Higher education dominance and siloed knowledge: a systematic review of flipped classroom research

To provide a context for flipped classrooms as a field of interest, some basic characteristics of the sample of the 530 most-cited publications are presented. The first characteristic concerns a substantial growth of publications starting in 2011 (Fig. 2). Here, we have not taken the 16% of publications that were false positives into account, as the aim of this part of the results is to characterise the emerging field of interest in relative terms.

Figure 2 shows that the rate of academic publications concerned with flipped classrooms increased substantially between 2011 and the end of 2015. This period started with two publications in 2011, followed by 16 publications in 2012, 76 publications in 2013, 136 publications in 2014, and 296 publications in 2015. This substantial increase in the number of academic publications concerned with flipped classrooms indicates, we will argue, an emerging movement of the flipped classroom as a field of interest. This is underscored in relation to higher education, where the proportion of those publications with ‘higher education’ in the title, keywords or abstract within the overall corpus is significant. In 2012, higher education publications represented 38% of the total number, while by 2015 they represented 73%. This indicates a particularly strong growth of interest in relation to higher education and suggests that higher education has come to dominate the scientific discourse around the flipped classroom phenomenon.

Geographically, the research on flipped classroom is dominated by publications from the United States. Despite some contributions from countries such as India and Malaysia, the Global South is generally underrepresented in our corpus, and largely English-speaking countries dominate (see Table 1).

Similar distributional patterns are seen in the research review by O’Flaherty et al. (2015), but our data highlights the relatively low number of contributions from European countries. While much of the U.S. dominance in publications could be accounted for by a general Anglo-Saxon dominance in academic publishing, it should also be noted that the flipped classroom approach was first popularised in secondary education by two teachers, Bergmann and Sams (2012a, 2012b), both of whom are active within the U.S. educational context. Similarly, as the second-most prolific country, Australia’s educational context has a strong tradition of technology-based and student-centred education forms that may well set the stage for strong engagement with the flipped classroom approach.

Overall, within the corpus, conference papers are the dominant form of publication, accounting for 55% of publications with journal articles making up the remaining 45%. However, this trend appears to be shifting, with a decrease in the relative dominance of conference papers in 2014 and 2015. The early dominance of conference papers over journal articles may be partly due to the flipped classroom being a rather new field of interest, where conference papers are used as a format for sharing preliminary results. Within our corpus, the most common publication venue is the proceedings of the American Society for Engineering Education’s (ASEE’s) Annual Conference and Exposition with 84 publications. The second-most common publication venue is the proceedings of the Frontiers in Education Conference of the Institute of Electrical and Electronics Engineers (IEEE) with 33 publications. This suggests that both ASEE’s Annual Conference and Exposition and IEEE’s Frontiers in Education Conference are publication venues used by researchers and practitioners to channel preliminary results to the engineering education community. It is not until the third-most common publication venue, PRIMUS: Problems, Resources and Issues in Mathematics Undergraduate Studies with 24 publications, that a journal appears. To sum up so far, the flipped classroom as a field of interest is currently engaged with sharing experiences and methods about the flipped classroom approach by providing preliminary results in conference proceedings and in journal articles, the former of which are slightly predominant. What is interesting is that such sharing of experiences and methods seems to be carried out within specific subject fields or educational domains, predominantly higher education, along with specific subject areas, such as engineering education, computer science and mathematics education.

The majority of the publications in our corpus are associated with a particular subject area, distributed as follows: social sciences (329 publications, 62% total), computer science (199 publications, 38% total), engineering (162 publications, 31% total), medicine (47 publications, 9% total) and mathematics (35 publications, 7% total). This implies that the most-cited publications related to flipped classrooms can be located in well-defined subject areas, and as illustrated before, commonly within the STEM (science, technology, engineering, mathematics) areas.

Based on co-occurring citations, that is, the interaction of publications as they refer to each other, we can see the dominance of STEM areas again, particularly engineering, natural science and medicine. Within the data, two strong clusters are formed: one in the subject-specific higher education domain (marked in red in Fig. 3) and another representing the domains of general education and educational technology (marked in green in Fig. 3). Within the educational technology domain, three journals dominate: Computers & Education, Educational Technology & Society, and British Journal of Educational Technology. Closely following these are the Internet and Higher Education and the Australasian Journal of Educational Technology. Again, the growing importance of journal publications within this field is made visible. Generally, there is a relatively low degree of interaction within the field as a whole, with the exception of some parts of the subject-specific areas and among certain conference publication venues. This suggests that much of the research discourse on flipped classrooms is constrained to clusters and in particular subject-specific silos, most often within the STEM higher education area (Fig. 3). The source data and co-citations within these fields suggests that studies on the flipped classroom phenomenon remain relatively local in character and tend not to refer to publications outside of their respective field clusters or get cited outside of those clusters. Only parts of the subject-specific domain show a strong discourse of internally citing earlier studies. This lack of interaction between fields, visible in Fig. 3, implies that the knowledge contribution of the flipped classroom as a field of interest has yet to stabilise. Academically and socially, the research is characterised by a low degree of interaction and a significant divide between the general education/educational technology areas and subject-specific domains.

To summarise this first result section based on descriptive-quantitative data from our corpus of the 530 most-cited publications, the current state of the flipped classroom as a field of interest can be described as growing fast with a slight preference for conference proceedings and a predominance of STEM and medical contributions. Furthermore, as the field of flipped classrooms seems to be fragmented with low interactivity, it also reflects trends in research within the wider field of educational technology. As argued by Selwyn (2012), research on educational technology is constituted by ‘different tribes’, each with a ‘particular interest and motives for studying technology and education’ (p. 213). Such tribal work leaves little ‘collective impetus for making the field anything more than the sum of its parts’ (p. 213). In the context of the flipped classroom, such isolated tribes and local subject area studies are manifested in the evident lack of interaction between subject fields. The data shows that the focus of most studies on flipped classrooms is subject-specific instructional designs.

In this section, we present those 31 publications cited 15 or more times in our corpus of 530 records from step 2. First, 37 publications were selected on the basis that they had received 15 or more citations. Six of these publications were identified as false positives that did not actually address the flipped classroom approach, resulting in a selection of the 31 most-cited publications on flipped classrooms. This list of most-cited publications confirms the predominance of American researchers as 30 of the 31 records has a first author affiliated with a U.S. institution. All published by U.S.-based researchers, the top 6 publications were cited 50 or more times. These publications are presented in Table 2.

The six most-cited studies use a variety of concepts to describe the teaching strategies employed, and most of them combine the concepts in their definitions. The most common concepts used are the flipped classroom approach (Mason, Shuman, & Cook, 2013; McLaughlin et al., 2014; Prober & Khan, 2013; Strayer, 2012; Tucker, 2012) and the inverted classroom (Gannod, Burge, & Helmick, 2008; Mason et al., 2013; Strayer, 2012). Some studies specifically acknowledge active learning (Mason et al., 2013; McLaughlin et al., 2014), and one study also names blended learning (Strayer, 2012). These two concepts can also be found in a wide range of studies across the whole corpus.

In the six most-cited studies, the flipped approach is generally defined as a strategy that ‘relies on technology to introduce students to course content outside of the classroom so that students can engage that content at a deeper level inside the classroom’ (Strayer, 2012, p. 171). The core idea is to ‘flip the common instructional approach’, enabling the classroom to be a place ‘to work through problems, advance concepts, and engage in collaborative learning’ (Tucker, 2012, p. 82) by mixing ‘the use of technology with hands-on activities’ (Gannod et al., 2008, p. 777). These definitions rest on the assumption that flipped teaching alters traditional instruction so that deeper levels of learning can take place in class rather than rely on homework (Gannod et al., 2008; Mason et al., 2013; McLaughlin et al., 2014; Prober & Khan, 2013; Strayer, 2012; Tucker, 2012). The most common advantages expressed in the six most-cited publications are that the flipped classroom approach focusses student-centred and collaborative, problem-based learning activities, thus enabling teachers to spend more time identifying student problems and knowledge gaps (Mason et al., 2013; McLaughlin et al., 2014; Prober & Khan, 2013; Strayer, 2012). They also indicate that the approach allows teachers to cover more course material (Mason et al., 2013) and engage directly with students when they are involved in in-depth learning activities (Gannod et al., 2008; Prober & Khan, 2013; Strayer, 2012). Moreover, the publications suggest that the approach makes it possible to foster a ‘shared responsibility between students and instructors’ (McLaughlin et al., 2014, p. 242), and overall, optimising classroom time is often specifically emphasised (Gannod et al., 2008; Mason et al., 2013; McLaughlin et al., 2014; Tucker, 2012). The flip is usually referred to as a video clip (Gannod et al., 2008; Mason et al., 2013; Tucker, 2012), but other interactive technologies or materials such as textbooks and handouts are also mentioned (McLaughlin et al., 2014; Prober & Khan, 2013; Strayer, 2012). The main advantage of using video clips to flip the classroom is described as enabling students to watch content as many times as needed (Prober & Khan, 2013).

Most of the six most-cited publications focus specifically on the opportunities created by implementing the flipped classroom approach. However, there are some exceptions, with some results showing that students were less satisfied with how classroom structures oriented them to the learning content of the tasks in a course using the flipped classroom approach (McLaughlin et al., 2014). Even if they collaborated more, the students also reported having to adjust to different ways of orienting to the learning activities, including flips, mini-lectures, collaborative work and discussions with the teachers, whereas in the traditional classroom, the structure supporting the learning tasks was well established (Strayer, 2012). For the flipped classroom approach to be successful, it is argued that it is important to rigorously structure classroom activities and flips so that they coherently support one another. If the flip and the classroom activities are not carefully aligned, flips may instead become a barrier to the students’ learning (Strayer, 2012).

Across the six most-cited studies, the flipped classroom approach is also discussed in terms of a potential solution for teachers to respond to the increasing pressure to implement and use digital technologies in their teaching (Tucker, 2012). In one publication, the pros and cons of using video clips produced by private companies are discussed (Tucker, 2012), and the novelty of the flipped classroom approach is also questioned when it is argued that teachers have a tradition of asking students to prepare before class by, for example, doing readings (Strayer, 2012; Tucker, 2012). Instead, the novelty of the flipped classroom approach is described as ‘the regular and systematic use of interactive technology in the learning process’ (Strayer, 2012, p. 172). Furthermore, the stability of the approach is also considered in the publications as formal schooling and higher education has a long history of implementing instructional approaches that are later abandoned, indicating that ‘there’s a real danger that flipping, a seemingly simple idea that is profound in practice, may be reduced into the latest educational fad’ (Tucker, 2012, p. 83).

As Table 2 shows, 4 of the 6 most-cited publications are journal articles. Among the 31 most-cited publications, 26 of the publications are journal articles, and only 4 are conference papers. This is in contrast somewhat to the slightly stronger conference paper discourse identified in step 1 of the review. Still, the second-most-cited publication in the field of flipped classrooms is a conference note on the development of the approach (Tucker, 2012). Although conference publications have high status in some disciplines, such as computer science, there is a clear preference for citing journal articles over conference papers in the field of flipped classrooms. An explanation for this can be that a journal article is often a more recent but also a more authoritative record of a study that has been validated by a more elaborate peer-review system. Of the 26 journal articles among the 31 most-cited publications (see Appendix 1), 4 were published in the American Journal of Pharmaceutical Education (10, 15, 16, 25), 3 in Academic Medicine (4, 7, 12), 2 in Phi Delta Kappan (23, 29), 2 in Computers & Education (31, 36), 2 in Internet and Higher Education (19, 28). The remaining 13 articles were published in 13 different journals. This shows that a significant proportion of the most-cited journal articles were published in well-established, high-ranked journals. In particular, it suggests that medical and pharmaceutical education is influential as empirical contexts within the field of flipped classroom research, with educational technology journals beginning to gain some influence. This differentiation between subject-specific areas and more general and educational technology–based fields further adds credence to the perception that the field of flipped classroom research is rather siloed and fragmented and therefore still evolving.

While they are fragmented in terms of the subject area of interest, the most-cited publications are relatively homogenous when it comes to educational level of interest. Publications addressing higher education contexts make up 25 of the 31 most-cited publications, compared to only 4 publications addressing K–12 schools (13, 20, 22, 28). This is an interesting finding since the flipped classroom approach is regarded to have first emerged within K–12 educational contexts (Ash, 2012). As will be discussed in the next section, many of these higher education publications rely on local case studies and aim to improve practice and student outcomes.

This section details an epistemological-qualitative analysis to understand the knowledge foundations of flipped classrooms as a field of interest. In this analysis, we have coded the subset of the 31 most-cited publications (see Appendix 1) with regard to the types of knowledge contributions made. These relate to the following research characteristics: focus, methods, data, theory use, conclusions and the type of text contribution (note that the same publication can appear in more than one category). The results of this analysis are presented in Table 3.

As already established, 25 of the 31 records are from the higher education sector, leaving only 6 of the records from other educational areas (2, 7, 13, 20, 22, 28). The character of most of the publications is that they are locally situated in terms of their sample or case. In addition, 26 of the 31 publications (all but 2, 7, 13, 20, 22, 28) are based on teachers reporting their own practices and the context of a single higher education course or classroom experience or, similarly, a specific curricular aspect within a disciplinary subject area. The number of (mostly student) respondents in empirical studies varies but is most commonly between 20 and 40 for a smaller higher education course, with a few examples having over 500. Such small sample sizes for local case studies might not be a problem if the flipped classroom approach used in these studies was fully theorised and operationalised in terms of a specified design (e.g., of content, pedagogy, technology use, etc.). However, basic descriptions are often lacking, making high-quality empirical research characteristics rare amongst the 31 most-cited publications. Commonly, the flipped classroom approach is taken for granted as effective in improving student learning, and the experimental setting or flipped classroom approach used is not fully described. Added to this, the choice of methods is generally inappropriate for showing improved student learning or making in-depth qualitative analyses of whether learning occurred. A considerable number, 19 records, makes use of comparisons of student results or motivation changes within small sample settings. In 16 references, basic survey data, like course questionnaires or examples from students’ pre- and post-tests, are often compared over two courses or between two courses (based on simple dichotomies of traditional versus flipped approaches). This methodological trend is often accompanied with the dismissal of traditional teaching as unsuccessful with the flipped classroom positioned as a solution. Thus, an argument regularly used in the 31 most-cited publications is that the flipped classroom approach improves student learning, but this claim is mainly evidenced by improved student test results or student self-reports of increased motivation. This suggests that the mainly positive results reported in the research may be an effect of the bias of self-reported studies undertaken by teachers themselves but may also be related to the rhetorical conviction and current hype around the flipped classroom approach. Only 4 of 31 references (2, 11, 13, 28) are opinion or reflection papers, but opinion-based arguments around the flipped classroom approach are also common within the overall sample. These arguments tend to describe the flipped classroom approach uncritically, without scientific consideration of empirical design or reference to earlier research. It should be noted that rather than explicitly using and referring to educational or learning theories, most studies instead refer to a mix of pedagogical terms or strategies (e.g., active, blended, inquiry-based, problem-based, flipped, student-centred learning, etc.) without describing or theoretically distinguishing them fully. Equally, these pedagogical terms or strategies are generally not considered in relation to the context in which a study was conducted, developed or reported. Empirically and theoretically, many of the most-cited flipped classroom studies do not draw on systematic or existing resources from other research or on related fields such as educational technology. Only one study explicitly stated the theoretical aim of developing the field of flipped classrooms (29).