Selection of Studies for Detailed Review

In order to review empirical research on Lesson Study across the globe, the authors conducted searches by title and abstract in the five academic databases aforementioned, and, in some cases, using different search terms, as detailed below. At this moment, three other articles were included from other sources. The complete screening process with database searches, the number of abstracts screened, and the full texts retrieved or rejected, as well as exclusion criteria are shown in Figure 1.

The first database to be searched was ERIC and the search term was “lesson study” for all peer-viewed papers which made their full text available with no definite time. This search led to a list of 3,376 papers. However, even though many of these contained the words “lesson” and “study” in the title or abstract, an expressive number proved not to relate to the Lesson Study approach, the object of this article. Then, two distinct searches were conducted with the term “lesson study methodology” and, next, with the terms “Lesson Study” combined with “technology”, again for all peer-viewed papers which made their full text available and with no definite time. Combining these two searches, a total of 294 articles were identified.

The next database to be consulted was REDIB. Two distinct searches were conducted using the following search terms: “metodología Estudios de Clase” combined with “educación” and then, another search using the terms “Lesson Study methodology” associated with “education”. Both searches were limited to articles in scientific journals and Education and scientific Education materials with no time restriction. The total number of articles recovered was 210.

In the following, a search was carried out in the ProQuest database using the search terms “Lesson Study” AND “Education”, limited to articles that were peer-reviewed with their full texts available. Due to the high number of publications identified (227,735), the search was narrowed. The second search was conducted using the search terms “Lesson Study” (in the title) AND “Education” AND “Culture”. The search was limited to full-text articles available and peer-viewed articles between 2016 and 2020 published in English. The total of articles identified was 94.

Two distinct searches were conducted in the SciELO database. The search term “Metodologia Lesson Study” was used and no limited time nor language was set. Next, the search terms “Lesson Study” AND “Educação” were used. Combining both searches, a total of 48 articles were identified.

The search words “Lesson Study” (in the title) AND “Education” were used to conduct a search in the Emerald Insight database for articles with full access. No definite time limit nor language restriction was set. A total of 15 articles were identified.

From the databases consulted, 636 articles were identified for screening. One of the documents was not available online, two were straight duplications, and five hundred fifty-one publications were not related to Lesson Study practice. Hence, a total of five hundred fifty-four articles were excluded at this stage. In the following, the authors proceeded to a full-text screening of the eighty-five remaining publications considered for eligibility.

Selected Studies for Eligibility

The selection of the studies for final inclusion and in-depth review involved the following criteria:

To address the aims of this systematic review, and in addition to the four criteria indicated above, the authors did a full-text screening of the eighty-five articles selected in the previous stage. As a result of this process, fifty-eight documents were excluded for the following reasons:

By this method, a total of twenty-eight articles met the criteria of this study and were included in the systematic review.

*LS = Lesson Study

Source: The authors. Adapted from Moher, Liberati, Tetzlaff and Altman (2009)

Figure 1 PRISMA flowchart of screening process with database searches, the number of abstracts screened and the full texts retrieved 

RQ1. How Has LS Been Implemented Around the Globe and What Characteristics Does it Present?

Since LS was brought to public attention in the late 1990s (^{LEWIS; PERRY; MURATA, 2006}; ^{TAKAHASHI; McDOUGAL, 2016}), this approach has reached various countries and all continents, such as Asia, Africa (^{FUJII, 2014}), Europe, in particular (^{DUDLEY et al., 2019}), and it has been increasingly investigated in the United States (^{LEWIS; PERRY; HURD, 2004}). In his editorial review article, ^{Elliott (2016}) acknowledges the International Journal for Lesson and Learning Studies for having published a rich resource of material involving the European, Asian, and North American contexts in various educational levels and across different school subjects and theories. This is consistent with the findings in this study that reveal that these regions alone concentrate more than 82% of the studies investigated, with a special remark on South America, where only one country was reported to have carried out one research, Chile.

Being the only country representing South America in this review, it is noted that the connection between Chile and LS did not happen unexpectedly. The implementation of LS in that country is the result of a program established between Chile and Japan, funded by JICA (Japan International Cooperation Agency), during three years (2006 to 2008) in which a total of three groups of 10 mathematics teachers belonging to eleven universities involved in initial and continuing training of teachers, and two members of the Ministry of Education participated in intensive courses in Japan for training in Japanese methodologies of LS (^{RAMOS-RODRÍGUEZ; MARTÍNEZ; PONTE, 2017}; ^{OLFOS; ISODA; ESTRELLA, 2020}).

After 15 years of the initial investment in the educational system in Chile, and despite having faced political interference from the government in the past, today the country has maintained the LS approach nationwide, its researchers have spread it by publishing books and through participation in congresses in Latin America, and during ten years, LS has been the theme of a doctoral program in the country. ^{Olfos, Isoda, and Estrella (2020}) confirm that meaningful improvements in ongoing professional development were observed by teachers from Chile who participated in training programs and implemented the LS model in their local institutions.

Several supports are essential if LS is to be introduced and sustained in a new country. LS has been introduced in most countries through the support of funding from the government or research agencies (^{FANG; WANG, 2021}). Therefore, it is critical that the administration at local, regional, and national levels should share their responsibilities to provide time and financial support “to enable teachers to collaboratively plan, teach, collect evidence, and revise lessons” (STIGLER; ^{HIEBERT, 2016}, p. 585). Lack of shared time, workload, and rigid schedules are some of the constraints identified in this review that could result in weak collaboration between the LS team members and, consequently, follow poorer results (see, for example, ^{HART, 2009}; ^{MON; DALI; SAM, 2016}; ^{ALAMRI, 2020}).

Although the goal of traditional Konaikenshu (school-based teacher training program), of which LS is a masterpiece, does not “exclusively focus on the development of students´ academic skills” (^{YOSHIDA, 1999}, p. 52) all the articles selected for this study pursued their goal in the context of investigating specific subject matters. Of the 28 identified articles, 19 (67.85%) aimed to implement LS involving mathematics. This is in line with what is observed in other recent international reviews, in which research referring to mathematics far outnumbers the publications when compared to other school subjects (e.g. ^{KANELLOPOULOU; DARRA, 2019}). This finding could be attributed to the fact that LS was introduced in the west through research involving mathematics presented at the Third International Mathematics and Science Study in 1999 in the United States (^{LEWIS; PERRY; MURATA, 2006}; ^{GROVE; DOIG, 2010}) in addition to the fact that American students presented a low performance in the Trends in International Mathematics and Science Study (^{SAITO, 2012}). These facts generated subsequent interest in addressing this problem. Similarly, LS was introduced in Chile through the agreement established in the collaboration program "Improving mathematical education in Chile supported by Japan” in 2005 (^{OLFOS; ISODA; ESTRELLA, 2020}). As LS could be recognized as an “organic system” (^{FUJII, 2014}) in which all teachers should work collaboratively driven by the same general purpose (STIGLER; ^{HIEBERT, 2016}), it is crucial that all school subjects are involved and specifically investigated through the LS approach so that it would produce more significant and sustained results. In compliance with ^{Yoshida (1999}), we argue that the impact of scant and isolated groups of teachers on students´ learning is limited.

LS is acknowledged for being a powerful professional development model that can foster teacher learning by bringing teaching close to practice (^{SUH; FULGINITI, 2012}; ^{KOTELAWALA, 2012}; ^{CAJKLER et al., 2013}, ^{DUDLEY et al., 2019}). However, it seems that traditionally, in Japan, most elementary and middle schools take part in LS, being less frequent in Japanese high schools (^{GROVES; DOIG, 2010}) and in higher education. In this same direction, data in this study reveal that most research, 12 of the 28 (42.85%), took place in elementary and middle school settings, while it was less investigated at the undergraduate level (25%). In line with this, another finding points out that LS was examined in the context of initial teacher education in only 21.43% of the studies, in contrast with in-service settings (78.57%).

The literature confirms that, successfully integrated, LS is central for prospective teacher learning (^{SOTO et al., 2019}, ^{HANDAYANI; TRYANTO, 2022}) as it “encourages team members to bridge the gap between theory and practice” (^{SCHIPPER et al., 2020}, p. 362). For example, ^{Suh and Fulginiti (2012}) discussed the challenge of providing the opportunity for pre-service teachers to experience the complexity of teaching and, at the same time, for experienced teachers to plan lessons along with novice teachers situated in practice. This positive experience allowed pre-service teachers a safe and authentic environment to lessen their anxiety, develop specific pedagogical knowledge and reflect on their practice.

Correspondingly, ^{Danday and Monterola (2019}), reported that an experimental group of pre-service teachers benefited from experiencing microteaching multiple-representation LS and demonstrated better effects on the enhancement of critical thinking. Both studies concluded that feedback from experts, collaboration among fellow pre-service teachers and students, and length of preparation and teaching influenced the positive results perceived. Therefore, we argue that LS is a prominent vehicle that should be introduced to teacher candidates beyond their preparation program to instill in them the lifelong commitment to reflective and collaborative practice necessary for their engagement in professional development.

In terms of the specific procedures and characteristics implemented to the LS model, 16 studies (57.14%) focused their evaluation on the process of LS, 8 studies (28.57%) focused on both the process and the outcome, 4 studies (14.29%) focused only on the outcome and 4 studies (14.29%) did not specify this aspect. This finding may be explained by the fact that the literature unanimously views LS as a complex process that focuses on the student, teacher, and organizational learning (e.g. ^{FUJII, 2014}; STIGLER; ^{HIEBERT, 2016}; ^{DUDLEY et al., 2019}), and much more than the “final product”, what matters is the process. This result might be connected with the controversial inclusion of teaching a revised lesson in the LS cycle.

In this review, 17 studies (60.71%) reported having refined and re-taught a research lesson, 7 studies (25%) did not use this resource and 4 studies (14.29%) did not specify it. Some experts are emphatic saying that teaching a revised lesson is not a common practice in LS procedures in Japan (^{YOSHIDA, 1999}; STIGLER; ^{HIEBERT, 2016}). Such a strategy is even ‘disrespectful’, as ^{Fujii (2014}) highlights, not only because a research lesson is unique and designed for a particular group of students, but also because students deserve the best and cannot be treated as scientific experiments. LS cannot be reduced to a means of producing more polished lessons (^{LEWIS, 2009}), since it is much more than the lesson itself (LEWIS et al., 2006). Rather, trying to obtain a perfect lesson is to deviate from its main purpose (^{TAKAHASHI; McDOUGAL, 2016}; ^{ELLIOTT, 2019}). Nevertheless, the overwhelming support for a high number of researchers conducting a revised lesson in this review comes from adaptions adopted in settings outside Japan. Chinese Lesson Study, for example, requires repeating the revised research lesson (^{Huang, Prince, Barlow, & Schmidt, 2017}), and in the west, ^{Dudley (2011}, 2013) advocates it as part of the LS cycle, whereas ^{Lewis (2009}) considers it optional. As Stigler and ^{Hiebert (2016}) explain, reteaching a lesson might be of crucial importance in contexts outside Japan because it unveils hidden aspects that are taken for granted in that country, but needed to be demonstrated elsewhere.

Regarding the inclusion of facilitators or knowledgeable others, 19 of the selected studies (67,86%) reported having had such a professional during the process of implementing LS, 6 studies (21.43%) did not include this professional, and 3 (10.71%) did not specify this information. As for the team members, 12 studies (42.86%) indicated that the professionals involved in the LS model did not have any experience with such an approach before, and the same number of studies did not specify it. As discussed in ^{Bocala (2015}), both novices and experts should work together in teacher learning models, since more experienced teachers and knowledgeable others can encourage and give support to novice participants and provide essential expertise to the whole process. Knowledgeable others have been defined as someone, usually outside of the planning team, who plays an important role in contributing during the planning process and the post-lesson discussion (^{DUDLEY, 2014}; ^{FUJII, 2016}; ^{TAKAHASHI; McDOUGAL, 2016}).

According to ^{Takahashi (2014}, p. 10), knowledgeable others are responsible for “(1) bringing new knowledge from research and the curriculum; (2) showing the connection between the theory and the practice; and (3) helping others learn how to reflect on teaching and learning.” In this respect, ^{Yoshida (1999}) informed that all the surveyed schools reported in his ethnographic study in Japan included an outside-of-school adviser as part of their organizational chart. Especially in countries where LS is still not embedded in their culture, these professionals are considered of high relevance (^{SEINO; FOSTER, 2020}), for “without them, knowledge and skills would remain empty concepts devoid of real application and use” (^{MEWALD; MÜRWALD-SHEIFINGER, 2019}, p. 229).

Being often acknowledged as a “teacher-led” collaborative research designed to develop teaching (^{LEWIS, 2002}; TAMURA; ^{UESUGI, 2020}), LS is mostly embedded in teachers´ cultural perspectives. This may explain why much less attention has been given to how students engage, perceive, and experience their learning. In the same vein, the results in this investigation indicated that only 3 studies (10,71%) demonstrated to have investigated the learners´ perspective, in contrast with 21 studies (75%) that looked at the teachers´ perceptions alone, and 4 studies (14.29%) that did not make this formation explicit.

In this respect, ^{Wood and Cajkler (2016}) argue that when learning challenges and approaches to learning are identified only through the teachers´ perspective, potential barriers remain between teachers and students. The authors found that listening to their students´ perceptions of what was difficult in the lesson and what they would change about it reflected on their learning outcomes. This finding led the authors to propose a participatory approach to LS as a vehicle to provide new insights into the complex nature and process of learning involving the students. In the study conducted by ^{Khokhotva and Albizuri (2020}), teachers are also informed on how and what students need to learn through the voice of students, which brings significant implications to teachers´ transformative learning, educational beliefs change, and improvements in the lesson design. These studies provide new perspectives on having students participate in the LS process and reveal that significant impacts are recognized due to the student´s active involvement.

Concerning the duration of the LS process indicated by the selected articles, the investigations occurred from three weeks to as far as 18 months, in an average of 1.8 institutions (SD=1.81). In contrast with cultures hungry for immediate results aiming at overnight improvements in test scores (^{CAJKLER et al., 2015}), LS is practiced with a view to long-term results in Japan. In ^{Yoshida´s account (1999}), for example, the average length of School-Based Teacher Development focusing on the same topic was about four years. ^{Takahashi and McDougal (2016}) warned about the decision of a school district of organizing the complete LS cycle on one single day and brought questions to whether this model could still be considered LS. As mentioned before, ^{Lewis, Perry, and Hurd (2004}) pointed out that LS favors a stronger connection between daily practice and the long-term goals the teachers pursued simultaneously. Although scholars vary in developing research in a different range of time lengths, what has unanimously agreed is that LS has the high potential to produce a long-term impact on teachers’ practice and substantial positive outcomes on their student´s academic success if sustained appropriately.

Among the pedagogical practices identified in the selected publications, professional collaborative practice is the one most used by the investigators (50%). More variation was identified concerning the learning or teaching theory associated with LS practices. Of the 28 articles, 23 different related theories were identified. This might be because LS is a flexible approach with room to be shaped according to cultural and organizational settings to meet the needs of teachers and students effectively (^{SCHIPPER et al., 2020}).

Such a flexible model, led ^{Wood (2020}) to discuss the dimension of variation concerning the theories described by articles recently published in the International Journal for Lesson and Learning Studies. She underscored the importance of research instruction “being explicit about the theory of learning informing lesson and learning studies and how its implementation leads to teacher learning” (^{Wood, 2020}, p. 93) and raised the question of which theory is the one to underpin Japanese LS. The concern with theoretical approaches associated with LS had already been discussed by ^{Norwich (2018}). The author examined variations within LS practices and their connections with related professional learning and teaching traditions. Norwich concluded that the wide variation of practices identified in his investigation makes it difficult to establish which features are necessary and sufficient “to distinguish LS practices from the other non-LS professional learning practices” (Norwich, 2018, p. 213) and suggested that a polythetic type of definition of professional learning and development practices might be useful to protect labels from being excessively used and deviated from its proper use.

RQ. What Cultural Aspects or Local Conditions May Have Influenced LS Outcomes?

The success credited to LS in its homeland is mostly a reflection of Japan´s cultural and social conditions and particular school system. Experts unanimously agree that, as teaching is a complex social activity, adaptations to LS procedures will be necessary to fit local needs and contexts (STIGLER; ^{HIEBERT, 2016}; ^{SCHIPPER et al., 2019}; ^{ADLER; MWADZAANGATI; TAKKER, 2023}). Pursuing this understanding, the above analysis reveals several aspects that should be taken into consideration for the effective implementation of LS outside Japan.

Time Constraints

Lack of shared time for the team to collaboratively conduct all the necessary activities to develop the LS model properly may simply refrain the implementation of this approach from the start. LS demands teachers´ involvement (^{LIM-RATNAM et al., 2019}) and, not surprisingly, more time for lesson planning than ordinary class preparation (^{CERBIN; KOPP, 2006}). “Protect their time”, advises ^{Dudley (2011}, p. 7). Or how else can teachers be highly committed to the improvement of their professional development if they are required to stay a hundred percent of their working hours teaching? How can they volunteer if they need to cope with 2 or 3 jobs and work 10 hours a day to increase their income and make ends meet every month? The factor of time seems to be a key component to be discussed and overcome if LS is to be sustained.

Financial Support

Connected to the issue mentioned above, providing financial support for teachers and research projects is crucial to enable the implementation and sustainability of this approach. ^{Takahashi and McDougal (2016}) confirm that the provision of adequate time for teachers to participate in the LS model through the use of funds is an important catalyst for LS implementation. Even when funding is not allocated officially, it is raised through other avenues to guarantee this approach is developed and spread (see, for example, ^{LEWIS; PERRY; HURD, 2009}).

Search for Quick Results

A quick result-oriented culture might not welcome LS practices as easily as they seek for better-examining results to happen overnight (^{CAJKLER et al., 2015}). In the LS movement, long-term goals are aligned with teachers´ daily practice to build strong collaboration and shared results. Therefore, rushing the process will not bring the expected results. In this sense, Stigler and ^{Hiebert (2016}) point out the reason why LS might not work in the North American context: “we underestimate the time and effort it will take to actually implement the idea in an effective way” (p. 586). In the present review article, the authors claim that these reasons identified by Stigler and Hiebert are the highest challenges that countries new to LS face in getting funding and collaboration from teachers, organizations, and public policies.

Choice for a Suitable Learning/Teaching Theory

This study has revealed that LS has been investigated alongside a variety of learning and teaching theories. This result might be related to the fact that LS is a flexible approach (^{SCHIPPER et al., 2020}) and, as indicated by ^{Norwich (2018}), LS and LS-related practices might be developed by adopting aspects from different traditions. What is essential, however, is that Japanese scholars make this model more explicit to the international academy (^{SAITO, 2012}) so that educators can choose an adequate theory to support their investigation in their local contexts. The question raised by ^{Wood (2020}, p. 98) as to “what is the learning theory that underpins JLS” (Japanese Lesson Study) might progress to “what are the adequate theories of learning that underpin LS?”

Inclusion of Knowledgeable Others

The inclusion of knowledgeable others in the LS model has been acknowledged as beneficial to all contexts, particularly the ones new to this process. While examining the projects that emerged in North American settings, ^{Takahashi and McDougal (2016}) considered that knowledgeable others’ contributions are essential to effective LS development not only during the post-lesson discussion but also during the planning process. In addition, the authors argued that, ideally, two different knowledgeable others should perform in each phase. In one respect, it is argued that the expertise of a knowledgeable other is not acquired through formal training, rather, it is developed through the participation in LS activities with colleagues and observation of a great number of final comments being given by peers and experts (^{Takahashi, 2014}). In settings where LS is not a common practice, the author recognizes that opportunities to develop due expertise are rare, however, contributions from knowledgeable others are even more necessary in these settings. To provide insight into the theme, Takahashi provides three steps that an inexperienced knowledgeable other might follow to ensure effective final comments. What is agreed unanimously is that research teams new to LS must not underestimate the importance of these professionals, in addition, researchers and organizations should collaborate to provide means for them to learn and develop this ability.

RQ3. What lessons inform future researchers or projects as to the necessary adaptations that should be made to develop a successful LS model?

Experts from Japan and China are becoming “more explicit about the theory behind lesson study, concerning both its goals and the methods for achieving them” (STIGLER; ^{HIEBERT, 2016}, p. 582). This is an increasing demand from foreign researchers who are focusing on understanding how LS occurs in its natural environment so that they can adapt it from there. The good news is that the LS process may not be intended to be used the same way everywhere. It is expected that LS differs across contexts (^{LEWIS; PERRY; MURATA, 2006}).

This study provides evidence that LS has been implemented around the globe and produced many positive results. It has demonstrated a variety of characteristics in the procedures that resulted in benefits for the students, teachers, and other stakeholders involved in the LS process. Such a flexible model permitted the adoption of a varied number of theories that sought to investigate the effects of LS across different subjects and school levels. It was also demonstrated that researchers use different strategies and characteristics of LS to suit the local needs and necessities.

In summary, observing and being faithful to the core aspects of LS is essential if it is to be implemented as such. However, the findings in this study reveal that several negotiable procedures can be reviewed in contexts beyond Japan to make it viable. It might have come the moment each cultural setting will have to decide which non-defining aspects will have to be revised and how LS will best suit their needs and conditions. What is certain, though, is that there are no shortcuts if we intend to point out a way to transform the education system on individual, regional, national, or global levels.