Investigating the Relationship among Students’ Interest, Flow and

Their Learning Outcomes in a Blended Learning Asynchronous

Forum

Shanyu Tan

, Zhi Liu

, Shiqi Liu

, Zhu Su

, Huanyou Chai

and Sannyuya Liu

1,2

National Engineering Research Center for E-Learning, Central China Normal University,

Luoyu Road 152, 430079 Wuhan, China

National Engineering Laboratory for Educational Big Data, Central China Normal University,

Luoyu Road 152, 430079 Wuhan, China

Keywords: Blended Learning, Asynchronous Forum, Learning Interest, Interest Mining, Flow Experience, Learning

Outcomes.

Abstract: Blended learning environment provide an important platform for university student learning. The use of text

data, generated from the asynchronous forum to explore students' intrinsic aspects of user posts in such com-

munities, is critical for adjusting teaching strategies. Therefore, information about their interest and flow in-

dicators has become important for educators to host online discussions. Flow experience is a sense of immer-

sive and feeling enjoyable and can reflect a person’s inner feelings. In order to explore the influence of learn-

ing interest and flow on the learning outcome, our study uses temporal emotion-aspect model (TEAM) to

mine student interest hidden in forum text data, and simultaneously uses a flow scale to measure the flow state

of students during their learning process. The results show that: 1) Interest topics unrelated to teaching content

are negatively related to learning outcomes. 2) Interest topics related to teaching content will provoke students’

ability to balance their skills and challenges, but have a negative effect on autotelic experience in the flow

experience. Interest topics related to entertainment have a negative effect on students' skills to meet the chal-

lenge, concentration and autotelic experience in discussion-based learning. Students may tend to lose self-

consciousness in the entertainment-centric discussion. 3) There influence factors between flow and learning

outcomes are loss of self-consciousness and concentration.

1 INTRODUCTION

Blended learning environment as a special campus

model, realize the organic integration of online course

resources and traditional classroom teaching on

campus, which help teachers optimize the teaching

methods and improve teaching quality. Blended

learning environments are courses targeted at

relatively fixed learning groups, so it can better

combine the online data, clicks, online duration, order

of lecture chapters, and offline research tools, such as

questionnaires, interviews, scales, etc. So, scientists

can conduct more accurate researches on students'

behaviors, emotions and other aspects.

Learning interest reflects a psychological

preference of learners in the process of learner

resource interaction. In general, learners’ interest can

be represented by their online behaviors in online

learning environments. The text data generated from

learning platforms provide researchers potential

opportunities to explore students’ interest, helping

teachers better control their courses procession,

which can facilitate timely intervention on students

with special needs.

Flow is an important indicator of describing

students’ subjective feelings in learning processes.

Having a flow experience means that when a student

takes part in an activity, he/she is devoted to the

current work and so calm that he/she forgets the time

passes, and feel a sense of space. Current researches

on flow focus on the following aspects: 1) Explore the

necessary conditions for arousing students' flow

experience in digital game learning; 2) The way to

improve the scale to measure flow more accurately.

In this way, educators can integrate digital games into

teaching more harmoniously and enhance students’

learning pleasure and performance.

In blended learning environments, asynchronous

discussion forums, as a communication tool, are used

Tan, S., Liu, Z., Liu, S., Su, Z., Chai, H. and Liu, S.

Investigating the Relationship among Students’ Interest, Flow and Their Learning Outcomes in a Blended Learning Asynchronous Forum.

DOI: 10.5220/0009339300340041

In Proceedings of the 12th International Conference on Computer Supported Education (CSEDU 2020) - Volume 2, pages 34-41

ISBN: 978-989-758-417-6

extensively to support students’ interaction and

engagement in online courses (Dringus & Ellis,

2010). Students have flexible time to control their

learning process in asynchronous forums. The defect

in asynchronous discussions is that communication is

not in time. Researchers have done a lot of work to

explore the influential factors of discussions, or

design tools to help students achieve collaborative

learning in asynchronous forums (Murphy, 2004;

Kear, 2004; Dennen, 2005). However, students’

interest, flow and their relationship with learning

outcomes are rarely studied. In view of the lack of

existing researches, this paper aims to explore

students' interest through the temporal emotion-

aspect model (TEAM) (Liu et al., 2019), revealing the

relationship among students' interest, flow and

learning outcomes. It can provide a reference for

teachers to understand students' interest and

indicators of their flow levels in blended learning

forums.

This paper is organized as follows. In the second

section, we review the methods of interest mining and

educational researches on flow. The third section is

the methodology and introduction of the experiment.

The fourth section summarizes the findings and

limitations of our study.

2 RELATED WORKS

2.1 Interest Mining

In online learning environments, students generate

large amounts of numerical and textual data. In a vir-

tual environment, Gu, Zhu, Zhao, & Zhang (2008)

used learners’ gazes, manipulations, gestures, dia-

logues and other behaviors to mine users' potential in-

terest via stages of web mining. Similarly, in E-learn-

ing and blended learning environments, the content of

posting in the course forum has also become a re-

search hotspot for many educational researchers. The

interaction between learners and forums is mainly to

browse, post and reply posts. In terms of the interac-

tion mode in forum postings, researchers analyzed the

social network and forum texts by calculating the

number of posts, time of posts and dialogue and other

basic learning records (Salter & Conneely, 2015; Liu

et al., 2018; Liu et al., 2019).

Based on the forum texts, this paper uses TEAM

to deduce students' implicit interest. TEAM stems

from the unsupervised emotional topic model named

Latent Dirichlet Allocation (LDA), which

automatically calculates the emotion-oriented aspect

probabilistic distributions over words for the overall

discussion. In fact, many researchers have proposed

relevant optimization based on LDA to quantify

factors of learning interest and preferences within

unstructured texts. For example, Jo & Oh (2011)

proposed SLDA (sentence-LDA), and then extended

SLDA to the Aspect and Sentiment Unification

Model (ASUM). Its outputs pairs of {aspect,

sentiment} called senti-aspects, automatically

discovering what aspects are evaluated in reviews and

how sentiments for different aspects are expressed.

Pengfei Wu, Shengquan Yu, & Dan Wang (2018)

used a learner-topic model, which combining learners

generated content and their dynamic interactions with

learning resources. They mined learners' knowledge

interest and collection interest, then combined them

to generate keywords.

2.2 Flow Experience

Flow experience refers to a positive experience pro-

duced by an individual when he/she is dedicated to

tasks. This experience makes people forget the pass-

ing of time, lose the sense of space and immerse

themselves in the enjoyment (Csikszentmihalyi,

1990). In education, the flow experience is often stud-

ied in gamification-based learning contexts, and flow

is regarded as a measure of design levels of digital

games (e.g. Kaur, Dhir, & Rajala, 2016; Perttula,

Kiili, Lindstedt, & Tuomi, 2017; Buil, Catalán, &

Martínez, 2018).

Csikszentmihalyi (2006) divides the state of flow

into nine dimensions: 1) Flow seems to occur when

individuals are well balanced the challenge of tasks

and their skills; 2) Individuals experience a

spontaneous and automatic sense when doing an

activity; 3) An activity has specific goals; 4)

Unambiguous feedbacks as to how well one is

performing; 5) A sense of control is needed; 6) A state

of focused concentration on things on hand; 7) Loss

of self-consciousness; 8) There is a distortion of the

transformation of time; 9) Individuals enter a state of

autotelic experience, showing activities are thought as

intrinsically rewarding. The first five dimensions can

be reduced to flow antecedents The remaining four

dimensions represent indicators of flow. The nine

dimensions have been the basis for different

researchers to measure flow levels and its indicators

(e.g., Jackson & Marsh, 1996; Jackson & Eklund,

2002; Kiili, 2005; Fu, Su, & Yu, 2009; Hamari &

Koivisto, 2014;).

Exiting researches have shown the relationship

between these dimensions (Buil, Catalán, &

Martínez, 2019). Moreover, students with higher flow

levels tend to obtain higher learning outcomes in

Investigating the Relationship among Students’ Interest, Flow and Their Learning Outcomes in a Blended Learning Asynchronous Forum

game-based learning, using eye-tracking technology

(Tsai, Huang, Hou, Hsu, & Chiou, 2016).

Prior researches have shown that gaming

experience, age, and gender have been verified to be

nothing to do with flow levels (Kiili, 2006). Hence,

an understanding of what elements provide flow

experience and what learners’ real interest in blended

learning environments asynchronous forum, can be

considered as critical factors for adjusting teaching

strategies to ensure learners’ active participation and

outcomes. Interestingly, existing researches

concerning flow experience have focused mainly on

games. To address this research gap, this study used

TEAM to mine students' interest and used a flow scale

to test students' flow levels, in an attempt to explore

the relationship between learners' interest, flow and

learning outcomes. Our study aims to answer the

following research questions:

(1) Does the mined interest topics reflect the real in-

terest of learners? What is the relationship be-

tween learner interest and learning outcomes in

the blended learning environment?

(2) Is there any relationship between learners’ inter-

est and flow experience in the asynchronous fo-

rum?

(3) Which indicators of flow are significantly re-

lated to learning outcomes?

3 METHODS

3.1 Participants

The data set of this paper is retrieved from the course

"Freshmen Seminar" on a university blended learning

platform, ranging from September 2018 to January

2019. This course is aimed at freshmen students ma-

joring in English. The purpose of offering this course

is to discuss the cultivation of professional skills and

to communicate some issues about career planning. A

total of 66 students enrolled in this course. Our col-

lected data included 4880 posts with a standard devi-

ation of 46.39 and an average of 73.94.

The data also included 66 valid questionnaires to

measure their levels of flow in the course. The aver-

age final grade of the students is 89.04 (on a scale of

0-100), and the standard deviation is 4.73.

3.2 Materials and Instruments

3.2.1 Temporal Emotion-aspect Model

This paper uses TEAM to calculate learners' interest

topics in the sense of probabilistic distribution.

TEAM assumes that the words in a single sentence

are drawn from one aspect and one emotion. The

same pair of {emotion, aspect} called emot-aspect.

TEAM belongs to an unsupervised model. So, it

doesn’t need manual aggregations on emot-aspect as-

sociations of the posts at the same time zone. And no

post-processing is required to calculate the emotion

orientations of different semantic units to aspects un-

der different emotion labels. It can output emotion-

specific aspect probabilistic distributions. TEAM uti-

lizes Gibbs sampling to estimate the hidden parame-

ters.

3.2.2 Flow Scale

This study translated and modified the Flow Scale for

Games (FSG) to measure students' state of flow in the

blended learning environments forum. The scale con-

sists of 25 questions, 23 single-choice questions, and

2 open-ended questions. The table uses a five-point

Likert scale, and scores of different questions are

combined to represent students' preferences towards

different dimensions of flow. Our analysis of the

learners’ questionnaire shows the internal con-

sistency to be 0.80 (Cronbach's alpha is 0.80), indi-

cating that our scale has good reliability.

3.2.3 Learning Outcomes

This study uses learners’ final grade as an operational

definition of learning outcomes because there is no

intervention in their performance. So, our hypothesis

is that, in the state of nature and non-intervention, the

final grade in the course reflects their consistent

learning habits and attitudes. And there will be a cer-

tain correlation with their learning outcomes. So

course’ final grade can represent learning outcomes.

4 RESULTS

4.1 Mined Interest Topics

Interest can be represented as a topical word that co-

occurs with positive emotions. This study uses the

positive emotion dictionary as a seed lexicon to cap-

ture the words related to a positive learning experi-

ence, by calculating the probability of each topic ap-

pearing in a forum post. Finally, we obtain 50 interest

topics, and make them as





12350

...TTTTT ，，

To examine the reliability and validity of the

obtained interest, we use a post-test questionnaire to

inquire about what topics they really were interested

in. Comparing the deduced interest with the self-

CSEDU 2020 - 12th International Conference on Computer Supported Education

reported interest collected by the questionnaire, we

can obtain that the average accuracy rate of the

interest mining reaches 0.799. From this matching

result, we can find that the modeled interest topics can

indeed reflect the interest of students to an extent.

Table 1 shows some of the partial results of mined

interest topics. For example, topic 7 represents

reading, and the keywords are “impression”, “book”,

“reading” and so on. Among them, underlined words

represent positive emotional words, such as

“profound”, “like”.

4.2 Interest Topics That Are Not

Related to Teaching Content Are

Negatively Related to Learning

Outcomes

The descriptive results of learnings’ interest and

learning outcomes are shown in Table 2. To explore

their relationship, multiple regression analysis is per-

formed by using interest as the independent variable

and learning outcomes as the dependent variable.

Since there are too many interest topics, it is not easy

to explore all the mined interest. So typical interest

topics, T6, T7, T8, T13, T44, T45 and T50, are se-

lected as the dependent variables. T13 and T50 talk

about the English level, and other topics don’t have

much to do with learning itself. T7 and T44 talk about

literature such as novels or movies. T6, T8, T45 talk

about professionalism, life values, mentality. In our

study, it proves that the model uses these topics to ex-

plain the learning outcome has 68.5% explanatory

power, and the adjusted R

indicates still has 64.7%

explanatory power.

It can be observed that T6, T7, T8, T44 and T50

are negatively correlated with learning effectiveness,

and T13 is positively correlated with learning

outcomes. Comparing and analyzing the above topics

with teachers' postings, it is found that the topics

are positively related to learning outcomes and

are significantly relevant to the teaching content of

the

teacher, such as T13 (β = 34.8, p = 0.001). The

Table 1: Interest topics vocabularies.

Interest topic Top 10 words with the highest probabilities

Reading books T7

印象/impression (0.034), 深刻/profound (0.029) 书/book (0.028) , 读/read (0.014),

文学/literature (0.015), 作品/works (0.011), 故事/story (0.011), 喜欢/like (0.010),

王子/prince (0.009), 小说/novel(0.009)

English learning T13

learning (0.035), study (0.022), know (0.017), English (0.017), agree (0.013), good

(0.013), important (0.012), strategies (0.008), plan (0.007), improve (0.007)

Mentality T45

literature (0.059), reading (0.059), great (0.055), 问题/question (0.014), 心态

/mentality (0.012), 锻炼/exercise (0.010), 强化/strengthen (0.010), 实践/practice

(0.010), 冷静/calm (0.008), 解决/solve (0.006)

Table 2: Regression coefficients of interest topics on learning performance.

Unnormalization coefficient Normalization coefficient

B S

Beta

(const)

91.726

***

0.978 93.742

T50 English level

-80.092

***

20.596 -0.327

***

-3.889

T13 English learning

34.800

9.664 0.303

3.601

T44 Literature

-587.084

***

155.492 -0.294

***

-3.776

T8 Life values

-42.516 22.754 -0.168 -1.868

T7 Novel

-109.380

30.995 -0.278

-3.529

T45 Mentality

-174.233

58.212 -0.237

-2.993

T6 Professionalism

-38.604

16.508 -0.209

-2.338

Note: ***p < 0.001, **p < 0.01, *p < 0.05

Investigating the Relationship among Students’ Interest, Flow and Their Learning Outcomes in a Blended Learning Asynchronous Forum

content of this topic is all written in English, sharing

their own English learning experience, content and

career planning. Topics that are negatively related to

learning outcomes and not related to the teaching

content of the teacher, e.g., T44 (β = -587.084, p =

0.000), mainly involve the real-life interest, such as

watching movies or reading novels. T45 (β = -

174.233, p = 0.004) talks about mentality, and so on.

However, some topics related to teaching content are

not strongly explanatory to the negative correlation

between teaching content and their grades. This issue

needs to be further studied in combination with

offline students’ learning situation.

4.3 Descriptive Result of Flow Levels

As can be seen from Table 3, students' flow tests

showed that the average flow score is 3.55, higher

than the median 3.00 (median of a 5-point Likert

scale), indicating that most of the students have a pos-

itive flow state in this course. And the Cronbach’s al-

pha estimate of the reliability of flow antecedents and

experience are found reasonable (α = 0.79, α = 0.64).

Only the reliability of loss of self-consciousness is

relatively poor.

Compared with the score of flow experience, the

score of antecedents is relatively lower. We can tell

that tasks assigned by teachers in the blended learning

environment are specific, and students are satisfied

with the teaching platform. The score of loss of self-

consciousness is lower than others (Mean = 2.32).

The explanation given by the students in the

questionnaire are summarized as follow: 1) The

discussion topic given by the teacher are too boring.

2) Some students are not interested in some

discussions. 3) Some students are disturbed by

external things such as mobile phones and urgent

matters. 4) Some students lose concentration and

inquiry into the problem.

4.4 The Relationship between Interest

and Flow

In order to explore the relationship between student

interest and flow, this study performs a Pearson cor-

relation analysis between flow components and key

interest topics.

As shown in Table 4, T6 representing professional

skills is positively correlated with Q1 representing

learning

challenges, with a correlation coefficient of

Table 3: Description of low dimensions included in the FSG (N = 66).

Element

Item number

Flow dimension Min Max Mean

Std

Dev.

Flow

antecedents

1, 10

Challenge 2.00 5.00 3.69 0.62 0.54

3, 12 Goal 2.00 5.00 4.06 0.59 0.47

4, 13 Feedback 2.00 5.00 3.71 0.75 0.65

6, 15 Control 3.00 5.00 4.10 0.51 0.55

2, 11 Playability 2.00 5.00 4.19 0.66 0.83

Indicators of flow

experience

5, 14, 19, 21

Concentration 1.25 4.25 3.12 0.64 0.57

8, 17 Time distortion 1.50 5.00 3.41 0.77 0.57

9, 18, 20, 22 Autotelic experience 1.75 4.75 3.40 0.84 0.90

7, 16 Loss of self-consciousness 1.00 4.50 2.32 0.75 0.42

Note: α = Cronbach’s alpha

Table 4: Pearson coefficients between interest and flow.

Interest content Interest topic Flow state r

Competence

Professionalism T6 Challenge Q1

0.245

English level T50 Autotelic experience Q18

-0.249

Entertainment

Novel T7

Challenge Q10

-0.252

Movie T44

Loss of self-consciousness Q7

0.445

Concentration Q19

-0.278

Autotelic experience Q22

-0.254

Note: **p <0 .001, *p < 0.05

CSEDU 2020 - 12th International Conference on Computer Supported Education

0.245 (p < 0.05). And T50 representing the English

level is negatively correlated with Q18 representing

autotelic experience, with a correlation coefficient of

-0.249 (p < 0.05). The correlation coefficient between

T7 for novel and Q10 for challenge is -0.252 (p <

0.05), the correlation coefficient between T44 for

entertainment and Q7 for Loss of self-consciousness

is 0.445 (p < 0.001), the correlation coefficient

between Q19 for concentration is -0.278 (p < 0.05),

and Q22 for autotelic experience reaches -0.254 (p <

0.05).

Based on the interest explored from this forum, it

is found that most types of interest topics do not affect

the level of flow, but the interest related to

professional skills’ development and entertainment

will significantly affect the level of flow. Discussions

involving relevant professional knowledge need to

apply a large amount of knowledge, and the discussed

problems are difficult, which requires students to

achieve a balance between challenges and skills. In

the process, students need to stop and think, which

will affect their sense of intrinsically rewarding to a

great extent. In the same way, when a learner focuses

on irrelevant content, he/she will ignore the teaching

content, not pay attention to the challenge of the task.

Because be interrupted by distraction, it results in

incoherent behaviors.

4.5 The Relationship between Flow and

Learning Outcomes

Similarly, we try to use correlation analysis to detect

the relationship between flow and learning outcomes.

The result shows that Q7, represents the loss of self-

consciousness, is negatively correlated with learning

outcome (r = -0.242, p < 0.05). Q19, that represents

concentration, is positively correlated with the learn-

ing outcome. (r = 0.278, p < 0.05). The result shows

that most of the flow indicators may not influence

learning outcomes. But, the remarkable factors about

flow and learning outcomes are self-conscious and

concentration. When focusing on what they are doing,

students may not be interrupted by the outside. In the

learning process, they need to be clearly aware of

their actions, and correct any deviations from their

goals.

5 DISCUSSIONS

In a long-time, leaning performance can only be rep-

resented through final grades. So, learning processes

seem to be a black box that no one knows what hap-

pened in it. Through exploring which factors may

dominate learning outcomes, teachers can compre-

hend their students’ learning processes and adjust

their teaching methods.

There is a phenomenon that when we are fully en-

gaged in current activities, we will forget the passage

of time. Even in a noisy environment, we can still feel

peace of mind and realize where we are and marvel at

the passage of time when the task is completed. For

this strange phenomenon, Csikszentmihalyi named it

to flow experience for the first time. Therefore, as a

perspective to evaluate students' intrinsic motivation,

flow levels may help researchers to explain students'

external performance, such as performance and be-

havior.

This present study innovatively introduced the

concept of flow into the forum discussions, attempt-

ing to examine the relationship between learners' in-

terest, flow and learning outcomes in a blended learn-

ing platform. To the best of our knowledge, our study

addressed the prior literature gap of examining stu-

dents’ flow experience. To test students’ flow levels,

we use Flow Scale for Games (eg., Hou, 2015; Hsieh

et al., 2016). The scale contains two aspects: flow an-

tecedents and indicators of flow experience.

Results show that combined with the post-mortem

test, TEAM (Liu et al., 2019) can be effectively used

to mine students' interest. In terms of interest and

learning outcomes, we find interest topics related to

teaching content is positively correlated with the

learning outcome, and vice versa. With regard to the

relationship between interest and flow, flow’s indica-

tor of loss of self-consciousness is negatively corre-

lated with their learning outcomes, and concentration

is positively correlated with their learning outcomes.

The thing maybe that students need to choose appro-

priate skills to deal with challenges related to the cul-

tivation of professional skills. Otherwise, they will

lose the sense of challenge if they are interested in en-

tertainment information, which distracts their atten-

tion from studies, interrupting the continuity of learn-

ing. According to our results, the relationship be-

tween flow and learning outcomes is influenced by

facts, loss of self-consciousness and concentration.

To improve students’ learning outcomes, here are

some pedagogical practices for teachers. Teachers

should pay attention to discussion processes in asyn-

chronous forums, and irregularly interact with their

students keeping discussion topics related to teaching

content. Teachers can enhance students’ learning out-

comes by adjusting the flow’s level. When feeling a

sense of loss of self- conscious, the major of students

Investigating the Relationship among Students’ Interest, Flow and Their Learning Outcomes in a Blended Learning Asynchronous Forum

may not immerse themselves in their studies but en-

tertainments. To avoid be attract, there may often

mind themselves to concentrate on their work. So, in

long-time discussions, teachers can provide some

useful information to attract them, or ask challenging

questions to inspire their curiosity.

Although a limited number of samples are

involved in the experiment, combined with the

postings of each student to describe their personal

flow experience, the results obtained in this study

after data analysis are explanatory and consistent with

the real feelings of students participating in the course

discussion. But it should be noted that this study has

some limitations. In the courses, posts account for

only 15% of the final grade. When trying to use the

mined interest topics in the forum as independent

variables to interpret students' learning outcomes,

small parts of interest topics are not very explanatory,

such as T50. Therefore, more factors need to be added

to explain overall learning outcomes.

ACKNOWLEDGEMENTS

This work was supported by the Research Funds from

National Natural Science Foundation of China (grant

number: 61702207, 61937001, 61977030,

31600918), National Key Research & Development

Program of China (grantnumber:2017YFB1401303),

Hubei Provincial Natural Science Foundation of

China (grant number: 2018CFB518).

REFERENCES

Buil, I., Catalán, S., & Martínez, E. (2018). Exploring stu-

dents’ flow experiences in business simulation games.

Journal of Computer Assisted Learning, 34(2),183–92.

https://doi.org/10.1111/jcal.12237

Buil, I., Catalán, S., & Martínez, E. (2019). The influence

of flow on learning outcomes: An empirical study on

the use of clickers: The influence of flow on learning

outcomes. British Journal of Educational Technology,

50(1), 428–439. https://doi.org/10.1111/bjet.12561

Csikszentmihalyi, M. (1990). Flow – The Psychology of op-

timal experience. 9.

Dennen *, V. P. (2005). From message posting to learning

dialogues: Factors affecting learner participation in

asynchronous discussion. Distance Education,

26(1),127-148.

https://doi.org/10.1080/01587910500081376

Dringus, L. P., & Ellis, T. (2010). Temporal transitions in

participation flow in an asynchronous discussion fo-

rum. Computers & Education, 54(2), 340–349.

https://doi.org/10.1016/j.compedu.2009.08.011

Hsieh, Y.-H., Lin, Y.-C., & Hou, H.-T. (2016). Exploring

the role of flow experience, learning performance and

potential behavior clusters in elementary students’

game-based learning. Interactive Learning Environ-

ments, 24(1), 178–193. https://doi.org/10.1080/

10494820.2013.834827

Fu, F., Su, R., & Yu, S. (2009). EGameFlow: A scale to

measure learners' enjoyment of e-learning games. Com-

puters & Education, 52, 101–112.

https://doi.org/10.1016/j.compedu.2008.07.004

Gu, R., Zhu, M., Zhao, L., & Zhang, N. (2008). Interest

mining in virtual learning environments. Online Infor-

mation Review, 32(2), 133–146. https://doi.org/

10.1108/14684520810879782

Hamari, J., & Koivisto, J. (2014). Measuring flow in gami-

fication: Dispositional flow scale-2. Computers in Hu-

man Behavior, 40, 133–143. https://doi.org/

10.1016/j.chb.2014.07.048

Jackson, S., & Eklund, R. (2002). Assessing flow in physi-

cal activity: The flow state scale-2 and dispositional

flow scale-2. Journal of Sport and Exercise Psychol-

ogy, 24, 133–150. https://doi.org/10.1123/

jsep.24.2.133

Jackson, S., & Marsh, H. (1996). Development and valida-

tion of a scale to measure optimal experience: The flow

state scale. Journal of Sport &Exercise Psychology, 18,

17–35.

Jo, Y., & Oh, A. H. (2011). Aspect and sentiment unifica-

tion model for online review analysis. Proceedings of

the Fourth ACM International Conference on Web

Search and Data Mining - WSDM ’11, 815.

https://doi.org/10.1145/1935826.1935932

Kaur, P., Dhir, A., Chen, S., & Rajala, R. (2016). Flow in

context: Development and validation of the flow expe-

rience instrument for social networking. Computers in

Human Behavior

, 59, 358–367. https://doi.org/

10.1016/j.chb.2016.02.039

Kaur, P., Dhir, A., & Rajala, R. (2016). Assessing flow ex-

perience in social networking site based brand commu-

nities. Computers in Human Behavior, 64, 217–225.

https://doi.org/10.1016/j.chb.2016.06.045

Kiili, K. (2005). Digital game-based learning: Towards an

experiential learning model. The Internet and Higher

Education, 8, 13–24. https://doi.org/10.1016/

j.iheduc.2004.12.001

Kiili, K. (2006). Evaluations of an Experiential Gaming

Model. Human Technology: An Interdisciplinary Jour-

nal on Humans in ICT Environments, 2(2), 187–201.

https://doi.org/10.17011/ht/urn.2006518

Liu, S., Chai, H., Liu, Z., Pinkwart, N., Han, X., & Hu, T.

(2019). Effects of Proactive Personality and Social

Centrality on Learning Performance in SPOCs: Pro-

ceedings of the 11th International Conference on Com-

puter Supported Education, 481–487.

https://doi.org/10.5220/0007756604810487

Liu, Z., Kang, L., Domanska, M., Liu, S., Sun, J., & Fang,

C. (2018). Social Network Characteristics of Learners

in a Course Forum and Their Relationship to Learning

CSEDU 2020 - 12th International Conference on Computer Supported Education

Outcomes: Proceedings of the 10th International Con-

ference on Computer Supported Education, 15–21.

https://doi.org/10.5220/0006647600150021

Liu, Z., Yang, C., Rüdian, S., Liu, S., Zhao, L., & Wang, T.

(2019). Temporal emotion-aspect modeling for discov-

ering what students are concerned about in online

course forums. Interactive Learning Environments,

27(5–6), 598–627. https://doi.org/10.1080/

10494820.2019.1610449

Murphy, E. (2004). Recognising and promoting collabora-

tion in an online asynchronous discussion. British Jour-

nal of Educational Technology, 35(4), 421–431.

https://doi.org/10.1111/j.0007013.2004.00401

Pengfei Wu, Shengquan Yu, & Dan Wang. (2018). Using a

Learner-Topic Model for Mining Learner Interests in

Open Learning Environments. Journal of Educational

Technology & Society, 21(2), 192–204.

Perttula, A., Kiili, K., Lindstedt, A., & Tuomi, P. (2017).

Flow experience in game based learning–a systematic

literature review. International Journal of Serious

Games, 4(1), 57-72.

Salter, N. P., & Conneely, M. R. (2015). Structured and un-

structured discussion forums as tools for student en-

gagement. Computers in Human Behavior, 46, 18–25.

https://doi.org/10.1016/j.chb.2014.12.037

Tsai, M.-J., Huang, L.-J., Hou, H.-T., Hsu, C.-Y., & Chiou,

G.-L. (2016). Visual behavior, flow and achievement in

game-based learning. Computers & Education, 98,

115–129. https://doi.org/10.1016/

j.compedu.2016.03.011

Investigating the Relationship among Students’ Interest, Flow and Their Learning Outcomes in a Blended Learning Asynchronous Forum