The Influence of Cognitive Presence, Social Presence and Teaching

Presence on Online Foreign Language Speaking Anxiety, L2

Motivational Self and Intended Effort-A Structural Equation

Modeling Approach

Qihui Chen

, Wenting Sun

and Xiaoling Wang

Institute of Modern Distance Education, Tongji University, Siping Road 1239, Shanghai, China

Department of Computer Science, Humboldt-Universität zu Berlin, Berlin, Germany

Department of College English, Zhejiang Normal University, Jinhua, China

Keywords: Community of Inquiry, Cognitive Presence, Teaching Presence, Social Presence, L2 (Second Language)

Motivational Self System, Ideal L2 Self, Ought-to L2 Self, General Self-Efficacy, Intended Effort.

Abstract: Although a large number of online foreign language oral learners has emerged, little is known about the

interactive mechanism between online learning environmental factors and learners’ self-relevant factors.

Using the quantitative method, 466 questionnaires of Chinese young adult learners were collected to test the

hypothesized relationships between cognitive presence (CP), teaching presence (TP), social presence (SP),

L2(second language) motivational self (including Ideal L2 self (IL2) and Ought-to L2 self (OL2)), general

self-efficacy (GSE), online foreign language speaking anxiety (AN) and intended efforts (IE) in a structural

equation model. The findings illustrate: (1) Online foreign language speaking learners are in middle anxiety,

high GSE and high IE level; (2) Online learning environment is an overall ecology in which CP, TP, SP, IL2,

OL2, and GSE are highly correlated with each other; (3) In terms of causality, TP and OL2 enhance AN, GSE

weakens AN, GSE and OL2 strengthen IE while AN reduces IE. To facilitate online oral foreign language

learning, designers of online learning platforms should consider providing the choice of learning partners and

more meta-cognition support to guide learners to deal with negative evaluations and manage learning

recordings.

1 INTRODUCTION

Along with online international conferences have

increased sharply, opportunities for online

communication in foreign languages soar. In this

case, the demand for online oral foreign language

learning has continued to rise. It is reported that the

annual growth rate of China’s adult English training

industry is as high as 25% (Lu et al., 2015). A large

number of online foreign language learning platforms

emerged, such as Duolingo, LingoDeer, Liulishuo,

Shanbei vocabulary English. However, the negative

experience impacted by anxiety reduces learners’

interest and enthusiasm for learning (Dirkx, 2008).

Compared with other subjects, foreign language

learners suffer more anxiety. The mismatch between

https://orcid.org/0000-0003-1445-2342

https://orcid.org/0000-0003-3413-2450

the real self and the expressive self induces the

anxiety of communicators (Horwitz, 1986). Learners

who have high anxiety levels tend to escape from the

foreign language communication scenario; on the

other hand, less engagement in speaking activities

reduces feedback about the progress of language

performance, which may trigger learners’ speaking

anxiety in return. This means that self-relevant

cognition and feedback from others influence the

anxiety of foreign language learners.

Regarding self-relevant cognition in foreign

language learning, Dörnyei's (2009) second language

(L2) motivational self system integrates L2 learning

motivation and identity, including Ideal L2 self (IL2)

and Ought-to L2 self (IL2). Based on this system,

Papi (2010) found that IL2 weakens English anxiety,

Chen, Q., Sun, W. and Wang, X.

The Inﬂuence of Cognitive Presence, Social Presence and Teaching Presence on Online Foreign Language Speaking Anxiety, L2 Motivational Self and Intended Effort-A Structural Equation

Modeling Approach.

DOI: 10.5220/0010979000003182

In Proceedings of the 14th International Conference on Computer Supported Education (CSEDU 2022) - Volume 1, pages 173-180

ISBN: 978-989-758-562-3; ISSN: 2184-5026

173

and OL2 enhances English anxiety. Besides, self-

efficacy is another important self- relevant factor in

the L2 motivational self system (Ueki & Takeuchi,

2012).

Regarding the analysis of feedback from others in

online foreign language speaking learning,

Community of Inquiry (CoI) theory can be considered

because this theory promotes researchers to explore

the dynamic process of online learning experience

through a collaborative constructivist perspective

(Arbaugh et al., 2008), which includes three closely

related online presences and one overlap section-

cognitive presence (CP), teaching presence (TP),

social presence (SP) and meaningful learning

experience (Garrison et al., 1999). Using the CoI

model to specify the learning experience involved in

online foreign language oral learning environment

can further reveal the interaction mechanism between

online learning environment and self-relevant factors,

and help improve the construction of online learning

environments which require highly human-computer

interaction. In this regard, three research questions

were proposed:

1. How do cognitive presence (CP), teaching

presence (TP) and social presence (SP) predict

online foreign language speaking anxiety (AN)

and intended efforts (IE)?

2. Whether CP, TP and SP influence L2 motivation

(including IL2, OL2) and self-efficacy

(especially general self-efficacy (GSE))?

3. How do L2 motivation and self-efficacy predict

online foreign language speaking anxiety (AN)

and intended efforts (IE)?

The next section proposes hypotheses based on a

theoretical foundation; methodology is shown in

section 3 and results in section 4. Section 5 gives

more details in discussion, implications and

limitations while section 6 gives conclusions and

future directions.

2 VARIABLES AND HYPOTHESES

The following section provides a quick introduction

about the used variables and describes hypotheses

among variables.

2.1 Variables Used

The most widely used definition of foreign language

anxiety is related to classroom language learning and

originated from the unique complex self-awareness,

concepts, emotions, and behaviors in the language

learning process, which includes communication

apprehension (CAN), fear of negative evaluation

(FAN) and test anxiety (TAN) (Horwitz et al., 1986).

Given the online language learning environment, this

study refers to this definition and subcategories but in

the online oral learning environment.

The L2 motivational self system proposed by

Dörnyei (2014) contains three main components: IL2,

OL2, and L2 learning experience, which has been

applied to different languages and cultural

environments (Ryan, 2009; Kong et al., 2018).

According to Dörnyei (2014), IL2 represents the ideal

image of L2 users that individuals hope to become in

the L2 field while OL2 self is affected by duties,

obligations or responsibilities, an attribute that others

believe that an individual should possess. The L2

learning experience is a specific contextual

motivation related to the immediate learning

environment and experience, for which CP, TP and

SP represent this factor in this study. Moreover,

adding other affective variables would improve the

prediction of the L2 motivational self framework,

such as L2 anxiety and self-efficacy (Kormos et al.,

2011; Dörnyei, 2014). By adding L2 anxiety and

intended effort, Papi (2010) extended this system,

which is one of the basic models of this research.

Referring to the definition of IE in the study of

Papi (2010), this variable was considered in this

study, which is defined as the time and energy that

learners plan to spend on the online foreign language

oral learning platform in the future. It is used as a

variable to measure learners' willingness to learn oral

foreign languages online.

GSE is defined as a personal belief, which refers

to an individual's comprehensive view of their ability

to perform in various situations (Judge et al., 1998).

In this regard, GSE focuses on the individual's

perception of the ability to meet task requirements in

different situations. Because online foreign language

oral learning involves multiple senses of self-

efficacy, such as computer self-efficacy, foreign

language oral self-efficacy, this study uses a wider

range of GSE to represent personal positive

inclinations and belief in the ability to complete tasks.

The experience process of CP is similar to the

learning process of foreign languages. Tobias (2013)

divided the language information processing process

into three stages: input, processing, and output.

Cleveland and Campbell (2012) pointed out that CP

describes the extent to which learners can construct

and confirm meaning through continuous reflection

and discourse. Ideally, CP includes triggering events,

exploring, integration and resolution. In this regard,

CP of this research refers to the degree of learners’

CSEDU 2022 - 14th International Conference on Computer Supported Education

174

knowledge understanding and construction in online

foreign language oral learning, including these four

subcategories.

Cleveland and Campbell (2012) emphasized that

teaching labor is the central organizational element of

online presence, which has the function of designing

and promoting online learning. Garrison et al. (1999)

determined three subcategories of TP from a large

amount of literature and inquiries: instructional

management, building understanding, and direct

instruction. The three subcategories of teaching

presence are highly similar to the comprehensible

input hypothesis of Krashen (1982) in the traditional

foreign language learning environment. Regarding

the definition and subcategories, TP of this study

refers to the guidance of the construction of foreign

language knowledge and social relations that learners

perceived during the online oral foreign language

learning process, including these three subcategories.

Cleveland and Campbell (2012) described SP

refers to the degree of social and emotional

connection between learners and others in an online

learning environment. Through a large number of

documents, Garrison et al., (1999) summarized that

SP includes three subcategories: emotional

expression, open communication and group cohesion.

These three factors are highly similar to the opposites

of willingness to communicate, group cohesion, and

fear of negative evaluation in the study of traditional

foreign language anxiety influencing factors.

Therefore, this study introduces SP into the online

foreign language oral learning anxiety influencing

factor model. It is defined as the strength of

expressing themselves and the strength of connection

with others that learners feel in the process of online

foreign language oral learning, including these three

subcategories.

2.2 Hypotheses

The CoI model regards online communities as places

that support online learning through the interaction

among CP, TP and SP (Thompson & MacDonald,

2005; Shea, 2006). The three factors were proved to

be interrelated when the theory was created (Garrison

et al., 1999), but additional empirical materials are

needed to test it (Arbaugh et al., 2008). Therefore, in

online foreign language oral learning, hypotheses

were established in this study:

H1: CP and SP positively influence each other;

H2: CP and SP positively influence each other;

H3: SP and TP positively influence each other;

In the process of foreign language oral learning,

learners are exposed to factors related to the possible

successful experience of language learning, including

courses, teachers, and peers. The path from foreign

language learning experience to foreign language

anxiety has been confirmed by many studies (Aida,

1994; Young, 1991). Here, hypotheses were

proposed:

H4: CP reduces AN; H5: SP reduces AN;

H6: TP reduces AN; H7: CP increases IE;

H8: TP increases IE; H9: SP increases IE;

Tobias (2013) pointed out that increased effort

can compensate for the negative effects of learner

anxiety in the three language stages of input,

processing, and output. However, in general, the

speed of the L2 information exchange process is too

fast to allow this kind of compensation. Therefore, a

hypothesis was established:

H10: AN reduces IE;

Research by Ryan (2009) showed that IL2 is

significantly related to immersive motivation and

explains expected effort. Although IL2 and OL2

express learners’ learning motivation from different

sources internally and externally, both are

motivations in essence and can promote the effort.

Therefore, hypotheses were established:

H11: IL2 and OL2 influence each other;

H12: IL2 increases IE; H13: OL2 increase IE;

Research by Papi (2010), Ueki and Takeuchi

(2012) showed similar results that IL2 has a

weakening effect on L2 anxiety, and OL2

significantly increases L2 anxiety. Here, similar

hypotheses were established:

H14: IL2 reduce AN; H15: OL2 increase AN;

MacIntyre et al. (1997) pointed out that students

with low self-confidence in the L2 learning

environment are more inclined to have negative

expectations of their behavior and become more

anxious when facing language learning tasks. A

similar hypothesis was established:

H16: GSE reduces AN;

Bong and Skaalvik (2003) demonstrated that self-

efficacy has an important motivational force. The

empirical study by Kormos et al. (2011) verified that

self-efficacy in L2 learning has a strong influence on

L2 motivation. Nielsen’s (2009) research has shown

The Inﬂuence of Cognitive Presence, Social Presence and Teaching Presence on Online Foreign Language Speaking Anxiety, L2

Motivational Self and Intended Effort-A Structural Equation Modeling Approach

175

that the stronger the self-efficacy of L2 learners, the

greater the expected effort. Hence, hypothesized:

H17: GSE is positively correlated with IL2;

H18: GSE is positively correlated with OL2;

H19: GSE increases IE;

Many studies have pointed out that the

environment has an important influence on the

formation and maintenance of students' L2 learning

motivation (Csizer & Dörnyei, 2005; Ryan, 2009).

Kormos et al. (2011) illustrated that the surrounding

environment of students: family, friends, peers,

teachers, and guidance materials, all contribute to the

learner’s goal setting, attitude information, self-

efficacy, and persistence in learning activities.

Therefore, hypotheses were established:

H20a: CP is positively correlated with IL2;

H20b: TP is positively correlated with IL2;

H20c: SP is positively correlated with IL2;

H21a: CP is positively correlated with OL2;

H21b: TP is positively correlated with OL2;

H21c: SP is positively correlated with OL2;

Bandura (1977), Markus and Nurius (1986) found

that mastering history, social comparison, attribution,

and evaluation of important others are all important

to learners’ sense of self-efficacy. In this study, CP,

TP and SP represent the knowledge exploration,

knowledge-guided help, and emotion exchange with

other participants experienced by learners on an

online foreign language oral learning platform.

Therefore, established hypotheses:

H22a: CP is positively correlated with GSE;

H22b: TP is positively correlated with GSE;

H22c: SP is positively correlated with GSE.

3 METHODOLOGIES

In this section, pedagogical design features are

represented through Liulishuo (a popular foreign

language oral learning APP) as an example. Then, the

research procedure, participants and research

instrument are displayed.

3.1 Pedagogical Design Features

For the description of the pedagogical design of

foreign language oral learning platform, Liulishuo in

China was chosen because this app has 120 million

downloads, 4.9 points (5-point rating system), and

employs intelligent speech recognition technology.

Based on the authors’ learning experience in

foreign language oral learning platforms, here are

some common learning activities shared by platforms

and represent in Liulishuo App. When the first

loading on this app, learners are recommended to

locate their English preference within several levels.

Then, according to users’ choice on preferred content

(e.g., grammar, vocabulary), interests (e.g., English

interview, daily conversation) and study schedule, a

learning plan is created. And after daily tasks, a

promotion can be clicked to share learning results on

social media. Besides basic sentences and vocabulary,

interesting topics can be added to a personalized

learning plan. Then, following and combined with the

voice recording of native English speakers, users

practice the speaking skills of relevant vocabulary

and sentences and obtain instrumental feedback on

inaccurate pronunciation. After one task, learners

gain a personalized radar chart of accuracy, rhythm,

fluency and pronunciation and weakness analysis in

pronunciation problems and sentence problems

combining recommended learning materials.

3.2 Procedure

First, an online foreign language speaking anxiety

questionnaire was organized based on the existing

ones. Then the survey was hosted on Wenjuanxing

(similar to SurveyMonkey) and was distributed in the

foreign language learning groups and universities

groups in the WeChat platform and QQ platform

(similar to WhatsApp and Facebook). After data was

collected, the reliability and validity of the

questionnaire were evaluated and confirmatory factor

analysis (CFA) of the hypotheses was verified.

3.3 Participants

The research mainly focused on Chinese young adult

learners. There were 466 valid questionnaires. The

average age is around 25 and the proportion of 18-30

years old is 89%. Males account for 38%, females

account for 62%. Among them, 74% are

undergraduate and postgraduate.

3.4 Research Instruments

Referring to existing questionnaires, mainly

Community of Inquiry instrument (Arbaugh et al.,

2008; Cleveland & Campbell, 2012), foreign

language anxiety scale (Horwitz et al., 1986), general

self-efficacy scale (Chen et al., 2001), L2

motivational self scale (Papi, 2010), online foreign

language speaking anxiety scale was developed. All

CSEDU 2022 - 14th International Conference on Computer Supported Education

176

questionnaire items were designed using a five-point

Likert scale, ranging from 5=strongly agree, to

1=strongly disagree. Educational technology experts

and English professionals were organized to translate

English version’s items into Chinese ones, and then

back to English scale to check the ambiguity of the

Chinese questionnaires. The Chinese questionnaires

can be found at the link 在线英语口语焦虑及影响

因素问卷 (wjx.cn).

4 RESULTS

In this article, reliability and validity were evaluated

in SPSS 22. The Cronbach's Alpha coefficient of the

overall questionnaire is 0.971, meaning that the

reliability of the questionnaire is relatively high. The

KMO (Kaiser-Meyer-Olkin) value of the overall

questionnaire is 0.967, and the approximate variance

of Bartlett's spherical test is 16397.127 (p<0.001),

indicating that the sample data is suitable for factor

analysis.

A confirmatory factor analysis (CFA) was

performed in AMOS 26 to measure the fitness

between the hypothesized model and observed data.

Root mean square error of approximation (RMSEA)=

0.056, Goodness-of-fit index (GFI)= 0.904, Adjusted

goodness-of-fit index (AGFI)= 0.874, Normed fit

index (NFI)= 0.943, TLI=0.958, Comparative fit

index (CFI)= 0.965, Parsimony normed fit index

(PNFI)= 0.776, Those indexes is consistent with the

goodness-of-fit indexes recommended by Hu and

Bentler (1999) and Wu (2009), meaning a relatively

good fit between the hypothesized model and the

observed data. In other words, the final version of the

online foreign language speaking anxiety model is a

reasonable representation of the data collected. More

detailed information about standardized factor

loading of variables can be found in table 1 and the

standard coefficient of paths in table 2.

4.1 Descriptive Statistics and

Measurement Model

The online foreign language speaking anxiety model

was tested, and measurement modeling results were

summarized in table 1. For short, results of

descriptive statistics were also combined in table 1.

The descriptive statistics results (details in Table

1) present that learners are in a state of moderate AN

(M=3.48). Comparatively, learners are at a relatively

high level of IE (M=3.91) and GSE (M=3.94). By

contrast, IL2 (M=3.68) and OL2 (M=3.64) are at the

mediate level. This displays that in the online

environment, the motivation to learn English is not

high though learners believe they can succeed in most

tasks. CP(M=3.72), TP (M=3.72) and SP (M=3.67)

also have similar levels, meaning that foreign

language learners may not receive large information

to promote content exploration, learning activities

organization and group cohesion online.

Table 1: Summary of the measurement model.

Latent variables Items SFL

Cognitive

Presence

(CP)(M=3.72)

CP1 0.842***

CP2 0.868***

CP3 0.883***

CP4 0.879***

Teaching

Presence

(TP)(M=3.72)

TP1 0.902***

TP2 0.918***

TP3 0.830***

Social Presence

(SP)(M=3.67)

SP1 0.907***

SP2 0.904***

SP3 0.887***

Ideal L2 self

(IL2) (M=3.68)

IL21 0.840***

IL22 0.829***

IL23 0.875***

Ought-to L2 self

(OL2) (M=3.64)

OL21 0.757***

OL22 0.829***

OL23 0.851***

General self-

efficacy

(GSE)(M=3.94)

GSE1 0.824***

GSE2 0.872***

GSE3 0.819***

Online foreign

language

speaking anxiety

(AN)(M=3.48)

CAN 0.858***

NAN 0.916***

TAN 0.781***

Intended Effort

(IE)(M=3.91)

IE1 0.863***

IE2 0.822***

IE3 0.825***

Note: M=Mean, SFL=Standardized factor loading,

***P<0.001, CP1=Triggering Event, CP2=Exploration,

CP3=Integration, CP4=Resolution, TP1=Instructional

Management, TP2= Building Understanding, TP3= Direct

Instruction, SP1=Emotional Expression, SP2=Open

Communication, SP3=Group Cohesion, CAN=

Communication Apprehension (CAN), NAN=Fear of

Negative Evaluation, TAN=Test Anxiety.

4.2 Structural Equation Model

The conducted CFA’s results were summarized in

table 2. The results present that all the mutual

influence relations have been proved (see table 2, H1-

H3, H11, H17, H18, H20, H21 and H22). Only a

small part of the causal relations has been proved (see

table 2, H10, H13, H15, H16 and H19). More detailed

analysis can be found in section 5.

The Inﬂuence of Cognitive Presence, Social Presence and Teaching Presence on Online Foreign Language Speaking Anxiety, L2

Motivational Self and Intended Effort-A Structural Equation Modeling Approach

177

Table 2: Summary of structural equation modeling.

Path SC H

otheses

CP->AN

(

)

-0.379 Not

TP ->AN

(

)

0.88*** Not

SP ->AN (H5) -0.145 Not

IL2->AN (H14) -0.014 Not

OL2->AN(H15) 0.527*** Supporte

GSE->AN

(

H16

)

-0.145* Su

orte

AN->IE

(

H10

)

-0.129* Su

orte

GSE->IE

(

H19

)

0.284*** Su

orte

CP->IE (H7) 0.279 Not

IL2->IE (H12) 0.116 Not

OL2->IE

(

H13

)

0.226** Su

orte

TP->IE

(

)

0.344 Not

SP->IE

(

)

-0.192 Not

TP<->CP (H2) 0.954*** Supporte

SP<->TP (H3) 0.928*** Supporte

SP<->CP(H1) 0.936*** Supporte

OL2<->IL2

(

H11

)

0.647*** Su

orte

OL2<->GSE

(

H18

)

0.662*** Su

orte

IL2<->GSE

(

H17

)

0.678*** Su

orte

IL2<->CP (H20a) 0.757*** Supporte

IL2<->TP (H20b) 0.727*** Supporte

IL2<->SP

(

H20c

)

0.774*** Su

orte

OL2<->CP

(

H21a

)

0.657*** Su

orte

OL2<->TP

(

H21b

)

0.612*** Su

orte

OL2<->SP (H21c) 0.692*** Supporte

GSE<->SP (H22c) 0.673*** Supporte

GSE<->TP (H22b) 0.650*** Supporte

GSE<->CP

(

H22a

)

0.672*** Su

orte

Note: ***P<0.001, **P<0.01, *P<0.05, SC=Standard

coefficient, “-” means negative relation, “->” means lead to,

“<->” means mutual influence.

5 DISCUSSION AND IMPLICATION

Among the influencing factors of IE, the path

coefficient of AN is -0.129 (see table 2). The research

results show that the anxiety experienced by learners

in the process of online foreign language speaking

learning belongs to obstructive anxiety. This result is

different from Shih (2019), in which AN has no

significant impact on IE. This may be related to the

sample. The sample of Shih (2019) are students from

senior high school in Taiwan and English is a

compulsory subject and an important subject of the

entrance college examination.

The strong correlation between CP, TP and SP

validates the close relationship described by Garrison

et al. (1999) in online foreign language oral learning.

IL2, OL2 and GSE are highly positively

correlated, which correlation coefficient is not less

than 0.65(see table 2). This echoes personality

psychology about the close connection between IL2

and OL2(Leary, 2007).

5.1 TP Enhances AN

The path coefficient of TP is 0.88 (see table 2), which

rejects H6. The research results show that the sense

of TP has the greatest impact on AN. It means that

although the sense of TP may enhance learners'

attention on online learning tasks, learners' sense of

tension and even anxiety are increased. This may be

related to the program setting of the online foreign

language learning platform. The teaching

management and direct guidance of the existing

foreign language oral learning platform are mainly

based on the procedure automatic response. Without

considering personal learning characteristics, the

repeated imitation process may increase the anxiety

of learners. At the same time, this result also verifies

the research results of Grant et al., (2013) about

negative effects generated in the traditional foreign

language learning environment may affect learners'

online learning experiences.

CP and SP have no significant impact on AN and

IE. The real-time interaction between learners is still

low due to the technical reasons of the online foreign

language learning platform and different individuals’

learning schedules. According to the authors’

experience in Liulishuo and Duolingo (two foreign

language speaking learning Apps), there is no

recommendation of learning partners to help

individuals practice and reflect on what has been

learned online.

5.2 IL2, OL2, GSE, CP, TP, SP

Promote Each Other

The path coefficients data show that IL2, OL2, GSE

are highly positively correlated with CP, TP, SP

respectively, for which the correlation coefficient is

not less than 0.61 (see table 2). The results of this

study support the argument of Cohen and Norst

(1989) about the close connection between language

and self.

5.3 OL2 Increase AN, GSE Reduce

AN, Both of Them Increase IE

The path coefficients from OL2 and GSE to AN are

0.527 and -0.145 respectively (see table 2), which

implies that OL2 and GSE have a significant impact

on AN. This indicates that external environmental

pressure strongly affects learners’ AN.

CSEDU 2022 - 14th International Conference on Computer Supported Education

178

OL2 and GSE have a significant impact on IE, and

the path coefficients are 0.226 and 0.284 respectively

(see table 2). This result supports the view that

instrumental motivation can increase learning

behavior (Gardner & MacIntyre, 1991) and Bandura's

(1977) view that people with low self-efficacy are

more likely to give up.

5.4 Implications and Limitations

Based on the research results, here are some

suggestions for the design of foreign language

learning platforms.

Recommend learning partners to practice and

reflect on how to use what has been learned online in

real life. Although the platform collects a large

amount of corpus and provides communication

templates, flexibility is lacking when real practices

are faced by real people in their lives.

Enrich the display forms of TP. In terms of course

setting, the use of concept maps to show the internal

connections between learning content and individuals’

preferences can enhance learners’ perception of the

structure of learning content, help learners clarify

their learning goals, and reduce additional cognitive

load (Hwang et al., 2011).

Strengthen learners' real-time recognition of their

anxiety, present different instructions according to the

degree of anxiety, guide learners to reduce their sense

of learning anxiety, and increase their attention to

learning content. If permits, wearable devices can be

used to identify the learner's heartbeat and other

physiological information, roughly estimate the

emotional intensity, and then propose different

reminders to calm the mood.

Guide learners to correctly deal with negative

evaluations. On the cognitive hand, learners need to

summarize the causes of errors and find ways to solve

them about specific tasks; on the meta-cognition hand,

learners need chances to compare their current and

previous learning results, and think about whether to

improve study plan further. This means that the online

platform should consider providing more learning

analytics data to support meta-cognition processing.

Some limitations affect the generalization of the

conclusions to a certain extent. First, there is no

comparative study between different learning

environments (e.g., online, face-to-face or blended)

and different technologies used (e.g., AR/VR, voice

chatbot). Then, this study also did not analyze the

specific functional applications of each learning

platform.

6 CONCLUSIONS

The descriptive statistics show that online foreign

language speaking learners are in middle anxiety,

high GSE and IE level. The structural equation

modeling shows that TP, OL2 increase AN while

GSE significantly reduces AN; GSE, OL2 increase IE

while AN reduces IE. The high correlation between

CP, TP, SP, IL2, OL2, and GSE proves that learners’

internal state and online learning environment is a

dynamic cyclic process. In this regard, personalized

or adaptive learning both from cognition and emotion

could improve the learning experience. Some

suggestions about human-computer interaction in

online learning platforms are recommended.

With the help of artificial intelligence and big

data, online education pays more attention to

individual learners. Smart tutors, smart teaching

assistants, etc. enrich the virtual learning

environment. Online learning at this stage and in the

future has higher requirements for human-computer

interaction and collaborative learning. But learning

still happens in people, and people have both reason

and emotional aspects. Therefore, online learning

with a strong interactive nature should pay more

attention to the emotional state of the learner, to

explore which emotions have a positive effect on the

learner of different features, and how to reduce the

emotions which hurt the learning experience.

Moreover, future studies are recommended to analyze

the design elements in an online learning environment

representing CP, TP, SP from a platform perspective.

Comparative studies of CP, TP, SP (representatives

and their relationships with learning performance and

evaluation) between different disciplines and

different difficulty levels are also good direction.

ACKNOWLEDGEMENTS

We would like to thank Pro. Raphael Zender

(Humboldt-Universität zu Berlin) for his feedback on

an earlier version of this paper and Zeyue Zhu (Tongji

University) for the collecting of the questionnaire.

REFERENCES

Aida, Y. (1994). Examination of Horwitz, Horwitz, and

Cope's construct of foreign language anxiety: The case

of students of Japanese. The modern language journal,

78(2), 155-168.

Arbaugh, J. B., Cleveland-Innes, M., Diaz, S. R., Garrison,

D. R., Ice, P., Richardson, J. C., & Swan, K. P. (2008).

The Inﬂuence of Cognitive Presence, Social Presence and Teaching Presence on Online Foreign Language Speaking Anxiety, L2

Motivational Self and Intended Effort-A Structural Equation Modeling Approach

179

Developing a community of inquiry instrument:

Testing a measure of the community of inquiry

framework using a multi-institutional sample. The

internet and higher education, 11(3-4), 133-136.

Bandura, A. (1977). Self-efficacy: toward a unifying theory

of behavioral change. Psychological Review, 84(2),

191-215.

Bong, M., & Skaalvik, E. M. (2003). Academic self-

concept and self-efficacy: How different are they

really?. Educational psychology review, 15(1), 1-40.

Chen, G., Gully, S. M., & Eden, D. (2001). Validation of a

new general self-efficacy scale. Organizational

research methods, 4(1), 62-83.

Cleveland-Innes, M., & Campbell, P. (2012). Emotional

presence, learning, and the online learning

environment. The International Review of Research in

Open and Distributed Learning, 13(4), 269-292.

Cohen, Y., & Norst, M. J. (1989). Fear, dependence and

loss of self-esteem: Affective barriers in second

language learning among adults. RELC Journal, 20(2),

61-77.

Csizér, K., & Dörnyei, Z. (2005). The internal structure of

language learning motivation and its relationship with

language choice and learning effort. The modern

language journal, 89(1), 19-36.

Dirkx, J. M. (2008). The meaning and role of emotions in

adult learning. New directions for adult and continuing

education, 2008(120), 7-18.

Dörnyei, Z. (2009). The L2 Motivational Self System. In

Motivation, language identity and the L2 self (pp. 9-42).

Multilingual Matters.

Dörnyei, Z. (2014). The psychology of the language

learner: Individual differences in second language

acquisition. Routledge.

Gardner, R. C., & MacIntyre, P. D. (1991). An instrumental

motivation in language study: Who says it isn't

effective?. Studies in second language acquisition,

13(1), 57-72.

Garrison, D. R., Anderson, T., & Archer, W. (1999).

Critical inquiry in a text-based environment: Computer

conferencing in higher education. The internet and

higher education, 2(2-3), 87-105.

Grant, S. J., Huang, H., & Pasfield-Neofitou, S. E. (2013).

Language learning in virtual worlds: The role of foreign

language and technical anxiety. Journal for virtual

worlds research, 6(1), 1-9.

Horwitz, E. K. (1986). Preliminary evidence for the

reliability and validity of a foreign language anxiety

scale. Tesol Quarterly, 20(3), 559-562.

Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). Foreign

language classroom anxiety. The Modern language

journal, 70(2), 125-132.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit

indexes in covariance structure analysis: Conventional

criteria versus new alternatives. Structural equation

modeling: a multidisciplinary journal, 6(1), 1-55.

Hwang, G. J., Wu, P. H., & Ke, H. R. (2011). An interactive

concept map approach to supporting mobile learning

activities for natural science courses. Computers &

Education, 57(4), 2272-2280.

Judge, T. A., Erez, A., & Bono, J. E. (1998). The power of

being positive: The relation between positive self-

concept and job performance. Human performance,

11(2-3), 167-187.

Kong, J. H., Han, J. E., Kim, S., Park, H., Kim, Y. S., &

Park, H. (2018). L2 motivational self system,

international posture and competitiveness of Korean

CTL and LCTL college learners: A structural equation

modeling approach. System, 72, 178-189.

Kormos, J., Kiddle, T., & Csizér, K. (2011). Systems of

goals, attitudes, and self-related beliefs in second-

language-learning motivation. Applied linguistics,

32(5), 495-516.

Leary, M. R. (2007). Motivational and emotional aspects of

the self. Annual Review of Psychology, 58, 317-344.

Lu S. L., Shao Y. J. & Feng c. (2015). Zhongguo zaixian

jiaoyu chanye lanpishu [Blue book of China's online

education industry]. Peking University Press.

MacIntyre, P. D., Noels, K. A., & Clément, R. (1997).

Biases in self‐ratings of second language proficiency:

The role of language anxiety. Language learning,

47(2), 265-287.

Markus, H., & Nurius, P. (1986). Possible selves. American

psychologist, 41(9), 954-969.

Papi, M. (2010). The L2 motivational self system, L2

anxiety, and motivated behavior: A structural equation

modeling approach. System, 38(3), 467-479.

Ryan, S. (2009). Self and Identity in L2 Motivation in

Japan: The Ideal L2 Self and Japanese Learners of

English. In Motivation, language identity and the L2

self (pp. 120-143). Multilingual Matters.

Shea, P. (2006). A study of students’ sense of learning

community in online environments. Journal of

Asynchronous Learning Networks, 10(1), 35-44.

Shih, H. J. (2019). L2 Anxiety, Self-Regulatory Strategies,

Self-Efficacy, Intended Effort and Academic

Achievement: A Structural Equation Modeling

Approach. International Education Studies, 12(3), 24-

35.

Thompson, T. L., & MacDonald, C. J. (2005). Community

building, emergent design and expecting the

unexpected: Creating a quality eLearning experience.

The Internet and Higher Education, 8(3), 233-249.

Tobias, S. (2013). Anxiety and cognitive processing of

instruction. In Self-related cognitions in anxiety and

motivation (pp. 45-64). Psychology Press.

Ueki, M., & Takeuchi, O. (2012). Validating the L2

motivational self system in a Japanese EFL context:

The interplay of L2 motivation, L2 anxiety, self-

efficacy, and the perceived amount of information.

Language Education & Technology, 49, 1-22.

Wu M. L. (2009). Jiegou fangcheng moxing: AMOS

decaozuo yu yingyong [Structural equation model: Tips

for practical application]. Chongqing University Press.

Young, D. J. (1991). Creating a low-anxiety classroom

environment: What does language anxiety research

suggest?. The modern language journal, 75(4), 426-

439.

CSEDU 2022 - 14th International Conference on Computer Supported Education

180