1. Introduction
On the basis of Dewey’s community-of-inquiry theory, the community-of-inquiry (CoI) theory of Garrison et al. was proposed in 2000 [
1]. It involves cognitive presence, teaching presence, and social presence, and it is a theoretical model and practical framework to effectively guide learners in deep learning [
1,
2]. CoI theory originated from the research of the online learning community, and it is widely adopted by researchers and educators for online and blended learning environments [
3]. At the same time, it is sometimes used for face-to-face learning environments in universities [
4]. Wang Q., from Peking University, said that the value of CoI would be significantly underestimated if it was only limited to the online environment, and that CoI theory was also effective at improving the teaching quality and enhancing the learning experience in traditional face-to-face teaching [
5]. This paper exactly discusses the application of CoI theory in the classroom settings of universities, with mainly face-to-face learning, and supplemented by online and offline blended learning.
Feedback is one of the factors that influence the academic achievement of learners. “Feedback literacy”, as an academic concept, was firstly proposed in 2012 [
6]. The research on it mainly focuses on how to provide feedback to learners, and on how learners absorb feedback, but it ignores the learner’s initiative in terms of self-feedback and feedback output, which can bring “mutual benefit of teaching and learning” [
6]. Feedback from students after learning is an important generative resource and is the basis for the promotion of teaching and learning. In this paper, the data from the learners’ feedback at the end of the course are used as the research material.
On the basis of the CoI framework and the learner-feedback data, this research examines the features of presence in classroom settings to assist teachers in optimizing course design and implementation for enhancing presence, improving learning experiences, and promoting high-quality learning. It also provides a new approach and perspective for qualitative text analysis, as well as a reference for related research and practices, by serving as an example of data analysis and data-based precision instruction.
2. Review of the CoI Framework
The CoI framework involves three elements: cognitive presence, teaching presence, and social presence. These presences are interconnected and influence one another [
7], and there is a significant positive correlation between them [
3,
8]. The CoI theory’s goals are to promote deep and meaningful learning, and to assist learners in engaging in collaborative constructive learning and critical reflective dialogue in order to cultivate critical and higher-order thinking. However, during the course of their ongoing investigation of CoI theory, the researchers discovered that the model’s constituent elements needed to be expanded further in order to improve the theory’s descriptive and explanatory power. Cleveland Innes et al. believe that emotional presence is another essential element of the CoI theory [
9]. Although social presence includes emotional response, emotional presence is independent to some extent. Later, Rienties et al. incorporated emotional presence into the CoI framework and reconstructed the CoI theoretical model [
10]. Emotion is the learners’ experiences of their attitudes towards objective things, and it plays an important role in facilitating thinking, judgment, memory, and information processing [
11]. Learners’ affective experiences are the motivation basis of cognition, and they are also a prerequisite for and a bridge to teaching effectiveness. In this paper, emotional presence is considered equally with cognitive presence, teaching presence, and social presence.
Cognitive presence is rooted in Dewey’s theory of practical inquiry, and it is associated with the learning results of cognitive skills, as well as with higher-order-thinking formation and the occurrence of deep learning. It has been suggested that teachers follow the four phases of the practical-inquiry model: a triggering event, exploration, integration, and resolution [
12]. A triggering event refers to asking questions and to changing the direction of a specific conversation to stimulate learners to start a new discussion or to engage in a new activity. Exploration involves reviewing previous knowledge, brainstorming, and exchanging information, which is manifested in the self-questioning of learners to seek opinions and explore the truth. Integration consists of combining thoughts in order to make them operational. Resolution actually refers to problem solving, and it includes analyzing problems, formulating potential solutions, and applying them. Garrison believes that cognitive presence is very important for achieving meaningful deep learning, and that questioning, exploring, integrating, and sharing play important roles in the construction of cognitive presence.
Teaching presence refers to the guidance of teachers and to their attention to learning [
5], and it is realized by the teachers’ activity planning, learning guidance, atmosphere creation, and direct instruction. Teaching presence includes three categories: design and organization, the facilitation of discourse, and direct instruction [
13]. Design and organization refer to setting goals, selecting methods, contents, and resources, and organizing activities and interactions. The facilitation of discourse involves providing support to stimulate learners’ construction and inquiry, such as establishing a safe and trustworthy atmosphere for discussion, assessing the learning process, or confirming the understanding level. Direct instruction means providing instruction with the teachers’ characteristics, which includes sharing specific topics directly with learners, providing problem-solving steps, and summarizing discussions. To explain the perception of the teaching presence, Wang Q. uses the word “see” [
5]. She believes that teaching presence indicates that teachers should try their best to let students see their efforts and their contributions [
5].
Social presence provides learners with a learning environment for collaborative interaction, knowledge construction, and critical thinking, where learners can communicate purposefully, show their personalities, and establish interpersonal relationships. Social presence is reliant upon emotional expression, open communication, and relationship cohesion [
14]. In this paper, we adopt the views of Cleveland-Innes et al. and Rienties et al., who take emotional presence as an independent essential element [
9,
10]. Therefore, social presence contains two main categories: open communication and relationship cohesion. Open communication refers to friendly participation and free expression in communication, which involves confirmation, praise, and some trivial expressions. Relationship cohesion refers to the gradual development of a cooperative–constructive relationship between teachers and students. Establishing a sense of belonging, maintaining community learning, greeting or addressing each other, and establishing relationship links with each other will strengthen this social communication and interpersonal relationship. Garrison finds that social presence does not directly contribute to deep learning and cognitive presence, but that it is useful for sharing critical views and for generating challenging ideas. It is a necessary condition for cognitive presence and deep learning [
3].
Emotional presence is an outward expression of emotion between individuals and individuals. It interacts with the learning technology, course content, students, and teachers, and it includes three categories: activity emotion, outcome emotion, and directed emotion [
9]. Activity emotion refers to the emotion in the process of inquiry, which is the emotional response to ongoing activity and process. Outcome emotion is expressed as the outcome of the inquiry and the result response to success or failure in problem solving. Directed emotion is an interpersonal emotion or mood. According to the CoI theory, there is a combined relationship between emotional presence and the cognitive, teaching, and social presences. These four presences influence each other, and they cumulatively promote meaningful deep learning.
The existing empirical studies that are related to the CoI framework mainly focus on the relationship between presences, how it is generated interactively [
3,
8,
15], the impact of presence on learning cognition and the learning effect [
14,
15,
16], and the analysis of the difference in the presences in different environments [
4,
17,
18]. Stefan Stenbom et al. conducted a transcript-coding analysis that is based on CoI theory on one-to-one online-mathematics-coaching discourse [
19]. Most of the previous studies focus on online or blended learning, while the face-to-face learning environment is less common; most of the adopted methods focus on quantitative research that is based on scale questionnaires, while the qualitative research is less common. This paper is set in a higher-education classroom, and the analysis is based on qualitative feedback data. These data are proof of each presence’s existence. The researchers identify the presences from the data and make them visible. Understanding the relationship between the cognitive, teaching, social, and emotional presences not only adds to the CoI framework, but it also assists college teachers in the optimization of their course design and in the implementation of precision teaching.
3. Materials and Methods
3.1. Research Question
The purpose of this paper was to analyze learners’ text feedback after the completion of the course by using the CoI framework in order to understand the characteristics of presence, and to provide support for university teachers to enhance presence and to improve the learning experience. The specific inquiries are as follows: (1) Is the CoI framework suitable for the classroom settings of colleges or universities? Does each presence exist in this case? (2) To what extent did the students experience the levels of CoI (cognitive, teaching, social, emotional) after participating in the class? (3) How do the levels of the CoI compare with respect to the categories in each dimension after participating in the class? (4) On the basis of the characteristics of presence, what support can be provided for university teachers to optimize course design and to implement precision teaching?
3.2. Participants and Data Collection
The work in this research was carried out on the basis of the course of Modern Educational Technology at Hangzhou Normal University, China. This course is an essential course for normal colleges or universities, and it is an important way to cultivate the informatization teaching abilities of preservice teachers. The course consists of 48 lessons that are divided into 16 weeks, with 3 lessons per week. The main contents of the course include an introduction, the acquisition and utilization of informatization learning resources, the design and development of multimedia courseware, the learning space and the smart-learning environment, the informatization teaching design and evaluation, cases of informatization teaching activities, and maker education. The course is primarily taught face to face, and it is supplemented by online and blended learning that is based on resources and tasks. It is currently the most widely used teaching method among university teachers.
A total of 1319 students from 17 majors, which are divided into 34 classes, learned the course of Modern Education Technology at Hangzhou Normal University in 2021. Among the students, there were 8 freshmen, 617 sophomores, 692 juniors, and 2 seniors. The course was taught by 13 teachers, 11 of whom were from educational technology, and 2 of whom were from computer science and technology.
The after-learning feedback from students is an important generative resource and is the foundation for the promotion of teaching and learning. Following the completion of the course, students were invited to submit their text responses to the following questions online: “What are your thoughts on the Modern Educational Technology course? What has worked well? What should be improved?”
A total of 13 of 34 classes, which represented 13 different majors, were chosen, with approximately 40 students in each class, and a total of 554 students. Each class was instructed by a different teacher. The 13 majors were composed of Chinese Language and Literature, Mathematics and Applied Mathematics, English, Physics, Chemistry, Science, Humanities Education, History, Computer Education, Arts Education, Physical Education, Elementary Education, and Preschool Education. After removing the meaningless feedback, 476 pieces of valid text feedback were collected. The number of words in each response ranged from 20 to 853, for a total of 57,599 words, and an average of 121 words for each student.
3.3. Research Method and Procedures
Text analysis is used in this paper. The validity and reliability of text analysis depends on a valid and reliable theoretical framework, and on the rationality of the coding scheme [
20]. The CoI framework has been extensively tested by scholars since it was proposed in 2000, which has ensured its high validity and reliability for text analysis [
21,
22]. Stefan Stenbom et al. used the CoI framework to analyze teacher–student conversations in a one-to-one online-mathematics-tutoring environment, and they provide detailed indicators for each category of the cognitive, teaching, social, and emotional presences [
19]. The messages of the teacher–student conversations were then classified as analysis units in the CoI framework. The learners provided both positive and negative feedback, as well as suggestions. As a result, this study first identified four thematic dimensions: “cognitive, teaching, social, and emotional”, and it then classified each dimension into three categories: positive, negative, and suggestion. The positive aspect reveals the presence’s existence. Each presence was classified on the basis of its category, and then the indicators were listed, and examples were given according to the meanings of the categories. During the development of the coding scheme, we partially adopted the coding scheme of Stefan Stenbom et al., but we also made modifications and adaptations for university classroom environments, as is shown in
Table 1.
Message, paragraph, and sentence units, as well as thematic units, have been commonly used in research that uses the CoI framework. The short sentence with complete meaning was chosen as the unit of analysis for this study. All the categories in
Table 1 were reviewed and were coded as either existing or not for each short sentence. This allowed a short sentence to be coded with more than one category to reflect the framework’s interrelationship of dimensions and categories. For example, the short sentence “Groups exchange feedbacks” has two codes and belongs to both the cooperative exploration of cognitive presence and the relational cohesion of social presence. A short sentence with one code resulted in the assignment of 1 point to a corresponding category, a short sentence with two codes resulted in the assignment of 0.5 point to each corresponding category, a short sentence with three codes resulted in the assignment of 0.33 point to each category, and so on. There were no short sentences with more than three categories found. Dimensional coding was accomplished by grouping codes at the category level with their corresponding dimensions. This was performed to allow for analysis at both the category and dimension levels.
The first and second authors participated in the coding. We first read all of the text feedback carefully in order to negotiate the coding scheme. The process of coding and quantization required subjective analysis and assignment by the researchers, which was difficult. To ensure reliability and consistency, we independently coded 50 identical pieces of text feedback. At the category level, the percent agreement of this study was 0.72, and, at the dimension level, the percent agreement of this study was 0.85. This was then followed by the coders each coding half of the feedback. Any uncertainties were determined through consultation and discussion.
4. Results
The coding scheme that is shown in
Table 1 was used in this paper, and the short sentence with the complete meaning was chosen as the analysis unit to encode 476 pieces of text feedback. In total, 1695 short sentences were coded, with 1124 of them having a code and being assigned 1 point to the corresponding category. The 459 short sentences had two codes, and each category was assigned 0.5 point. The 112 short sentences had three coding categories, and each category was assigned 0.33 point. There are a total of 2378 instances of various categories.
Table 2 presents the study’s findings at the dimension level. For 13 classes, the mean value is the sum divided by 13. The sum of the positive, negative, and suggestion values shows that the teaching dimension is the most common, followed by the emotional and cognitive dimensions, and the social dimension is the least common. According to the sum of the positive value, emotional presence is the most common, followed by cognitive presence and teaching presence, and social presence is far less common than the other three. Furthermore, the negative and suggestion values of the teaching dimension were significantly higher than the other ones.
As for the positive part of the four dimensions, which is, namely, each presence, the general linear model was used for multivariate variance analysis, and Wilks’ Lambda = 0.102, F(3, 10) = 29.209, p < 0.001, and η2 = 0.898. The results show that the four presences are significantly different at the significance level of 0.001. As for the negative part of the four dimensions, the general linear model was used for multivariate variance analysis, and Wilks’ Lambda = 0.285, F(3, 10) = 8.367, p = 0.004 < 0.05, and η2 = 0.715. This indicates that the negative part of the four dimensions is significantly different at the significance level of 0.05. As for the suggestion part of the four dimensions, the general linear model was used to conduct a multivariate test, and Wilks’ Lambda = 0.235, F(3, 10) = 10.857, p = 0.002 < 0.05, and η2 = 0.765. This indicates that there are significant differences in the suggestion part of the four dimensions at the significance level of 0.05.
Table 3 presents the descriptive statistics of the cognitive presence. Resolution is much more common than the other categories, followed by exploration and integration, with the least common being triggering events. Multivariate variance analysis by using the general linear model showed Wilks’ Lambda = 0.364, F(3, 10) = 5.814,
p = 0.015 < 0.05, and η
2 = 0.636. The results show that there are significant differences among the four categories of cognitive presence at the significance level of 0.05.
Table 4 presents the descriptive statistics of the teaching presence. The category of design and organization is significantly more common, while facilitation is much less common than the others: Wilks’ Lambda = 0.123, F(2, 11) = 39.289,
p < 0.001, and η
2 = 0.877. The results indicate that there are significant differences among the three categories of teaching presence at the significance level of 0.001.
Table 5 presents the descriptive statistics of the social presence. The values of open communication and relationship cohesion are both small, and the former is smaller: Wilks’ Lambda = 0.699, F(1, 12) = 5.164,
p = 0.042 < 0.05, and η
2 = 0.301. This implies that there are significant differences between the two categories of social presence at the significance level of 0.05.
Table 6 presents the descriptive statistics of the emotional presence. Activity emotion and outcome emotion are more common, while directed emotion is relatively rare: Wilks’ Lambda = 0.353, F(2, 11) = 10.059,
p = 0.003 < 0.05, and η
2 = 0.647. This indicates that the three categories of emotional presence are significantly different at the significance level of 0.05.
5. Discussion
Cognitive presence has the second highest value of the four presences. The most common category is resolution, followed by exploration and integration, and the least common is triggering events. Cognitive presence is mainly achieved through the steps of triggering events, exploration, integration, and resolution. A triggering event is generally caused by teachers, and learners explore and integrate in order to further reach the stage of resolution. The goal of teaching activities is to assist students in solving problems or completing complex tasks. Because learners may be more concerned with the outcome than the process, resolution is more common, which implies the existence of the other categories. The triggering event is the smallest, and it has an influence on the level of cognitive development. Teachers should carefully design triggering events to promote learners’ cognitive development. It was found that practice-related triggering events were more likely to cause cognition to reach the resolution stage [
23]. Triggering events can thus be combined with problems, practices, applications, operations, and so on.
The value of the teaching presence is relatively higher, the category of design and organization is significantly more common, and facilitation is significantly less common than the others. By deeply analyzing the text feedback, it was found that learners enjoy the flipped classroom, case teaching, learning by playing, cooperative learning, project-based learning, and other new design and organization forms. Rodgers believes that trust is the foundation of facilitation [
24]. Teachers can promote dialogue and discussion from the “relationship” level of psychology by encouraging every learner to integrate into the learning space as a facilitator and by bringing out the students’ teaching presence.
The negative value and the suggestion value of the teaching dimension are much larger than the other three dimensions, which indicates that there is much room for improvement. By deeply analyzing the text feedback, it was found that the main problem focuses on the requirement of practical teaching. Some words, such as “technology, software, production, video, download, media, courseware, tools, editing, operation, PPT, electronic whiteboard”, which are related to physical technologies, are repeatedly mentioned, and they refer to the learner’s instrumental rationality (i.e., a focus on physical technology, function-oriented, and treating technology as a substantive means and tool). Contemporary university students naturally adapt to the informatization learning style, and they are willing to use information technology. That they prefer physical technology is understandable. On the one hand, in order to meet the needs of the digital generation, teachers can incorporate physical technology into the teaching content and learning activities. On the other hand, it is also necessary to guide learners’ cognition to a higher level, and to help learners to understand that technology not only supports teaching as a physical technology, but it also solves how to conduct it, it applies technology to redesign teaching activities, and it innovates teaching modes [
25]. Teachers can use activity strategies that combine practical experience with cooperation and mutual assistance to fully mobilize the enthusiasm and initiative of learners, create a friendly and trusting psychological environment in the classroom, and promote the development of the teaching presence.
Social presence is less common than the other three presences. We discovered some statements while reading the text feedback, such as “I hope that the theoretical part will not only be taught by the teacher, but we will be able to participate in it”, and “More interaction with students and a more active atmosphere are expected”. The lack of social presence could be attributed to a lack of interaction between teachers and students. Garrison believes that the social presence in the inquiry-learning community should be enhanced [
7]. The essence of teaching is the embodiment of teacher–student and student–student interactions, which are a two-way street between teachers and students. For example, the content or form of interaction is relatively dull, and some teachers do not know how to interact better. On the one hand, from the level of “tactics”, teachers can increase interactive links and innovative interaction forms. Teachers can improve interaction before, during, and after class by using interactive tools, such as barrage, the UMU interactive platform, the Rain Classroom platform, and Blue Ink Cloud Class. Teachers, on the other hand, must recognize that, in this age of information overload, learning is no longer a one-way transfer of knowledge. Through interaction and communication, it is necessary to “see” students and “see” each other, as well as to gain insight into inner needs and external emotions, and then respond accordingly. Teachers use incentive measures to increase learners’ enthusiasm and willingness, guide the conversation, and ensure effective communication. Then, the students can feel the value and fun of mutual peer assistance. Teachers create conditions for inquiry-based high-quality interaction and higher-order cognitive development by creating social presence.
Positive emotional experiences can stimulate the learners’ interest in learning, encourage them to participate in learning with responsibility and professionalism, and encourage them to investigate the power and value of self-growth [
26]. The results show that emotional presence is the most common, with the largest value, which is much larger than the other presences; the negative value of the emotional dimension is the second largest value; and the suggestion value is the smallest. Students have the courage to express their emotions in text feedback, and they can state whether their emotions are positive or negative. Some words that appear frequently in the text feedback are “feel, interesting, interested, thankful, vivid, useful, beneficial, and like”. Emotional presence is likely the most sensitive indicator of all the presences, and it may act as an “amplifier” that potentially influences other presences. The values of activity emotion and outcome emotion are higher among the three categories of emotional presence, while the value of directed emotion is lower. The findings show that learners have more positive emotional responses to ongoing activities and to the outcomes of problem-solving successes or failures, and they also have positive emotional responses to interpersonal emotions, but these are relatively fewer. Learners in Chinese culture are reserved, and they rarely express interpersonal emotion. When they participate in the learning, they pay more attention to the subject matter, as well as to their own emotional experiences. Teaching is more than just the processing and transmission of cognitive information; it is also the experience and communication of emotional information, in which emotion and reason coexist and rely on one another.
6. Conclusions
On the basis of the CoI framework, this study coded 476 pieces of text feedback, which included 1695 short sentences, from 13 classes on Modern Educational Technology. The findings indicate that the CoI framework is also appropriate for use in college or university classrooms. All presences exist, and the data support the existence of emotional presence as well.
There are significant differences between the dimensions and between the categories of each presence. The value of cognitive presence is large, while the negative value and the suggestion value of the cognitive dimension are relatively smaller. Resolution is more common than others, exploration and integration are also more common, while triggering events are the least common. Teaching presence receives the most attention. Moreover, the negative and suggestion values of the teaching dimension are much larger than in the other dimensions. The category of design and organization is significantly more common, and the facilitation of discourse is much more uncommon than in the other categories. The positive and negative values of the social dimension are the smallest among the four dimensions, and the positive value is much smaller than the others. Emotional presence is the most common, with the largest value, which is much larger than the other presences; the negative value of the emotional dimension is the second largest; and the suggestion value is the smallest. All these characteristics of presences help teachers to optimize course design and implementation for enhancing presence, improving learning experiences, and promoting high-quality learning.
The work in this paper adopts the text-analysis method to manually code and assign values. It is difficult, and it inevitably brings about subjective bias. In order to overcome this shortcoming, a more rigorous coding mechanism will be adopted to further reduce the subjective error of manual coding in our future work.
The individual factors of learners play an important role in cognitive participation and knowledge acquisition. Some scholars are committed to studying learners’ self-efficacy, self-regulated learning, and metacognition, and they put forward the dimension of the learning presence [
8,
27]. This dimension was not considered in this paper. It can be expected that the results of this research can be expanded and enriched in different learning environments by combining the cognitive, teaching, social, emotional, and learning dimensions.
Data-based knowledge discovery and viewing and improving teaching from the perspective of students’ feedback have brought methodological and epistemological improvements to researchers. However, when facing the complex system of classroom teaching, some researchers are still reluctant to go too far and produce significant effects. In our future work, we can further study how to use the CoI framework to improve teaching in a positive direction.