. Multiculturalism is a critique of the emphasis on ethnic
dwell in the mountains or flatlands, with the former being equality by liberalism without considering cultural differences,
especially lacking in their informatization level (Tzou et al., 2019; which results in the mainstream culture assimilating or rejecting
Janica, 2021). Moreover, indigenous students often lag behind in marginal cultures and the latter being in a disadvantageous
their learning performance and have a low willingness to learn position. Multiculturalism advocates not only universal equality
because of the influence of innate environmental factors and and non-assimilation, but more importantly, the recognition
differences in cultural cognition. Their lack in informatization of one another’s cultural differences (Modood and May, 2001;
skills in turn lead them to choose the service or nursing industry Race, 2015). This is based on the spirit of multi-ethnic groups
after graduation, and only a minority undertakes IT-related being recognized, so that the rights to their unique cultural
work (Lopez and Bobroff, 2019). This situation becomes a characteristics are maintained and preserved.
vicious cycle because there are few indigenous students who Multicultural education is an indispensable key in education
are able to bring the IT knowledge that they have learned to create the awareness on multi-ethnic groups. The key goal is to
back to their hometowns to drive the advancement of their have the participation and integration of multi-ethnic students as
tribe’s informatization. In the long run, the urban–rural digital the basic criterion during classroom teaching activities (Ladson-
gap and the indigenous tribes’ lack of informatization will Billings, 2004; Banks and Banks, 2019). In many countries around
keep widening. the world, research on education in science have gradually
In the spirit of diversity, equality, autonomy, and respect, improved by taking into consideration cultural and ethnic
the relevant government departments have been promoting characteristics. The perspective of multiculturalism has also been
education-related efforts tailored to the indigenous peoples. The adopted to identify appropriate teaching methods and materials
aim was to establish a cultural model for indigenous education, when teaching science to indigenous people, which involve
strengthen self-identity and acceptance, and avoid the decline incorporating their perspectives or acquired knowledge (Writer,
and loss of their unique cultures (Battiste, 2018). However, 2008).
there is insufficient awareness of the cultural differences and There are various approaches to multi-ethnic educational
perception of academic achievements. Many efforts wanted reforms internationally. These aimed at the establishment and
to take into account the tribes’ cultural characteristics and maintenance of a multi-ethnic educational culture through the
level of cultural acceptance, causing such efforts to fail setting up of relevant systems, participation of multi-ethnic
to keep up with the overall trends (Eades et al., 2022). groups in educational decision-making, ensuring educational
Therefore, the focus of research in multi-ethnic IT and equity, and development of multicultural curriculum designs
programming education should be finding ways of using AI (Nikolaeva and Savvinov, 2016). Teachers bear an important
assisted-learning systems to effectively identify the cognitive responsibility in the transformation of multicultural education.
understanding and learning performance of students from A key factor is maintaining respect for and acceptance of
different ethnicities with regard to learning programming, and multiculturalism, as well as establishing a curriculum that is
then establish an adaptive programming framework that matches suitable for application in the context of multi-ethnic world views
the multiculturalism situation. and values (Blair, 2002; Denessen et al., 2007).
To further understand the relationship between learning Some studies on the way indigenous people learn pointed
performance and cognition of programming lessons among out that the important reasons behind their poor academic
multi-ethnic students, an AI-assisted programming module performance can be classified into the six categories of “cultural
was designed to record the students’ learning process during factor,” “parents’ socioeconomic status,” “racial discrimination,”
programming lessons, which facilitated analysis of their “issues pertaining to educational funding,” “issues with teaching
performance. The cognitive learning framework was used as resources,” and “issues with students’ abilities”(Duff and Li, 2009;
the basis of this qualitative study to examine the impact of Nikolaeva and Savvinov, 2016; Schwab, 2018). Among these,
multi-ethnic students’ connection between self-expectancy, the cultural factor relates to the fact that current educational
effort expectancy, and social value on their overall learning courses and teaching materials are still designed with emphasis
performance. The research findings were used to construct on the mainstream culture but lack design considerations for
a theoretical foundational model for multi-ethnic education multi-ethnic groups. Consequently, when students of marginal
in IT programming, so as to create a multicultural learning ethnic groups participated in the related courses, they often

Frontiers in Psychology | www.frontiersin.org 3 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

had a sense of psychological estrangement and uncertainty over problems, which is a means of cultivating self-thought. It can be
the future. Because they did not understand how learning a explained in terms of intrinsic and extrinsic abilities (Ezziane,
course could help them realize their definition of self-worth, they 2007; Sulisworo and Suryani, 2014). Intrinsic ability refers to
often generated negative learning emotions and produced poor the ability to think about and clarify problems, analyze the
academic performance. The main factor behind the resolution information needed in the current environment, and interpret
of this situation is the establishment of multicultural education. and reorganize information into useful facts through analysis.
It is only through building a multi-ethnic educational and Extrinsic ability is knowing where to find the information and
instructional design that integrates the multi-ethnic students’ how to obtain it, and using one’s self-learning ability to obtain
tribal experiences and values as part of the overall design that they information with which to solve the related problems. Students
can identify with the curriculum (Sylva et al., 2010). not only need to enrich their overall technological literacy, but
The main bottleneck for multi-ethnic education is having should also have complete IT literacy encompassing the three
insufficient relevant basic research. Currently, most practices dimensions of knowledge, ability, and mode of thinking and
in the fields of sociology and language have responded to the action (Brandt, 2001; Markauskaite, 2006).
tribalism of multi-ethnic students. However, in terms of teaching Indigenous students are disadvantaged by their relative lack
practice in fields related to science and IT, it is difficult to of living resources and cultural education, which makes it even
respond using a culture of equality. Haberman and Post (1998) more pertinent for them to cultivate the relevant scientific
mentioned that different ethnic groups should be encouraged to and IT literacy (Martin, 2006; Perso and Hayward, 2020). In
design localized multicultural courses according to their cultural addition to having the ability to master and control scientific
tendencies. Berry and Kalin (1995) also echoed that lessons on and technological developments, they can use such capabilities
culture, language, and multiculturalism should be implemented to overcome limitations when seeking employment and enhance
in parallel during multicultural teaching. Doing so will ensure their employability (Pramuda et al., 2019). Moreover, they need
that various ethnic groups have equal opportunities to participate to use their own technological thinking abilities and creativity to
in cultural and other activities, on the premise of multiculturalism address the difficulties faced by the various tribes and promote
being respected and appreciated, and without their own cultural their overall development based on their respective situation and
characteristics being lost. conditions (Li et al., 2021).
With the popularity of AI technology in recent years,
programming lessons have gradually evolved from technological
Information Technology Literacy and literacy to technological and programming literacy. Students’
Programming Capabilities creativity can be stimulated through the process of learning
With technological advancements, people have begun changing programming, allowing them to build their self-learning
their living habits in response to technological changes. Physical ability, exercise their logical thinking ability, cultivate their
interactions in the forms of chatting, writing letters, and shopping computational thinking, and hone their problem-solving ability
have gradually transformed to be based on computers and (Serin, 2011; Grover and Pea, 2013). With life’s problems
mobile phones. In the field of education, there has also been becoming increasingly complex, computational thinking can be
the gradual promotion of technological and digital education. applied to effectively analyze complex issues. Making operational
Mobile computer devices allow learners to undertake learning concepts process-oriented can also help solve problems and
anytime and anywhere, and the teaching and learning methods be applied to analyze and understand humans’ problem-
involved have gradually become increasingly reliant on computer solving behaviors.
technology (Marcum, 2002). However, such a novel and Computational thinking is a fundamental problem-thinking
impactful way of learning is actually associated with the key factor skill that encompasses the entire thinking process and involves
of socioeconomic status. After all, such learning opportunities the formulation of problems and solutions. Computational
are often available only to families or regions of a relatively high thinking can be used as a standard solution for overall problems
socioeconomic status. In contrast, students from low-income via computer operations or programming (Aho, 2012; Shute
households, marginal groups, and even indigenous tribes have et al., 2017). However, it goes beyond that and is actually an
difficulty accessing such learning opportunities (Loyer, 2018). essential skill for everyone, not just information or computer
To popularize the relevant IT training in society, many scientists, or even students who major in computer science.
colleges and tertiary institutions have begun introducing IT This is because it gives students of all disciplines the requisite
literacy as a general education lesson. This is so that college ability to incorporate computing techniques and methods when
students gain a holistic education to respond to IT developments. solving problems. Barr and Stephenson (2011) subdivided
Kosslyn et al. (2007) pointed out that general education not the competency dimension of computational thinking and
only strengthens the value of self-existence, but also encourages introduced it into other courses (such as mathematics, science,
students to critically think about their own environments social studies, language, and art) to examine the concept of
against the backdrop of the developments achieved by human integrating the cultivation of computational thinking abilities, as
civilization. They should also have the ability to deal with relevant well as the corresponding goals in terms of learning behaviors.
issues and develop problem-solving skills (Serin, 2011). Some research has found that combining it with the concept
Information technology literacy refers to empowering learners of computational thinking can effectively make it easier for
with the ability to use IT to pursue knowledge and solve students who are not in the professional IT field to understand

Frontiers in Psychology | www.frontiersin.org 4 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

the method of programming operations, thereby arousing their expectancy of learning success. Emotional motivation refers
interest in acquiring programming skills (Barr et al., 2011; to how much learners like the behavior of learning, or their
Mannila et al., 2014). emotional response during learning. MacIntyre and Blackie
(2012) introduced another category of learning motivation in the
form of volitional motivation, which refers to whether learners
RESEARCH METHOD persist and implement the learning behaviors acquired during
the learning process. Indigenous students often live adjacent to
Presenting the Research Model their fellow tribespeople and they learn the appropriate behaviors
The aim of this study was to make a theoretical assessment of and actions by observing the latter. Although this environmental
multi-ethnic students’ learning expectancy and motivation for element is not affected by the teaching materials that the students
programming lessons, which affect their learning performance come across, it produces in them the phenomena of self-
for the subject. The cognitive development theory was used as the motivation and -adjustment. People construct the corresponding
basis to describe the process of their knowledge acquisition, and educational methods based on their observations and through
their performance at and changes during information processing. self-learning (Bandura, 1989). The overall perspective also has
It includes the following periods: sensory action, pre-operation a relative impact on indigenous students’ expectancy through
thinking, specific operation, and formal operation (Gopnik and organizational expectancy and industrial structure, forming a
Wellman, 1994; Huitt and Hummel, 2003). Cognitive growth has tripartite mode of interaction.
different effects on students’ various abilities. Cognition must The social cognitive career theory (SCCT), which was derived
be the building block of self-learning and the formation of self- from this form of social cognitive theory, also affects learners’
cognition. The experience of others can only serve as reference career choices. During research on students’ interests in different
or past anecdotes during learners’ exploration process. According fields, additional considerations are given to the disguised forms
to research, there are some adults who have yet to reach the of learning interests, achievement abilities, and goal selection
formal operation period. They are unable to achieve that stage (Lent et al., 2002). In this study, the SCCT was used to examine
because the conditions when they were growing up prevented the impact of indigenous students’ cognitive motivation when
them from learning about, or allowed them to avoid contact with, they learn programming.
new circumstances and objects. This situation is also more likely
to occur during the learning situation of students from marginal The Module Used for Analysis
cultures. This was a phenomenon that needed evaluation in this The module developed in this study examined the students’
study because it affects the teaching of multicultural students. situation when they learn programming in order to analyze their
Educational psychologists have pointed out that different learning performance. In this study, Python 3.0 was adopted
people have dissimilar learning motivations and strategies. as the standard programming language for learning and used
For example, some students prefer to find the answers by to collect data during programming lessons. Typically, when
themselves, whereas others want their teachers to give them students learn to write programs, they would conduct functional
prompts so that they can solve the problems quickly. Their self- validation after their programs have been compiled by a compiler,
learning expectancy and motivation affect the overall learning and then perform unit testing to ensure that what they have
factors. Learners can improve their overall learning efficiency learned is executed correctly.
by changing their learning motivation, which then enhances When learning programming, learners will encounter various
their learning effectiveness. Learning motivation can be divided erroneous situations during the compiling stage. These include
into physiological and psychological motivation. Physiological syntax or logic errors including writing errors for the program’s
motivation is the primordial need that drives one to engage functional module and failures in the definition of variables,
in learning behaviors; psychological motivation is based on which prevent the compiler from completing its execution. Its
self-will, motivation for achievement, and expectancy. Research platform mainly captures and compiles records of failures. After
on learning motivation is predominantly focused on analyzing obtaining the compilation of recorded failures, a textual search
psychological motivation. Many studies have pointed out that was performed and the erroneous information was recorded.
when students feel that the relevant knowledge and cognition are Next, text mining was used to search for the relevant syntax error
of value or can help them gain a sense of achievement with their in the “substitute search word,” before comparing the possible
group, their intrinsic motivation is raised. This increases their errors for that word generated from the database of line numbers
positive learning attitudes, which include level of commitment, and keywords, and categorizing those errors. The other check was
awareness of self-worth, satisfaction level with the teaching, and to introduce an empty function set to generate a data defect, so
positive feedback between teachers and students. that the detection of its internal error and address could produce
Pintrich et al. (1993) first proposed the theoretical model of the empty data error function before comparing its line number
expectancy and motivation, and conducted research on learning and string characters.
through the three main motivations of value, expectancy, and The unit test, which is performed after the basic compilation
emotion. Value motivation refers to the value obtained by is completed, analyzes the Python code through Pylint. Its
learners’ interaction with the relevant learning objectives, and analytical suite uses the PEP 8 coding style for detection, and
is their reason for learning. Expectancy motivation refers to lists whether the codes being detected and validated have violated
the ability that learners expect to gain after learning, or the any PEP 8 specifications or contained common errors. The data

Frontiers in Psychology | www.frontiersin.org 5 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

being reported are labeled as C and W for convention and and indigenous people, in information programming courses.
warning, respectively. There are six levels of Pylint results in The subjects were aged between 18 and 22, with a male to female
total. After using Pylint to validate the results of executing the ratio of 80–16, and all had basic information programming skills.
codes, information of the relevant results and line numbers are Before each class, teachers will introduce the basic functions
recorded. Finally, these are matched with the coding coverage to of this course to ensure that each participant has basic ability
predict the program’s operating or functional errors. Detection and knowledge. Data collection and survey with cognitive
of the coding coverage is mainly to confirm whether the overall questionnaires were conducted over six programming lessons.
codes allow the coding process to be executed completely and for The purpose was to examine the relationship between the
purpose of memory validation. These form the basis for further expectancy of self-worth of students from multi-ethnic groups
estimating whether the codes will generate more errors. and their learning motivation and process, as well as understand
There are many tools for detecting coding errors that use their performance when learning programming. PLS-SEM was
error location methods to identify the corresponding errors. used to analyze the relevant data because of the small sample
In this study, the coding coverage and Pylint detection results size (Hair et al., 2017). Given that there were six indicators
were used to predict the programming results. The module in this study, 60 samples was the minimum number required
compares the results with pre-loaded correct templates, and then to meet the evaluation criteria according to the 10 times rule
use path analysis to predict and compare the two execution (Chin and Newsted, 1999).
paths, and determine the coverage and number of warning lines.
After deconstructing the entire program, it will determine the
status of the results. The module records the user’s overall time Analysis of the Results Recorded by the
spent on programming practice, completion of the functions, Module
number of compilations made, corresponding correction data, The module recorded each student’s process of learning program,
data coverage, and program check line count. their correct answers (error rate), answering time, and the
In each lesson, the students were given 30 min to write a number of syntax errors, logic errors, and warnings. First, the
functional program for the IoT. Items with syntax errors and correct rates for completing the programming task at various
the related compilation time and type of errors were recorded response times were analyzed shown as Figure 1. The correct
through the program’s compiler interface. Items with logic rate at 5 min was only 2%. This was because the students in this
errors and warnings were analyzed using the program’s logging category had given up on the task at the early stage and simply
function and Pylint. made their submissions. The majority of the students completed
their task in approximately 20–25 min, with the correct rate
being approximately 71–85%. When the response time reached
RESEARCH FINDINGS 30 min, the correct rate dropped to 60% and the standard
deviation (SD) increased to approximately 12%. Upon checking,
In this study, 96 students who studied in information engineering this was determined to be caused by those students who had not
were invited as participants, with 62 and 34 being Han Chinese completed the task but continued working on it.

FIGURE 1 | Statistics on correct answer rates.

Frontiers in Psychology | www.frontiersin.org 6 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

A detailed analysis was made of the recorded errors Next, bootstrapping (with 5,000 iterations) was used to
corresponding to the various time points and the statistical results calculate the related t-value, p-value, and R2 value. The overall
are shown in Figure 2. The number of syntax errors gradually results are shown in Figures 3, 4, which indicate that their
decreased with time, but the number of logic errors did not emotional cognition did not correspond to the test results for
decrease significantly. There were more warnings at the 15– learning motivation. In the figures, ∗∗ and ∗∗∗ means the
20-min period, but these gradually decreased at the 25–30-min ∗ t > 1.96 (p < 0.05), ∗∗ t > 2.58 (p ≤ 0.01), and ∗∗∗

period. The SD decreased significantly at the 30-min time point. t > 3.29 (p ≤ 0.001). There were also differences between
A possible reason was the students making a final attempt to the Han Chinese and indigenous students in terms of their
check with one another before the allocated time was up. expectancy cognition. For Han Chinese students, the relevant
indicators of learning motivation are value cognition, self-
expectancy, and social expectancy. Especially in the value
Learning Motivation and Effectiveness of cognition, the t value of the influence on motivation is 4.642,
Han Chinese Versus Indigenous Students which is more significant than self-expectation and social
The learning motivation and performance of Han Chinese and painting expectations; the path coefficient is 0.2424, which is
indigenous students were analyzed in terms of values, expectancy, similar to the social expectation of 0.242. In learning effectiveness,
emotions, and social expectations. SmartPLS was used to analyze values, learning motivation and social expectancy also showed
the data from the questionnaires. The first step was to eliminate significant correlations. In the learning effectiveness indicators,
data for those questionnaire items with factor loadings below the factor loading of error rate, grammatical errors and function
0.7. The expectancy cognition will be 0.684 in Cronbach’s alpha warnings are higher than 0.6, while those of answering time,
and CR will be 0.524 and social expectancy will be 0.425 in logical errors and coverage ratio are lower than 0.6 and are
Cronbach’s alpha and CR will be 0.337. However, because this excluded. For indigenous students, the relevant indicators of
item of poor validity will affect the overall effect, it will be learning motivation are value cognition and social expectancy.
removed. After analysis, the Cronbach’s alpha and rho_A of all the Specifically, their self-expectancy regarding IT programming did
items were greater than 0.7. Although the composite reliability not significantly correlate with their learning motivation. The
(CR) for the item was slightly below 0.7, it was still within impacts of social expectancy and value cognition on learning
the acceptable range (Fornell and Larcker, 1981). The average motivation varied among the Han Chinese students, for whom
variance extracted (AVE) of all the items was greater than 0.5, the impact of social expectations was not as strong as that
which was in line with the convergent validity for variations. for the indigenous students. For the latter, there were stronger
The heterotrait-monotrait ratio (HTMT) was used to determine correlations between social expectations and their learning
the discriminant validity of the various facets. All the facets met motivation and performance. In the learning effectiveness
the requirement of discriminant validity < 0.9 (Henseler et al., indicators, the factor loading of error rate, grammatical errors,
2016). logical errors, and function warnings are higher than 0.6,

FIGURE 2 | Statistics on the detailed analysis of recorded errors.

Frontiers in Psychology | www.frontiersin.org 7 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

FIGURE 3 | Impact of Han Chinese students’ learning motivation on their learning performance.

FIGURE 4 | Impact of students’ learning motivation on their learning performance.

while those of answering time and coverage ratio are lower results, it can be found that the index value cannot reflect the
than 0.6 and are excluded. According to the experimental learning effect obviously because some students who answered
results of Further discussion on why the factor loading is too the question too short or gave up waiting for the answer for
low in the answering time. According to the experimental a long time, it can be found that the index value cannot

Frontiers in Psychology | www.frontiersin.org 8 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

reflect the learning effectiveness obviously because some students in the service industry upon graduation, and job
who answered the question too short or gave up waiting for opportunities in the IT industry are lacking in their tribes.
the answer for a long time. The coverage ratio speculation may Therefore, through the popularization of information
not be correctly reflected in the learning effectiveness due to the from the indigenous tribes, it will be more able to
occurrence of plagiarism in each stage. help the indigenous students to understand the value of
information education.
iii. The impact of learning motivation on their learning
RESEARCH LIMITATIONS AND performance was different for both groups of students.
DISCUSSION For the Han Chinese students, their learning motivation
affected their overall performance in IT learning; for the
In this study, a deconstruction programming module was indigenous students, even though they might have learning
developed to analyze and examine the learning performance of motivation, the presence of economic or other factors
multi-ethnic students when they attend programming lessons. ultimately affected their learning performance. In contrast,
In this study limitations, the object of this study does not fully there were cases where their learning motivation was not
explain the value perception of all aboriginal students in learning. obvious but their learning performance ended up being
The research process was explained through the recruitment excellent. For future studies, the experimental process
instructions. It cannot be denied that for some participants, can be implemented using a general and foundational
the process of learning programming might possibly have been learning course for programming instead and add relevant
affected by the research (for example, delay in submission of culturally responsive teaching methods to help enhance
task, checking with other participants and asking them questions, the motivation of indigenous people to learn programming
and the success rate). Although four weeks of training on basic education.
computing abilities were conducted for the participants at the
early stage of the study, the impact of other factors such as
the cognitive load of IoT module circuit on their learning
effectiveness cannot be completely ruled out.
DATA AVAILABILITY STATEMENT
i. In terms of learning motivation, both groups of students The raw data supporting the conclusions of this article will be
showed correlations between their value cognition and made available by the authors, without undue reservation.
social expectations, although social expectations had a
stronger impact on the indigenous students’ learning
motivation than that of the Han Chinese students. The AUTHOR CONTRIBUTIONS
indigenous students were more concerned about whether
the learning of IT programming would affect their families C-WT was in charge of the course program teaching and research
or tribes, and this concern was also reflected in terms of process implementation. Y-WM designed and developed the AI-
the impact on their learning outcomes. Interviews with assisted platform. Y-CC was responsible for data analysis and
some of the indigenous participants revealed that they discussion. Y-HL proposed the overall research model and data-
were aware of the positive impacts of IT on their family’s oriented research design. All authors contributed to the article
finances, and that their tribes will have a better future and approved the submitted version.
with development in informatization. However, it is worth
noting that when their tribe or family had imposed a
negative expectancy on them, it would cause them to give FUNDING
up easily or even disrupt their learning in IT.
ii. There was a correlation between the Han Chinese This study was supported in part by the Ministry of Science and
students’ value and expectancy cognition and their learning Technology under grant numbers MOST 110-2511-H-143-002
motivation. This was not the case for indigenous students. and MOST 109-2511-H-143-002.
However, the learning motivation of both groups of
students correlated with their value cognition. During
the interviews, some indigenous students concurred that SUPPLEMENTARY MATERIAL
IT programming could lead to job opportunities and
increase in self-worth. However, they lacked expectancy The Supplementary Material for this article can be found
of their future employment. The main reasons were online at: https://www.frontiersin.org/articles/10.3389/fpsyg.
that indigenous students are predominantly employed 2022.868698/full#supplementary-material

REFERENCES Bandura, A. (1989). Regulation of cognitive processes through perceived self-

efficacy. Dev. Psychol. 25:729. doi: 10.1037/0012-1649.25.5.729
Aho, A. V. (2012). Computation and computational thinking. Comput. J. 55, Banks, J. A., and Banks, C. A. M. (2019). Multicultural Education: Issues and
832–835. doi: 10.1093/comjnl/bxs074 Perspectives. John Wiley & Sons.

Frontiers in Psychology | www.frontiersin.org 9 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

Barr, D., Harrison, J., and Conery, L. (2011). Computational thinking: a digital age Janica, E. V. (2021). “Math & crafts, educational activities: 400 indigenous kids
skill for everyone. Learn. Lead. Technol. 38, 20–23. learning math from engineers and scientists,” in Proceeding of the 2021 IEEE
Barr, V., and Stephenson, C. (2011). Bringing computational thinking to K-12: Integrated STEM Education Conference (ISEC), (IEEE), 45–50. doi: 10.1109/
what is involved and what is the role of the computer science education ISEC52395.2021.9764076
community? ACM Inroads 2, 48–54. Juškevičienė, A., Stupurienė, G., and Jevsikova, T. (2021). Computational thinking
Battiste, M. (2018). "Reconciling indigenous knowledge in education: promises, development through physical computing activities in STEAM education.
possibilities, and imperatives," in Dissident Knowledge in Higher Education, eds Comput. Appl. Eng. Educ. 29, 175–190. doi: 10.1002/cae.22365
M. Spooner and J. McNinch (Regina, SK: University of Regina Press), 123–148. Kanemune, S., Shirai, S., and Tani, S. (2017). Informatics and programming
Berry, J. W., and Kalin, R. (1995). Multicultural and ethnic attitudes in Canada: an education at primary and secondary schools in Japan. Olymp. Inform. 11,
overview of the 1991 national survey. Can. J. Behav. Sci. Rev. Can. Des Sci. Du 143–150. doi: 10.15388/ioi.2017.11
Comport. 27:301. doi: 10.1037/0008-400X.27.3.301 Kosslyn, S. M., Liu, D. R., Menand, L., Petersen, R. A., Pilbeam, D. R., Simmons, A.,
Blair, M. (2002). Effective school leadership: the multi-ethnic context. Br. J. Soc. et al. (2007). Report of the task force on general education. Retrieved November
Educ. 23, 179–191. doi: 10.1080/01425690220137701 12:2013.
Brandt, D. S. (2001). Information technology literacy: task knowledge and mental Ladson-Billings, G. (2004). New directions in multicultural education. Handbook
models. Libr. Trends 50, 73–86. Res. Multicult. Educ. 2, 50–65.
Chin, W. W., and Newsted, P. R. (1999). Structural equation modeling analysis Lai, Y. H., Chen, S. Y., Lai, C. F., Chang, Y. C., and Su, Y. S. (2021). Study
with small samples using partial least squares. Statist. Strat. Small Sample Res. 1, on enhancing AIoT computational thinking skills by plot image-based VR.
307–341. Interact. Learn. Environ. 29, 482–495. doi: 10.1080/10494820.2019.1580750
Denessen, E., Bakker, J., and Gierveld, M. (2007). Multi-ethnic schools’ parental Lee, J., Finn, J., and Liu, X. (2019). Time-indexed effect size for educational research
involvement policies and practices. Sch. Community J. 17, 27–44. and evaluation: reinterpreting program effects and achievement gaps in K–
Duff, P. A., and Li, D. (2009). Indigenous, minority, and heritage language 12 reading and math. J. Exp. Educ. 87, 193–213. doi: 10.1080/00220973.2017.
education in Canada: policies, contexts, and issues. Can. Mod. Lang. Rev. 66, 1409183
1–8. doi: 10.3138/cmlr.66.1.001 Lent, R. W., Brown, S. D., and Hackett, G. (2002). Social cognitive career theory.
Eades, O., Toombs, M. R., Cinelli, R., Easton, C., Hampton, R., Nicholson, Career Choice Dev. 4, 255–311.
G. C., et al. (2022). The path to eldership: results from a contemporary Li, J., Brar, A., and Roihan, N. (2021). The use of digital technology to enhance
indigenous Australian community. Gerontologist 62, 607–615. doi: 10.1093/ language and literacy skills for Indigenous people: a systematic literature review.
geront/gnab062 Comput. Educ. Open 2:100035. doi: 10.1016/j.caeo.2021.100035
Essel, H. B., Nunoo, F. K. N., and Ahiaklo-Kuz, N. A. Y. (2017). Development of an Li, Y. (2018). Current problems with the prerequisites for flipped classroom
integrated art and visual programming framework for ghanaian basic schools teaching—a case study in a university in Northwest China. Smart Learn.
based on a 21st century skill deficiency diagnostic on two basic school subjects. Environ. 5, 1–23. doi: 10.1186/s40561-018-0051-4
J. Educ. Hum. Dev. 6, 89–98. doi: 10.15640/jehd.v6n4a10 Lin, Y. T. (2019). Impacts of a flipped classroom with a smart learning diagnosis
Ezziane, Z. (2007). Information technology literacy: implications on teaching and system on students’ learning performance, perception, and problem solving
learning. J. Educ. Technol. Soc. 10, 175–191. ability in a software engineering course. Comput. Hum. Behav. 95, 187–196.
Fornell, C., and Larcker, D. F. (1981). Structural equation models with doi: 10.1016/j.chb.2018.11.036
unobservable variables and measurement error: algebra and statistics. J. Mark. Lopez, J. H., and Bobroff, K. L. (2019). Rooted indigenous core values: culturally
Res. 18, 382–388. appropriate curriculum and methods for civic education reflective of native
Gates, B. (2008). Strengthening American competitiveness for the 21st century. american culture and learning styles. Multicult. Perspect. 21, 119–126. doi:
Yearbook Natl. Soc. Study Educ. 107, 95–98. doi: 10.1111/j.1744-7984.2008. 10.1080/15210960.2019.1606640
00173_1.x Loyer, J. (2018). Indigenous Information Literacy: Nêhiyaw Kinship Enabling Self-
Gong, D., Yang, H. H., and Cai, J. (2020). Exploring the key influencing factors Care in Research. Sacramento, CA: Library Juice Press.
on college students’ computational thinking skills through flipped-classroom Lye, S. Y., and Koh, J. H. L. (2018). Case Studies Of Elementary Children’s
instruction. Int. J. Educ. Technol. Higher Educ. 17, 1–13. doi: 10.1186/s41239- Engagement In Computational Thinking Through Scratch Programming
020-00196-0 Computational Thinking in the STEM Disciplines. Cham: Springer, 227–251.
Gopnik, A., and Wellman, H. M. (1994). The Theory. In an Earlier Version Of doi: 10.1007/978-3-319-93566-9_12
This Chapter Was Presented At The Society For Research In Child Development MacIntyre, P. D., and Blackie, R. A. (2012). Action control, motivated strategies,
Meeting, 1991. Cambridge: Cambridge University Press. and integrative motivation as predictors of language learning affect and the
Grover, S., and Pea, R. (2013). Computational thinking in K–12: a review of the intention to continue learning French. System 40, 533–543. doi: 10.1016/j.
state of the field. Educ. Res. 42, 38–43. doi: 10.3102/0013189X12463051 system.2012.10.014
Gürer, M. D., Cetin, I., and Top, E. (2019). Factors affecting students’ attitudes Mannila, L., Dagiene, V., Demo, B., Grgurina, N., Mirolo, C., Rolandsson, L.,
toward computer programming. Inf. Educ. 18, 281–296. et al. (2014). “Computational thinking in K-9 education,” in Proceedings of the
Gushwa, M., Bernier, J., and Robinson, D. (2019). Advancing child sexual abuse Working Group Reports of the 2014 on Innovation & Technology in Computer
prevention in schools: an exploration of the effectiveness of the enough! Online Science Education Conference, 1–29. doi: 10.1145/2713609.2713610
training program for K-12 teachers. J. Child Sexual Abuse 28, 144–159. doi: Marcum, J. W. (2002). Rethinking information literacy. Library Quart. 72, 1–26.
10.1080/10538712.2018.1477000 doi: 10.1086/603335
Haberman, M., and Post, L. (1998). Teachers for multicultural schools: the power Markauskaite, L. (2006). Towards an integrated analytical framework of
of selection. Theory Pract. 37, 96–104. doi: 10.1016/j.jtbi.2017.06.032 information and communications technology literacy: from intended to
Hair, J. F., Hult, G. T. M., Ringle, C. M., Sarstedt, M., and Thiele, K. O. (2017). implemented and achieved dimensions. Inform. Res. Int. Electr. J. 11:n3.
Mirror, mirror on the wall: a comparative evaluation of composite-based Martin, A. J. (2006). A motivational psychology for the education of
structural equation modeling methods. J. Acad. Mark. Sci. 45, 616–632. doi: indigenous Australian students. Austr. J. Indig. Educ. 35, 30–43. doi: 10.1017/
10.1007/s11747-017-0517-x S1326011100004142
Henseler, J., Hubona, G., and Ray, P. A. (2016). Using PLS path modeling in new Modood, T., and May, S. (2001). Multiculturalism and education in britain: an
technology research: updated guidelines. Ind. Manage. Data Syst. 116, 2–20. internally contested debate. Int. J. Educ. Res. 35, 305–317. doi: 0.1016/S0883-
doi: 10.1108/IMDS-09-2015-0382 0355(01)00026-X
Hsu, T. C., Chang, S. C., and Hung, Y. T. (2018). How to learn and how to Nikolaeva, A. D., and Savvinov, V. M. (2016). Multi-ethnic school in the russian
teach computational thinking: Suggestions based on a review of the literature. federation: the preconditions of formation and development (a case study of a
Comput. Educ. 126, 296–310. doi: 10.1016/j.compedu.2018.07.004 national region). Int. Electr. J. Math. Educ. 11, 3405–3414.
Huitt, W., and Hummel, J. (2003). Piaget’s theory of cognitive development. Educ. Nouri, J., Zhang, L., Mannila, L., and Norén, E. (2020). Development of
Psychol. Interact. 3, 1–5. computational thinking, digital competence and 21st century skills when

Frontiers in Psychology | www.frontiersin.org 10 June 2022 | Volume 13 | Article 868698

Tsai et al. AI-Assisted Multiculturalism Programming Lessons

learning programming in K-9. Educ. Inquiry 11, 1–17. doi: 10.1080/20004508. Sulisworo, D., and Suryani, F. (2014). The effect of cooperative learning, motivation
2019.1627844 and information technology literacy to achievement. Int. J. Learn. Dev. 4, 58–64.
Perso, T., and Hayward, C. (2020). Teaching Indigenous Students: Cultural doi: 10.5296/ijld.v4i2.4908
Awareness and Classroom Strategies for Improving Learning Outcomes. London: Sylva, T., Chinn, P., and Kinoshita, C. (2010). A culturally relevant agricultural and
Routledge. doi: 10.4324/9781003117728 environmental course for K–12 teachers in hawaii. J. Natl. Res. Life Sci. Educ.
Pintrich, P. R., Smith, D. A., Garcia, T., and McKeachie, W. J. (1993). 39, 10–14. doi: 10.4195/jnrlse.2008.0040k
Reliability and predictive validity of the motivated strategies for learning Tzou, C., Suárez, E., Bell, P., LaBonte, D., Starks, E., and Bang, M. (2019).
questionnaire (MSLQ). Educ. Psychol. Measure. 53, 801–813. doi: 10.1177/ Storywork in STEM-art: making, materiality and robotics within everyday acts
0013164493053003024 of indigenous presence and resurgence. Cognition and Instruction 37, 306–326.
Pramuda, A., Kuswanto, H., and Hadiati, S. (2019). Effect of real-time physics doi: 10.1080/07370008.2019.1624547
organizer based smartphone and indigenous technology to students’ scientific U.S. National Science and Technology Council (2016). National Artificial
literacy viewed from gender differences. Int. J. Instruct. 12, 253–270. doi: 0. Intelligence Research and Development Strategic Plan. Washington DC:
29333/iji.2019.12316a Networking and Information Technology Research and Development
Race, R. (2015). Multiculturalism and Education. London: Bloomsbury Publishing. Subcommittee.
Sangwin, C. J., and O’Toole, C. (2017). Computer programming Writer, J. H. (2008). Unmasking, exposing, and confronting: critical race theory,
in the UK undergraduate mathematics curriculum. Int. J. Math. tribal critical race theory and multicultural education. Int. J. Multicult. Educ.
Educ. Sci. Technol. 48, 1133–1152. doi: 10.1080/0020739X.2017.13 10, 1–15. doi: 10.18251/ijme.v10i2.137
15186
Saputro, A. D., Atun, S., and Wilujeng, I. (2019). The impact of problem Conflict of Interest: The authors declare that the research was conducted in the
solving instruction on academic achievement and science process skills among absence of any commercial or financial relationships that could be construed as a
prospective elementary teachers. Ilkogretim Online 18:2. potential conflict of interest.
Scherer, R., Siddiq, F., and Viveros, B. S. (2020). A meta-analysis of teaching
and learning computer programming: effective instructional approaches Publisher’s Note: All claims expressed in this article are solely those of the authors
and conditions. Comput. Hum. Behav. 109:106349. doi: 10.1016/j.chb.2020. and do not necessarily represent those of their affiliated organizations, or those of
106349 the publisher, the editors and the reviewers. Any product that may be evaluated in
Schwab, R. (2018). Why Only One In Three? The Complex Reasons For Low this article, or claim that may be made by its manufacturer, is not guaranteed or
Indigenous School Retention. Canberra, ACT: Centre for Aboriginal Economic endorsed by the publisher.
Policy Research (CAEPR), The Australian National University.
Serin, O. (2011). The effects of the computer-based instruction on the achievement Copyright © 2022 Tsai, Ma, Chang and Lai. This is an open-access article distributed
and problem solving skills of the science and technology students. Turkish under the terms of the Creative Commons Attribution License (CC BY). The use,
Online J. Educ. Technol. TOJET 10, 183–201. distribution or reproduction in other forums is permitted, provided the original
Shute, V. J., Sun, C., and Asbell-Clarke, J. (2017). Demystifying computational author(s) and the copyright owner(s) are credited and that the original publication
thinking. Educ. Res. Rev. 22, 142–158. doi: 10.1016/j.edurev.2017. in this journal is cited, in accordance with accepted academic practice. No use,
09.003 distribution or reproduction is permitted which does not comply with these terms.

Frontiers in Psychology | www.frontiersin.org 11 June 2022 | Volume 13 | Article 868698