The flipped classroom has become a popular way of structuring a classroom in the digital age (Han, 2022, p. 2). Because of access to the internet, many teachers have started to structure their classes so that students learn new concepts and ideas before class (Lo and Hew, 2017). Class time then becomes a period of active learning, where students can get real-time feedback and collaborate with each other (Han, 2022, p. 2). 

Fig. 1: Flipped Classroom instructional model. (n.d.). https://www.polyflip.eu/Flipped-classroom-in-theory/General-framework-terminology-and-pedagogy

Part of many flipped classroms is self-paced or self-regulated learning (I will be using the terms interchangeably based on what is used in the study), which often takes the form of project-based learning, inquiry-based learning, and/or collaborative learning (Bautista, 2015, p. 185). Self-paced learning allows students to take responsibility for their learning and to be able to be reflective on their experience (Lee & Chang, 2024). 

This topic is important as many teachers are moving toward student-centered learning and a more universal design for learning (UDL) approach. At my school, we are really trying to emphasize UDL principles and make learning accessible to all students. What I have noticed in my own classroom, as well, is that some students might get content while I'm lecturing really quickly, while others need more time. As I've noticed this more and more over the past year, I have contemplated moving to a more self-paced, largely flipped classroom. We also live in a world of diverse learners (we always have; it is just more acknowledged now) and thus are in need of differentiated instruction. One goal of research here then was to help me understand both the advantages and disadvantages of the flipped classroom and self-pacing. . The main dimensions of the project are the structure of self-pacing and flipping, the social learning nature of these pedagogies, and the overall effectiveness of these strategies. 

There are a few relevant theories for self-pacing and a flipped classroom. Most of them fall into the category of social learning theories, where learning is based on social interactions. In addition to social learning, the other main theories are what I am calling reflexive learning for this paper - they revolve heavily around an ongoing cycle of learning through a few processes but with a heavy emphasis on reflection being crucial for learning. 

Social Learning/Constructivists 

The biggest theory the flipped classroom is based on is constructivist theory. This theory posits that "learning is an active process of constructing meaning and knowledge through social interactions" (Larsari and Abouabdelkader, 2024, p. 15). Additionally, they contend that students should actively participate in their learning through interactions and experiences (Larsari and Abouabdelkader, 2024, p. 15). In the flipped classroom, student interaction is one of the most important components (Lai & Hwang, 2016, p. 127). 

While Piaget is an important constructivist, the most important for the flipped classroom and self-pacing is Vygotsky, specifically his Social Development Theory (SDT) (Bautista, 2015, p. 186). For Vygotsky, social interactions were crucial for learning because he believed they came before true cognitive development (Bautista, 2015, p. 186). Vygotsky posited that there are people who know more than the learner, called More Knowledge Other(s) (MKO) who model the behavior for the learner (Bautista, 2015, p. 187). Often, these are teachers, but it could be a learner’s peer or even technology, which today would probably take the form of digital learning technologies (Bautista, 2015, p. 187). Finally, and perhaps the most important idea from SDT for this paper, is the Zone of Proximal Development (Bautista, 2015, p. 187).

Fig. 2: UAF CTL Staff. (2017). Lev Vygotsky’s concept of the Zone of Proximal Development. https://ctl.uaf.edu/2017/02/20/encourage-your-students-to-step-into-the-zone/

The first zone for Vygotsky is the learning that students can do on their own, or the ability to problem-solve on one’s own (Bautista, 2015, p. 187). The zone of proximal development (ZPD) is the area where a learner needs guidance from a teacher and/or peers (Bautista, 2015, p. 187). Outside of this zone, Vygotsky believed no learning could occur (Bautista, 2015, p. 187).

Video 1: Sprouts. (2020). Vygotsky's Theory of Cognitive Development in Social Relationships [Video]. YouTube. https://www.youtube.com/watch?v=8I2hrSRbmHE&vl=en

This video shows the Zone of Proximal Development as well as describing Vygotsky's theory well. I liked the example the video used of twins as it is a nice contrast. I also like the visual of a young swimmer in the shallow area for what a learner could do by themselves, with a parent for slightly deeper waters (the zone of proximal development), and then the third zone was clearly for more advanced/adult swimmers (beyond the learner's current capabilities). 

Additionally, Bandura’s Social Learning Theory (SLT) is based on the idea that learning is when humans learn behaviors by observing others and then performing those actions (Bautista, 2015, p.186). Additionally, the environment students learn in also impacts learning, as much of learning is context-based (Larsari and Abouabdelkader, 2024, p. 16). Bandura also broke up learners into observational learners and modelers (Larsari and Abouabdelkader, 2024, p. 16).

Fig 3: Kurt, S. (2019). Learning through Observation. https://educationaltechnology.net/social-learning-theory-albert-bandura/

As seen above, these factors all influence each other (Larsari and Abouabdelkader, 2024, p. 16). One problem with Bandura's theory for the flipped classroom is that he said observational learning can occur without the learner really changing behavior, aka learners can imitate but might not always learn that behavior (Larsari and Abouabdelkader, 2024, p. 16).

The last of the social learning theories is the Situated Learning Theory by Lave (Bautista, 2015, p. 187). Just like Bandura and Vygotsky, Lave said that learning takes place within social interactions through various activities, especially inquiries (Bautista, 2015, p. 187). Learners are situated in an environment that fosters practice and interaction with peers, and through each inquiry, students/learners become more engaged in learning (Bautista, 2015, p. 187).

The flipped classroom and self-paced learning build on these social learning theories in many ways. Firstly, the flipped classroom focuses on learning by doing, but specifically, doing this among other students (Cheng et al., 2018, p. 796). Based on constructivist approaches, the "flipped classroom instructional approach discusses learning as active engagement with the content, instructor, and other students" (Cheng et al, 2018, p. 796). For Vygotsky specifically, students would experience the ZPD in the classroom where they could have teachers and peers as the MKO if they needed guidance. Additionally, self-pacing would help students gauge where they were at and if they were ready to move out of the can-do for themselves zone with specific content.

Reflexive Learning Theories

The flipped classroom also builds on Zimmerman’s self-regulated learning (SRL) theory. SRL theory focuses on three different types of learners – metacognitive, motivated, and strategic (Brenner, 2022). Metacognitive learners know what their strengths and weaknesses are for learning (Brenner, 2022). Motivated learners believe that effort can help them succeed, even with challenging tasks (Brenner, 2022). Strategic learners are those who know a lot of learning strategies and can use them in different situations (Brenner, 2022). Zimmerman also sees SRL as a cycle of planning (or forethought), performance, and evaluation, especially self-evaluation (Brenner, 2022; Bautista, 2015, p. 185). Part of self-regulated learning also involves learning from social interactions, tying back to Bandura and Vygotsky (Brenner, 2022).

Fig. 4: De Laet, T., Dujardin, R., Langie, G., Sijmkens, E., Van de Broeck, L., De Cock, M., De Keyzer, J., Sermeus, J., & Van Soom, C. (2024).Zimmerman’s Model of Self-Regulated Learning: https://set.kuleuven.be/LESEC/vision2/24/self-regulated-learning

A large part of SRL theory is the cyclical nature of the process of learning, which can be seen above. (Brenner, 2022).

In addition to Zimmerman, Kolb's Experiential Learning Theory (ELT) has a cyclical nature (Larsari and Abouabdelkader, 2024, p. 16). Similar to Zimmerman, learning is a cycle of thinking (planning), acting (performance), and reflecting (self-reflection) (Larsari and Abouabdelkader, 2024, p. 16). However, these are broken up into four stages instead of just three like Zimmerman (Larsari and Abouabdelkader, 2024, p. 16). Kolb uses "concrete experience, reflective observation, abstract conceptualization, and active experimentation" (Larsari and Abouabdelkader, 2024, p. 16). This reflection piece is crucial to Kolb and ELT as it creates knowledge (instead of knowledge only coming from experience) (Larsari and Abouabdelkader, 2024, p. 16). In the image below, you can see the cyclical and reflective nature of Kolb's theory. 

Fig. 5: McLeod, S. (2024). The Experiential Learning Cycle. https://www.simplypsychology.org/learning-kolb.html

Other Relevant Theories

There are a few other theories that are also relevant to the flipped classroom and self-pacing. The first is Dewey's Active Learning Theory (Ettien & Touré, 2023, p. 54). The key for Dewey was engagement in the learning process, which involves hands-on learning and real-world interactions (Ettien & Touré, 2023, p. 54). Dewey, and later Bolduc, also emphasized the importance of learning by doing and that there is a reflective nature to it as well (Ettien & Touré, 2023, p. 55). 

Another is Self-Determination Theory (SDT). SDT believes that people are naturally curious and interested in learning (Niemiec & Ryan, 2009, p. 133). In education, it focuses more on intrinsic motivation rather than teachers putting pressures and rewards/punishments on students (Niemiec & Ryan, 2009, p. 134-135). A person's intrinsic motivation is fueled by their desire for autonomy, which in a classroom, will then in turn increase a student's competence (Niemiec & Ryan, 2009, p. 135). As seen below, the external factors would be the grades and tests that teachers have pressure to put on students, but SDT says that students will learn better when they can foster those feelings on the left (the intrinsic motivation) (Niemiec, Ryan, 2009, p. 134-135). 

Fig. 6: Nickerson, C. (2023). Intrinsic and extrinsic motivation. https://www.simplypsychology.org/differences-between-extrinsic-and-intrinsic-motivation.html

Critics: Cognitivism instead of Constructivism

Some of the biggest critiques of social learning theories are that they just have not been tested in reality a lot. Many of the studies and articles I found talk about no real sense of effectiveness them and the need for greater research on this topic (Brauer & Tittle, 2012, p. 157; Brenner, 2022; Bautista, 2015, p. 184). Some specific critiques along the lines of mixed effectiveness include a lack of theory in the studies done; it is not clear what in-class activities actually lead to critical thinking since there are a variety of activities; some more abstract concepts cannot be easily grasped outside of in-class instruction; and the mixed impact on students, especially in diverse learning populations (Song et al., 2017, p. 180). Additionally, some researchers have stated how this theory is predicated on the idea that students will willingly and effectively work together, which is not always the case (Palaigeorgiou & Papadopoulou, 2019, p. 820). There just seem to be limited practical applications in studies, or at least many of the studies feel like more research needs to be done (Palaigeorgiou & Papadopoulou, 2019, p. 821). However, there are people who do have more specific critiques of constructivism. They tend to be much more focused on brain architecture and how that affects learning (Kirschner et al., 2006, p. 77).

One of the critiques of constructivism is that it does not match the nature of working memory (Kirschner et al., 2006, p. 77). Working memory, aka short-term memory, can only retain a few elements within a time period (Kirschner et al., 2006, p. 77). Some critics say that because of the nature of working memory, constructivist learning, which tends to be inquiry or problem-based, creates a cognitive overload for learners (Kirschner et al., 2006, p. 77). Part of the problem is that problem-based learning is often very taxing on working memory (Kirschner et al., 2006, p. 77). Additionally, they argue that the ability to learn information for long-term memory is hampered because the available working memory is being used to do the searching for knowledge (Kirschner et al., 2006, p. 77). Some of the critics do not disagree about the social nature of learning that constructivists put forth, but rather their critique is of the way that they say teachers should take and use that information (Kirschner et al., 2006, p. 78) Specifically Kirschner et al. (2006) are arguing that minimally guided instruction does not actually help students learn (p.78). This leads psychologists, curriculum designers, teachers, and more to assume that the skills teachers learn when learning how to teach the material and to research the material are the same ones the students will get through this inquiry; however, the critics argue this is not what actually happens as the two skills are separate from each other (Kirschner et al., 2006, p. 78-79).

Some of these critics, including Kirschner and Sweller, one of the other authors on that paper, have supported cognitive load theory (CLT) as an alternative to constructivist theories (Kirschner et al., 2018, 213). CLT says “cognitive load refers to the total working memory resources required to carry out a learning task” (Kirschner et al., 2018, 214). Processing of new information is what taxes the cognitive load of working memory, which in turn will affect how well a student retrains that information (Kirschner et al., 2018, 214). While there are limits on what students can learn at a time, generally, there are no limits on how much information can be stored in long-term memory (Centre for Education Statistics and Evaluation, 2017, p. 2). Because of this, CLT is more supportive of explicit instruction rather than guided instruction (Centre for Education Statistics and Evaluation, 2017, p. 2).

Fig. 7: Main, P. (2017). Cognitive Load Model. https://www.structural-learning.com/post/cognitive-load-theory-a-teachers-guide

Cognitive load is divided into two types (sometimes three depending on the source) – intrinsic and extraneous, the third being germane (Kirschner et al., 2018, 214). Intrinsic cognitive load deals with the complexity of the information being presented, which in itself is related to the interrelatedness of elements (Kirschner et al., 2018, 214). Extraneous cognitive load is the elements unrelated to the task that put a strain on working memory, like how information is presented in the task (Kirschner et al., 2018, 214). (Topolovčan, 2023, p. 194). Extraneous cognitive load hampers learning because it does not lead students to develop these neural pathways, often called schema in CLT (Centre for Education Statistics and Evaluation, 2017, p. 3). Germane is sometimes considered part of intrinsic because it relates to how instruction is designed, with well-designed instruction facilitating learning (Centre for Education Statistics and Evaluation, 2017, p. 3). 

Although many believe constructivist theories and cognitive load theories are opposed to each other, not all researchers agree. Skulmowski and Xu (2022) discuss how the two can actually be aligned. Additionally, the lack of motivational theory, which I mentioned above with self-determination theory and which goes well with constructivist theories, is often thought to be one of the problems with CLT (Skulmowski & Xu, 2022, p. 177). They suggest a constructive alignment for CLT design, focused on three main ideas: gearing instruction toward specific objectives, creating cognitively engaging activities, and using performance-based assessments (Skulmowski & Xu, 2022, p. 187). The combined theory they put forth is cognitive load alignment, which also allows for more differentiation among students (thereby solving one of the problems I will discuss in section 4) (Skulmowski & Xu, 2022, p. 188). Below is an example given in their study wherein cognitive load must align with the germane processing load that a task will require (below in the cognitive processing and learner characteristics) (Skulmowski & Xu, 2022, p. 188). Then, teachers must create assessments that are appropriate to the cognitive processing style so that there is an accurate assessment of learning (Skulmowski & Xu, 2022, p. 188).

Fig. 8: (Skulmowski & Xu, 2022, p. 188)

Critiques of Self-Regulated Learning 

In addition to critiques of social learning, there are also some specific critiques of self-regulated learning. First, some students do not have the skills to work independently right away (Palaigeorgiou & Papadopoulou, 2019, p. 820). These students need more scaffolding support from the teacher before they can self-regulate at all. Another big critique of self-regulated learning is that because instructors are moving around the classroom, they can not always monitor the various discussions happening between groups (Palaigeorgiou & Papadopoulou, 2019, p. 820). 

For a flipped classroom and self-paced learning, these critiques highlight the difficulties some learners might have with external stimuli or trying to find the information themselves. Sometimes, with problem-based or inquiry-learning, which is central to both social learning/constructivism and the flipped classroom, it is hard for some students to know where to start. I see that in my class sometimes with something that requires a little more thought or searching - some students know how to search and tweak their searches as need be, but others need my guidance and more step-by-step directions. I also see this in my students with the motivation needed for self-pacing - I have some who want to learn just to learn, but others only want to do it for a grade (and thus are less motivated to go beyond the surface).

The self-paced classroom is one where students are allowed to proceed at their own pace (Bautista, 2015, p. 184). Part of the reasoning is that because of learning differences - different ways of learning, different rates of learning, etc. - students can do better if they are not all put at the same pace (Bautista, 2015, p. 184). Additionally, this allows for more interactive thinking and learning during classtime (Lo, 2018, p. 793). A key component of self-pacing and the flipped classroom are videos made outside the classroom, usually hosted on some sort of digital learning system. This supports a flipped classroom as it allows teachers to be more curators of digital content (Cope & Kalantzis, 2021, p. 39). Teachers can also create activities that are more interactive and support the content, rather than only just delivering content in a traditional format (Cope & Kalantazis, 2021, p. 40). This focus on student agency can be seen in the infographic below from Cope and Kalantzis (2021, p. 42).

Fig. 9: (Cope and Kalantzis (2021, p. 42).

One study on self-paced learning was from Bautista (2015), who surveyed high school chemistry sstudents in a self-paced classroom. The learning here was not 100% self-paced - there were lectures and laboratory activities as well as homework (Bautista, 2015, p. 185). However, students were given the learning modules ahead of time so they knew what to expect (Bautista, 2015, p. 185). Bautista (2015) was hoping to analyze if there was a significant difference in scores between students in a self-paced classroom and students in a traditional classroom (p. 185). Bautista (2015) conducted this study at Quirino State College in the Phillippines, in a high school chemistry classroom. He used a questionnaire to survey students' performance on a test and included several factors that assessed self-paced learning, including students' motivation (Bautista, 2015, p. 189). The results showed that the students in the self-paced classroom performed better on the test, and that students with more motiviation in this class did even better (Bautista, 2015, p. 189-191). While Bautista (2015) concluded that while yes, there were better scores in the self-paced classroom, these effects were only significant for high ability students; for students who had less ability, the impact was much smaller (p. 191). 

While Bautista (2015) focused more on self-paced learning, a study by Lai and Hwang (2016) looked more at self-regulating in the flipped classroom to try to demonstrate the impact that the skills of self-regulation can have students' performance in the flipped classroom. One of their arguments is that many students lack the fundamental skills needed in order to do the self-regulating outside of the classroom that a flipped classroom requires (Lai & Hwang, 2016, p. 127). They argue that flipped classrooms have enormous benefits like promoting active learning, emphasizing student-centered learning, and utilizing class time more effectively, but that these advantages only happen if students know how to self-regulate (Lai & Hwang, 2016, p. 128). This goes back to self-determination theory (SDT) as only those students who have the motivation to learn and the desire to have autonomy (Niemiec & Ryan, 2009, p. 135).

In a meta-analysis of flipped classrooms, Lo and Hew (2017) analyzed fifteen case studies of flipped classrooms in K-12 settings, which is different to many studies that focus more on university settings. They do make comparisons between the two, highlighting the similarity in benefits and drawbacks between both (Lo & Hew, 2017). The biggest difference between the two was in the operational challenges, which will be addressed in the next section (Lo & Hew, 2017). Based on their critical review of these studies, the authors put forth their own model of a flipped classroom, represented in the image below. 

Fig. 10: (Lo & Hew, 2017). (fig. 3. label is from their paper)

Their model allows for more flexibility than other flipped classrooms, as it allows for teachers to cover more complex topics in class if needed or alter plans based on students' needs (Lo & Hew, 2017). Additionally, they provide guidelines for effectively implementing the flipped classroom, based on their review of other studies (Lo & Hew, 2017). These guidelines are:

1. Teachers communicate clearly with students before switching to a flipped classroom to ensure they understand the arrangement (Lo & Hew, 2017)
2. Scaffolding and modeling how to learn in the flipped classroom (Lo & Hew, 2017)
3. Using the Cognitive Theory of Multimedia Learning (CTML) to design instructional videos (Lo & Hew, 2017)
4. Maintaining the same level of workload as before flipping the classroom (Lo & Hew, 2017)
5. Creating some sort of online forum or other means where students can pose questions since they cannot ask them during class like a traditional class (Lo & Hew, 2017)
6. Preparing teachers adequately and giving them feedback on their practices (Lo & Hew, 2017)
7. Teachers should slowly build up resources and videos they create to avoid being overwhelmed by the amount of work required (Lo & Hew, 2017)
8. Ensuring equity for technology access to support students who might have more limited resources (Lo & Hew, 2017)
9. Using gamification to encourage student learning (Lo & Hew, 2017)
10. Investment by school in IT and other resources that will support flipped classrooms (Lo & Hew, 2017)

Lo & Hew (2017) conclude that flipped classrooms do seem to encourage active learning. However, they do also concede that there were mixed feelings from students surveyed about the flipped classroom (Lo & Hew, 2017). They also note the limitation of its application in the K-12 setting as there were few studies for them to review (most being in the higher education setting) (Lo & Hew, 2017). 

Lo (2018) continues studying flipped classrooms by looking at digital learning tools in the flipped classroom. Lo (2018) builds on the six pillars of educational technology from J.M. Spector (p. 798). These pillars are communication, interaction, environment, culture, instruction, and learning (Lo, 2018, p. 798). The table below shows the challenges that Lo (2018) identified with instructional videos and potential solutions to these problems based on CTML (p. 800). 

Table 1: (Lo, 2018, p. 800). 

Lo (2018) also recommends computer-generated feedback and technology to track student completion (p. 801-802). The other most relevant suggestion is creating a culture of learner-centered instruction (Lo, 2018, p. 803). This goes back to Zimmerman's SRL as students need to be encouraged to go through this process of inquiry and class activity with a willingness to fail and grow from failures (Lo, 2018, p. 803-804). 

One theme that appeared in many articles was the overall effectiveness of the flipped classroom. To analyze this, I looked at a meta-analysis of the flipped classroom's impact on learning outcomes by Cheng, Ritzhaupt, and Antonenko (2018). Their meta-analysis looked at 55 case studies (Cheng et al., 2018, p. 805). Of those 55, 14 discussed the negative effects of the flipped classroom and 41 were in favor of implementing a flipped classroom (Cheng et al., 2018, p. 805). However, as they acknowledge, not all of these positive studies had statistically significant results - only 14 of them had significant statistics in favor of flipped classrooms (Cheng et al., 2018, p. 805). Cheng, Ritzhaupt, and Antonenko (2018) also found differences in effectiveness across disciplines - arts and humanities had the greatest positive effect, while engineering had a negative effect, although not significantly so (p. 814). Interestingly, while K-12 and undergraduate showed positive effects from a flipped classroom, studies on graduate students showed a negative effect, although again, not necessarily statistically significant (Cheng et al., 2018, p. 817).

One of the purported effects of the flipped classroom and self-pacing is the benefit for diverse learners. There is increased student-teacher interaction in a flipped classroom, so students who need more help can get it (Altemueller & Lindquist, 2017, p. 343). Additionally, teachers can create different levels of activities that meet the varied needs of their students (Altemueller & Lindquist, 2017, p. 343). One study that looks at this is Goedhart et al. (2019). They looked at students in a graduate course in Amsterdam where half the students were in a flipped classroom and the other half were in a traditional classroom (Goedhart et al., 2019, p. 297). They used questionnaires to gauge students' performance through the course as well as their opinions about features of the classroom (Goedhart et al., 2019, p. 302). They found that students in the flipped classroom were more prepared for class than those in the traditional setup and that this pre-class preparation allowed for better self-pacing (Goedhart et al., 2019, p. 306). They did see some students did not feel challenged enough in group activities and recommended that activities be differentiated based on cognitive load and prior knowledge (Goedhart et al., 2019, p. 306). This study helps show the benefits that self-pacing and a flipped classroom can have on students, especially for diverse learners.

Another study that focused on self-regulated learning is from Palaigeorgiou and Papadopoulou (2019) who studied the role of educational videos and tablets in increasing students' self-regulation. Additionally, they argue that interactive educational videos can overcome cognitive overload (Palaigeorgiou & Papadopoulou, 2019, p. 807). Their study looked at elementary students in Greece who were learning about heat transfer in physics (Palaigeorgiou & Papadopoulou, 2019, p. 809). Students were paired up to share a tablet and had a worksheet to fill out as they watched (Palaigeorgiou & Papadopoulou, 2019, p. 809). The teacher was there to provide support whenever students needed it, but was in a more passive role, while the students were in a more active role (Palaigeorgiou & Papadopoulou, 2019, p. 809). The interactive videos they used had the same interactive elements - pointers/text labels, embedded questions (both predictive and inductive), internal and external video links, reflective pauses, and the ability to jump to different steps (Palaigeorgiou & Papadopoulou, 2019, p. 811). The control group for this study was classrooms of the same age that were taught in a more traditional method, with the teacher instructing directly (Palaigeorgiou & Papadopoulou, 2019, p. 811). The study showed that students who were in the self-regulated classroom with interactive videos performed better on the post-test than the students who were in a traditional classroom (Palaigeorgiou & Papadopoulou, 2019, p. 815-820). Many students surveyed also appreciated that they felt they could work through questions together, whereas in traditional lecture settings, they cannot discuss (Palaigeorgiou & Papadopoulou, 2019, p. 816-817).

Another meta-analysis on the flipped classroom came two years after Cheng, Ritzhaupt, and Antonenko (2018) from Strelan, Osborn, and Palmer (2020). They looked at 174 studies across various education levels and disciplines (Strelan et al., 2020). Cognizant of the many factors that influence learning, the authors in this meta-analysis purposely took into account (Strelan et al., 2020). They found that, overall, there was a positive effect from the flipped classroom (Strelan et al., 2020). Compared to Cheng et al. (2018), there was more of a moderate effect found in this meta-analysis (Strelan et al., 2020). Interestingly, for disciplines, humanities again saw the biggest effect, but engineering saw a more moderate effect (health sciences had the lowest effect from flipped classrooms here) (Strelan et al., 2020). Strelan et al. (2020) also addressed the aspect of self-pacing in pre-class activities, finding that it could actually reduce cognitive load if done properly, which provides a counter to the arguments of CLT. Their overall conclusion was that there is a moderate positive effect from a flipped classroom, but that it varies across disciplines, perhaps because of the prior scaffolding some classes naturally do (because of the subject matter) (Strelan et al., 2020). 

Uus, Seitlinger, and Ley (2020) studied self-directed learning (SDL) in a middle school classroom in Estonia. The authors start the article by acknowledging Cognitive Load Theory (CLT) and the impact it can have on students' learning, as well as the differences in students' self-control, especially with attention regulation (Uus, Seitlinger, 2020). They tested students using a dinosaur task where students had to identify features of dinosaurs and were given feedback after each attempt (Uus, Seitlinger, 2020). They found that students with higher working memory capacities were able to resist distractions in test test more easily (Uus, Seitlinger, 2020). Although they concluded this, they did acknowledge that some of these problems could be overcome by teaching students metacognitive skills as well as scaffolding some of the skills needed for self-directed learning (Uus, Seitlinger, 2020).

Deng, Feng, and Shen (2023) also studied improving flipped classrooms with the use of videos that featured questions. Their study was conducted among sophomores at a Chinese university, who watched videos that had questions embedded throughout, not just at the end (Deng et al., 2023, p. 12681-12682). However, the authors posited that there needed to be an extra step - teachers should review the questions at the start of class to ensure student understanding (Deng et al., 2023, p. 12684). The results of their study showed that this intervention (the review of questions at the start of class) improved student learning and led to more regular learning (Deng et al., 2023, p. 12695).

One of the most recent meta-analyses I looked at was from Lee and Chang (2024), who analyzed articles that talked about self-directed learning (SDL) in digital learning environments. They looked at 17 studies that used SDL in the test group and traditional learning methods in the control groups (Lee & Chang, 2024). Their meta-analysis showed that SDL had a positive impact on knowledge acquisition, especially when there was increased peer interaction (Lee & Chang, 2024). Compared to traditional teaching, students consistently had better learning performances; however, many students across the studies acknowledged that a combination of self-paced learning and traditional instruction, often called blended learning, was the best combination (Lee & Chang, 2024).

The other meta-analysis from 2024 I looked at was from Li et al. Their goal was to analyze the effectiveness of flipped classrooms and more specifically, to find what factors influenced the effectiveness (Li et al., 2024). They reference the meta-analysis from Lo and Hew (2017) that I researched above, as well as another from Bond (2020) as being two of the few reviews on K-12 flipped classrooms (Li et al, 2024). However, they mention that both of these were qualitative and did not quantitatively measure effectiveness (Li et al., 2024). In their meta-analysis of 129 studies, they found that the effect size was .59, excluding outliers (Li et al., 2024). This means that for a standard deviation of 10 on an assessment, students in a flipped classroom scored 5.6 points higher than students in a traditional classroom (Li et al., 2024). They also found that flipped classrooms can benefit students in low- and middle-income areas and that there was no statistically significant difference between disciplines (Li et al., 2024). An interesting thing though is they were trying to figure out what factors influence effectiveness, but their data showed that there was nothing statistically significantly different in terms of pre-class vs. in-class activities, design, etc. (Li et al., 2024). This suggests perhaps more focused research is needed, but also maybe that the underlying principle of the flipped classroom is more effective as a baseline for students. 

Overall, the literature seems to point to the general effectiveness of a flipped classroom and self-pacing. However, the statistical significance and what leads to the effectiveness of this seem varied. 

Overall, the flipped classroom seems to have mixed results. As it relates to social learning like Vygotsky, it succeeds immensely. The flipped classroom is designed for students to interact with each other and to foster problem-based or inquiry-based learning. Even the self-regulated learning piece emphasizes the social nature of learning. In many studies (Bautista, 2015; Lai & Hwang, 2016; Tullis & Benjamin, 2011), there were clear benefits to self-pacing and the flipped classroom. For instance, Tullis and Benjamin (2011) found that a self-pacing group of university students from the University of Illinois outperformed those who had fixed-rate learning modules (p. 109).  Additionally, there seems to be a potential for great benefits for diverse learners (potential because it would depend on teachers creating lessons with tiered activities and truly differentiating) (Goedhart et al., 2019). 

For many of the studies that found the flipped classroom to be successful, there was still a need for teachers to scaffold skills for students to be successful in both self-pacing and a flipped classroom. This ties back to Vygotsky's Zone of Proximal Development, as seen in the image below. If students do not gain those skills through scaffolding, they will be in the blue circle above what they cannot do. However, when teachers scaffold, this can help ensure a successful self-paced, flipped classroom.

Fig. 11: Sakamoto, S. (2021). Scaffolding Strategies. https://connect.ajet.net/2021/03/08/scaffolding-strategies-for-teaching-english/

Despite some success, there are some limitations to self-regulated learning and the flipped classroom. One of the arguments I saw repeatedly was that high-achieving students did really well in a self-regulated environment, but that many low-achieving students did not significantly improve (Ahmed & Indurkhya, 2020). Related to this was students' motivation level, as some students lack the motivation to self-pace (Jackson, 2018, p. 35-36). Proponents of self-pacing, though, say that teachers can help students develop a mindset about learning that would overcome this obstacle (Jackson, 2018, p. 36). 

Some of the biggest challenges of a flipped classroom were detailed in Lo and Hew (2017). They broke the challenges up into three larger categories - student-related, faculty, and operational challenges (Lo & Hew, 2017). There were not as many faculty challenges, which mostly focused on limited materials available and the front-end time of starting the flipped classroom process (Lo & Hew, 2017). 

Table 2: Lo and Hew (2017) (table numbers above are from their paper)

The two tables above from Lo and Hew (2017) show the big challenges they identified in their review of flipped classroom studies. Some of these challenges though have been addressed by other theories, such as the cognitive theory of multimedia (Mayer, 2014, p. 43). 

There were also mixed feelings and results about other elements, such as equity in the classroom. Some authors argued that the flipped classroom allowed for greater equity as it would allow for greater choice and a variety of learning opportunities (Ahmed & Indurkhya, 2020). However, others pointed to the equity issues that could arise because of the technological component of the flipped classroom that takes place at home. These issues can be seen in the second table above from Lo and Hew (2017), as students' IT resources are the number one challenge from the literature they reviewed. 

Additionally, to the challenges, I think one of the ideals of the flipped classroom and that self-pacing is predicated on is that students will actually do the self-learning outside of the classroom (not just that it is too complex). In a review of the literature on the flipped classroom, Akçayir and Akçayir (2018) found that the biggest problem reported by teachers was that students were not completing the self-learning at home (p. 341). Since the whole point of a flipped classroom is to do more interactive and engaging activities in the classroom, if the learning does not happen before that, it defeats the purpose if there is no learning before that (Akçayir & Akçayir, 2018, p. 341). Perhaps, that goes back to some of the other results discovered - that this method tends to work best for those students who are higher achieving and perhaps more motivated to learn intrinsically. 

Akçayir and Akçayir (2018) also posed an interesting dilemma for the future of the flipped classroom - most of the studies conducted just focus on one classroom that is flipped (p. 343). Having multiple flipped classrooms could pose issues for some students, as they stated their biggest problem was the additional time it added on (Akçayir and Akçayir, 2018, p. 341). I think this is one of the biggest areas that needs more research. Can all classrooms be flipped and students still be successful in all of them? Additionally, I think more research could be done in a K-12 setting. It seemed like more was headed this way, but a majority of studies I found focused on the university setting. Is there something developmentally that makes it harder for K-12 students (at any level) to be successful in a flipped classroom? These studies could also maybe help illuminate the murkiness of the results I found, especially as discussed in review articles like Akçayir and Akçayir (2018), who reported on both the positives and negatives of flipped classrooms or self-regulated learning. 

Perhaps, some of these challenges could be overcome with increased or better usage of digital technologies. For instance, Skulmowski and Xu (2022) discuss how gamification in learning can increase motivation in students (p. 179). Overall, they found that interactive digital technology in general increases students' learning (Skulmowski & Xu, 2022, p. 180). However, too much interactivity can also cause cognitive overload (Skulmowski & Xu, 2022, p. 180). Therefore, limiting the technology to low or middle levels of interactivity is optimal for peak learning (Skulmowski & Xu, 2022, p. 180). Additionally, the use of AI is being studied more and more within the context of the flipped classroom. In a study by Huang, Lu, and Yang (2023), they used AI to give personalized video recommendations to the test students; the control students were still in a flipped classroom, just with no AI video recommendations. They found that there was increased engagement because of the personalized videos, which increased motivation in some students (Huang et al., 2023). Interestingly, the students who were most affected were those of medium motivation - the high motivation already had mastered the concepts and the low motivation had little willingness to engage with the technology (Huang et al., 2023). This suggests though that AI can be used in a flipped classroom to help improve student learning. I liked this idea of personalized videos. I think chatbots that provoke critical thinking could also be effective if used correctly.

Besides Skulmowski and Xu (2022) (to a small degree too) and Lo and Hew (2017) to a greater degree, not many studies really gave strategies on how to implement self-pacing or even the flipped classroom effectively. They repeatedly mentioned the need for scaffolding as seen above and defined both well, but very few clearly delineated how teachers achieve these successes themselves. For instance, Deng, Feng, and Shen (2023) talk about the importance of embedding questions within videos to improve flipped classrooms, even though they admit that there is little guidance for teachers - they write, "there is a paucity of evidence-based studies that inform instructors on how to improve the effectiveness of video-based flipped learning" (p. 12693). Educators can read about these successful studies and be inspired, but then they are left without many practical resources for something that requires a lot of front-loading work (with the creation of videos, activities, etc.). 

Based on the studies I looked at, I am still contemplating the use of the flipped classroom and self-pacing. If I did this in my classroom, I would do recorded lectures for them to watch at home. Based on the research I read, I would incorporate questions that were embedded in the video and more interactive elements. Then in the classroom, I would have activities that students could do at their own pace, with various check-in points to make sure they are getting the correct material. There are some who I know would do great at this, but a lot of those students are also the ones who want to ask me questions and be more engaged with the teacher's help. Unfortunately, the less motivated ones, who might struggle more with self-pacing, are less likely, at least at my school, to ask for help. My other big hesitation is the consistent refrain I saw of the time on the front end it took teachers. If I move in this direction, it would have to be next year, as I think I would want the summer to work on it. Not discussed in any of the literature I read either was the parental reaction. I bring this up because at my school, one teacher who has flipped her classroom has received some pushback from parents. While some understand and are supportive, others are not. Perhaps, that is somewhere else the research needs to go - outside of just student perceptions of these topics, but also parental perceptions and how those would affect the efficacy of teaching. It would also be worth my time to investigate specific technology that I have not experienced or do not have access to (behind an institutional paywall that the school could maybe pay for access to). Altemueller and Lindquist (2017) did provide a list of technologies that could be used in a flipped classroom, so I plan on exploring those more. 

I'm left with similar feelings I had to begin with - intrigue but trepidation. The main difference is now I have a greater understanding of various strategies to implement them. However, the effectiveness still remains elusive, especially what factors influence effectiveness. I might slowly introduce some of these ideas instead of jumping right in.